Review of mobile wholesale voice call terminationmarkets

EU Market Review

15 May 2003

2

Contents

Summary 4

Chapter 1 Introduction 7

Chapter 2 The services considered in this review 12

Chapter 3 Definition of the relevant market 15

Chapter 4 Assessment of significant market power 42

Chapter 5 Detrimental effects arising from SMP in the

termination market 51

Chapter 6 Regulatory option appraisal and proposed remedies 59

Chapter 7 Proposed charge controls 97

Chapter 8 Consultation details 114

Annex A Notifications of proposals and proposedsignificant market power conditions of entitlement 116

Annex B Assessment of market power 169

Annex C Alternative solutions 176

Annex D Cost of Capital 189

Annex E Explanation of LRIC and LRIC+ target charge 199

Annex F Evaluation of surcharge for the network externality 209

3

Annex G Fair target charge and the level of charge control 218

Annex H Charge cap mechanism 222

Annex I Treatment of ported numbers 227

Annex J List of questions for consultation 231

Annex K Glossary 233

4

Summary

A new regulatory regime

S.1 A new regulatory framework for electronic communications networks andservices will enter into force in the UK on 25 July 2003. The basis for the newregulatory framework is five new EU Communications Directives that are designedto create harmonised regulation across Europe and aimed at reducing entrybarriers and fostering prospects for effective competition to the benefit ofconsumers.

S.2 The new Directives require National Regulatory Authorities (NRAs), such asOftel, to carry out reviews of competition in communications markets, to ensurethat regulation remains appropriate in the light of changing market conditions.

Markets considered in this review

S.3 The markets considered in this review are those in which services are soldand purchased by communications providers in order to complete end-to-endcalls. The public electronic communications network (PECN) provider needs topurchase services to complete end-to-end calls when calls leave its network andare then terminated on another PECN provider’s network. These are calltermination services. This particular review considers wholesale voice callsterminated on individual mobile networks. However, through its roamingagreement, 3 provides 2G termination services to its subscribers via anotheroperator’s network. Therefore, in respect of 3, this review also considerswholesale 2G voice call termination provided to the subscribers of 3. Marketdefinition is discussed in detail in Chapter 3. This product market definition isbroadly consistent with the European Commission's Recommendation.

S.4 Definition of the retail market and assessment of retail market power are notincluded because they are not directly relevant to this review. Interactions with,and effects of regulation of wholesale termination on the retail market are,however, considered as part of this review.

Initial conclusions on Significant Market Power

S.5 The Director General of Telecommunications (the 'Director') has consideredwhether significant market power (SMP) is held by any communications providersin the relevant markets, taking into account the European Commission’sGuidelines on market analysis and the assessment of significant market powerand Oftel’s SMP Guidelines (available at:www.oftel.gov.uk/publications/about_oftel/2002/smpg0802).

S.6 As a result of this, the Director’s initial view is that each mobile networkoperator ("MNO") in the UK has significant market power in a separate market forvoice call termination on its network, and in the case of 3 wholesale 2G voice call

5

termination provided to its subscribers, and can be expected to have market powerfor at least the next three years. The key factors informing this initial view are:

• the existence of the calling party pays ("CPP") arrangements and mobile users’low sensitivity to the price of incoming calls;

• each MNO has a 100% share of the relevant market and is thus a monopolist;• the purchasers of mobile termination lack sufficient countervailing buyer power

to constrain charges to the competitive level; and• termination charges on 2G networks having been persistently and significantly

above costs, which demonstrates the ability and incentive of MNOs to setprices above the competitive level.

Regulatory remedies

S.7 The Director considers it appropriate to propose a number of remedies on allMNOs terminating voice calls from fixed or mobile calls on their network.

S.8 In summary the Director’s proposals are:

S.9 In respect of Vodafone, O2, T-mobile and Orange for their 2G call terminationservices, to require that they:

a) provide network access for the purposes of 2G call termination;b) not unduly discriminate in the provision of such access;c) publish a Reference Offer;d) give prior notification of price changes; ande) reduce termination charges in line with proposed charge controls.

S.10 In respect of Inquam, to require that it publish prices and give priornotification of price changes.

S.11 In respect of 3, and the 2G voice call termination services that it re-sells, arequirement that it sets charges on the basis of forward-looking long runincremental costs of providing that service.

S.12 The Director also proposes that there should be no ex-ante regulation of 3Gvoice call termination services.

S.13 In respect of the charge control proposed by the Director, 2G voicetermination provided by Vodafone, O2, Orange and T-Mobile should be subject toan RPI-X charge control, to last until 2006 (details of the proposed control can befound in Chapter 7 and Annex G). This should not be unfamiliar to any networkproviders as it broadly follows proposals already set out by the Director in hisstatement Review of the charge controls on calls to mobiles in September 2001. Italso follows the findings and recommendations of the Competition Commission,Report on references under section 13 of the Telecommunications Act 1984 on the

6

charges made by Vodafone, O2, Orange and T-Mobile for terminating calls fromfixed and mobile networks, of January 2003.

S.14 The Director believes that his proposals are consistent with the newregulatory framework and are a proportionate response to all MNOs’ SMP in theprovision of wholesale mobile voice call termination.

Consultation

S.15 The Director is seeking comments on his proposals by 24 July 2003. Oncehe has considered all responses, the Director intends to publish his finalproposals. Comments should be sent to Lawrence Knight, Oftel, 50 Ludgate Hill,London EC4M 7JJ or sent to lawrence.knight@oftel.gov.uk

S.16 The formal draft notification of the proposed market definition, SMPdesignation and specific conditions can be found in Annex A.

7

Chapter 1

Introduction

A new regulatory regime

1.1 A new regulatory framework for electronic communications networks andservices will enter into force in the UK on 25 July 2003. The basis for the newregulatory framework is five new EU Communications Directives ("the Directives")as follows:

• the Framework Directive (Directive 2002/21/EC);• the Access Directive (Directive 2002/19/EC);• the Authorisation Directive (Directive 2002/20/EC);• the Universal Service Directive (Directive 2002/22/EC); and• the Privacy Directive (Directive 2002/58/EC).

1.2 The new regulatory framework is designed to create harmonised regulationacross Europe and aimed at reducing entry barriers and fostering prospects foreffective competition to the benefit of consumers.

1.3 The Framework Directive provides the overall structure for the new regulatoryregime and sets out fundamental rules and objectives which read across all thenew directives. Article 8 of the Framework Directive sets out three key policyobjectives which have been taken into account as relevant in the preparation ofthis consultation document, namely promotion of competition, development of theinternal market and the promotion of the interests of the citizens of the EuropeanUnion. The Authorisation Directive establishes a new system whereby any personwill be generally authorised to provide communications services and/or networkswithout prior approval. The general authorisation replaces the existing licensingregime. The Universal Service Directive defines a basic set of services that mustbe provided to end-users. The Access and Interconnection Directive sets out theterms on which providers may access each others’ networks and services with aview to providing publicly available electronic communications services. Thesefour Directives must be implemented in the UK in other EU Member States on 25July 2003. The fifth Directive on Privacy establishes users’ rights with regard tothe privacy of their communications. This Directive was adopted slightly later thanthe other four Directives and has an implementation date of 31 October 2003.

Implementation

Communications Bill

1.4 In the UK, it is intended to implement the four main Directives through a newCommunications Act. The Communications Bill was introduced into the House ofCommons on 19 November 2002 and is available atwww.communicationsbill.gov.uk. The latest version of the Communications Bill is

8

that which completed its passage through the House of Commons on 4 March2003 and was introduced into the House of Lords on the following day. It can befound at that website. References to the Communications Bill in this document arereferences to that version of the Bill. The Bill may continue to be subject to changeas it proceeds through Parliament.

1.5 It is intended that the Communications Bill will receive royal assent by 25 July2003. However, in the event that the Communications Bill does not receive royalassent by 25 July 2003, the Government has acknowledged that implementationwill need to occur by Statutory Instruments made under the EuropeanCommunities Act 1972 for an interim period until the Bill enters into force. Further,if the Communications Bill does receive royal assent by 25 July 2003, it isexpected that OFCOM will not be ready by the summer to assume all of its dutiesforeseen by the Communications Bill. Should that be the case, theCommunications Bill makes specific provision to enable Ofcom’s functions to becarried out by the Director for a transitional period. For these reasons, thisdocument refers to the Director rather than Ofcom.

EC Directives

1.6 Article 16 of the Framework Directive provides that analysis of the relevantmarkets for the purpose of the new regime should be carried out as soon aspossible after the adoption of the European Commission’s Recommendation onmarkets ("the EU Recommendation"). The EU Recommendation was adopted on11 February 2003. This market review is the preparatory work required as to theanalysis of markets required by the new regime. In the UK, the ElectronicCommunications (Market Analysis) Regulations 2003 (SI 2003/330) (“theRegulations”) confirm the Director’s ability to carry out this preparatory work.Pursuant to the Regulations, Oftel is consulting on the market identification,designation of SMP (or not) and the SMP conditions proposed.

1.7 Oftel currently regulates the industry through conditions attached to licences.The existing licensing regime must end on 24 July 2003. Oftel is currentlyconsulting on a set of draft General Conditions for those obligations which willapply from 25 July 2003 to all persons providing electronic communicationsnetworks or services. However, new specific obligations arising from SMP canonly be imposed after 24 July where National Regulatory Authorities (NRAs) haveidentified operators as having significant market power in discrete marketsfollowing a market review. Before completing such reviews and imposing suchobligations, NRAs are required to conduct national consultations (under Art. 6 ofthe Framework Directive) and to notify the EC and other NRAs (under Art 7)allowing a minimum of a month for any representations. Certain access relatedconditions set under clause 42(5) of the Bill are also subject to this consultationprocess.

1.8 Oftel had planned to complete the market reviews in summer 2003, includingthe required notifications to be in a position to implement the full new regulatoryregime including all the specific conditions after 25 July 2003. However, at the

9

CoCom meeting of 9 April 2003, the EC advised, contrary to earlier indications,that NRAs could not make notifications of the outcome of market analyses to themor other NRAs before 25 July 2003.

1.9 Oftel has now considered the implications of the EC’s views. It means thatspecific obligations (particularly those arising from market analyses or in the formof certain types of Access Related conditions) cannot be imposed until thenotification period has expired, ie not before the end of August at the earliest. Inorder to avoid a regulatory lacuna, this means that existing conditions in theseareas will need to be carried over post July 2003 until this notification process iscomplete. The carry-over of such conditions in these circumstances is compatiblewith the EC Directives and provision for this purpose is included in Schedule 18 ofthe Communications Bill (amendments to which were tabled on 30 April 2003 inthe House of Lords to make such provision).

1.10 Oftel has set up a project to identify and advise operators of conditions andlinked decisions such as determinations, consents etc, which need to be continueduntil the market analyses and necessary consultations are complete. Furtherinformation on the procedure to be put in place will be advised to the industry indue course. Any queries should be addressed to the project manager, PeterDavies (peter.davies@oftel.gov.uk).

1.11 As a result of the EC’s revised view on the timing of notifications, Oftel nowproposes to make its first formal notifications under Arts. 6 and 7 of the FrameworkDirective after 25 July 2003 and then allowing a period of a month for response.Taking into account the time for Oftel to consider the representations made andprepare a final statement, the earliest Oftel expects to introduce new obligationsarising from the market reviews will be September 2003. However the EC can, inlimited circumstances, require a further two months to respond. If they invoke sucha requirement for a particular review, then the completion of that review will befurther delayed accordingly.

Market reviews

1.12 The new Directives include the requirement that national regulatoryauthorities (NRAs), such as Oftel, should carry out reviews of competition incommunications markets, to ensure that regulation remains proportionate in thelight of changing market conditions. Oftel already carries out market reviews aspart of its long term strategy, focusing on effective competition as the best meansto deliver a good deal for consumers.

1.13 As discussed above, Oftel is now conducting a series of market reviewsunder the Regulations. The timing of this review is subject to the issues set out inparagraphs 1.4 – 1.11 above.

1.14 Each market review has three stages:

10

• definition of the relevant market or markets;• assessment of competition in each market, in particular whether any

companies have Significant Market Power (SMP) in a given market; and• assessment of the options for regulation and proposal of appropriate

regulatory obligations where there has been a finding of SMP.

1.15 More detailed requirements and guidance concerning the conduct of marketreviews are provided in the Directives, the Communications Bill, the Regulations,and in additional documents issued by the European Commission and Oftel. Asrequired by the new regime, Oftel will take the utmost account of the two EuropeanCommission documents discussed in paragraphs 1.11 and 1.12 of this marketreview.

Recommendation on relevant product and service markets

1.16 The EU Recommendation has identified a set of product and service marketswithin the electronic communications sector, which are to be reviewed by NRAs.The EU Recommendation seeks to promote harmonisation across the EuropeanCommunity by ensuring that the same product and service markets are subject toa market analysis in all Member States. However, NRAs are able to regulatemarkets that differ from those identified in the EU Recommendation where this isjustified by national circumstances and where the Commission does not raise anyobjections. Accordingly, NRAs are to define relevant markets appropriate tonational circumstances, taking the utmost account of the product markets listed inthe EU Recommendation.

Significant Market Power Guidelines

1.17 The European Commission has also issued guidelines on market analysisand the assessment of SMP (“the SMP Guidelines”). Oftel has produced additionalguidelines on the criteria to assess effective competition, which can be found athttp://www.oftel.gov.uk/publications/about_oftel/2002/smpg0802.htm. Thesesupplement the SMP Guidelines and replace Oftel’s effective competitionguidelines issued in August 2000.

Regulatory option appraisal

1.18 When considering the appropriate level of regulation if a finding of SMP isfound, Oftel will also give consideration to its regulatory option appraisal guidelinespublished in June 2002. These can be found at:http://www.oftel.gov.uk/publications/about_oftel/2002/roa0602.htm.

1.19 As explained in paragraph 1.14, there are three distinct stages to thisconsultation document. First, in chapter 3, Oftel defines the relevant market(s)considered in this review. Secondly, in chapter 4, Oftel assesses whether anynetwork provider has SMP in the relevant market or markets. As a result of Oftel’s

11

conclusions in chapter 4, chapters 5, 6 and 7 consider regulatory options andremedies for the Mobile Network Operators (MNOs).

Notification

1.20 Annex A contains the formal notification of the proposals made by theDirector as a result of the review, including the market(s) defined, the designationof SMP and the conditions proposed as a result of the market analysis. Thisimplements the provisions of Regulation 5 and 6 of the Regulations.

1.21 This document, including the formal notification in Annex A, has been madeaccessible to the European Commission and to the regulatory authorities in otherMember States in accordance with the scheme of the Directives.

12

Chapter 2

Services considered in this review

Details of services

2.1 In order for a customer of one public electronic communications network(PECN) to be able to speak with the customer of another such communicationsnetwork, the providers of these networks need to connect with each other.Network access services (which in this context refer to mobile voice calltermination) are those services that PECN providers need to purchase from eachother in order to enable the caller to speak with the intended recipient of the call onanother provider’s network. These services are referred to as wholesale servicesas they are sold and purchased by network providers rather than retail customers.Retail customers purchase retail services and these may involve use of a numberof wholesale services and one or more network.

2.2 If wholesale mobile voice call termination services were not readily available,retail customers could only communicate with retail customers connected to thesame network (known as 'on-net' calling). This would distort competition throughreduction in consumer choice and welfare. In order to maximise welfare, networkproviders need to purchase voice call termination from other network providers sothat their customers can contact the customers of the other network provider.

2.3 In the case of wholesale mobile voice call termination, this concerns callsbetween different mobile networks ('off-net' calls) and calls from a fixed network toa mobile network ('fixed-to-mobile' calls). When a call is made to a mobile phone,whether from a fixed line or from a mobile on another network, the call passesfrom the originating operator (A) to the terminating operator (B). The terminatingoperator charges a fee for connecting the call to its customer – the terminationcharge. This charge is paid by the originating operator and passed on to the callerin the retail price they pay for their call.

2.4 This consultation is only considering wholesale mobile voice call termination.The following mobile services are being covered in separate Oftel market reviews,following the EU Recommendation placing them in separate markets:

Fixed ormobile operator

(A)

Mobile operator‘(B)

Interconnection

13

• access to mobile networks, enabling the provision of retail voice services /calls made by retail customers from mobiles; and

• the wholesale service of international roaming, provided by UK mobilenetworks to foreign mobile networks, so that foreign mobile customers canuse their mobiles whilst in the UK.

Existing regulation

Charge Controls

2.5 Following an investigation by the Monopolies and Mergers Commission (nowrenamed the Competition Commission - ’the CC’) in 1998, which concluded thatthe termination charges of Vodafone and Cellnet (since renamed O2) were toohigh in relation to cost and against the public interest, controls were imposed onthe termination charges of Vodafone and O2. Their termination charges wereimmediately brought down to a ceiling of 11.7 pence per minute, which was thenreduced by RPI-9% in the each of the subsequent two years until March 2002.

2.6 In 2000/2001, Oftel undertook a review of the controls on Vodafone and O2and the level of competition in mobile voice call termination in general. At the endof that review, Oftel proposed that the termination charges of all four mobileoperators – Vodafone, O2, Orange and T-Mobile (or One 2 One as it was thenknown) – be reduced by RPI-12% each year for four years until March 2006. Themobile operators rejected the proposed licence modifications to bring into forcethese controls. As the Director still believed that controls were necessary toprotect consumers, in January 2002 he asked the Competition Commission toinvestigate whether, in the absence of controls, the termination charges of the fourmobile operators would be against the public interest and, if so, whether this couldbe remedied by way of a modification to their licences.

2.7 In February 2002, the Director, with their agreement, modified the licences ofVodafone and O2 so that the existing controls of RPI-9% on their terminationcharges were ‘rolled over’ for one year to March 2003, in order to protectconsumers pending the outcome of the CC investigation.

2.8 In its December 2002 report (Reports on references under section 13 of theTelecommunications Act 1984 on the charges made by Vodafone, O2, Orangeand T-Mobile for terminating calls from fixed and mobile networks), the CCconcluded that:

• the termination charges of the four mobile operators operate against thepublic interest;

• current termination charges are 30-40% above a fair charge;• consumers pay too much for calls from fixed lines to mobiles and from one

mobile network to another;• the high cost of termination deters people from calling mobiles; and

14

• those who make more calls to mobiles, either from a fixed line or anothernetwork, unfairly subsidise other mobile owners who mainly receive calls ormake on-net calls.

2.9 The December 2002 CC report recommended that:• each MNO should reduce the level of the total termination charge by 15 per

cent in real terms before 25 July this year;• O2’s and Vodafone’s charges should be subject to further reductions of

RPI-15% between 25 July 2003 and 31 March 2004 and for each of the twosubsequent financial years to March 2006; and

• Orange’s and T-Mobile’s charges should be reduced by RPI-14% between25 July 2003 and 31 March 2004 and in each of these subsequent two timeperiods.

2.10 The CC’s report is in line with the Director's conclusions in his statement ofSeptember 2001. The Director has accepted the CC’s recommendation of a one-off cut of 15% in real terms by July. Formal licence modifications to the MNO’slicences were put out to public consultation on 28 February 2003 (available at:www.oftel.gov.uk/publications/licensing/2003/fixmob_voda0203and www.oftel.gov.uk/publications/licensing/2003/fixmob_orangetmob03). Thosemodifications were made by the Director on 4 April 2003. Arrangements for anycontrol of termination charges after July 2003 are considered in this market reviewundertaken by Oftel under the requirements of the new European Directives.

Significant Market Power

2.11 Vodafone and O2 have been are at present designated as having significantmarket power (SMP) under the EC Interconnection Directive (Directive 97/33/EC).As a consequence of that designation, Vodafone and O2 are required to enter intointerconnection agreements with other Schedule 2 operators, not to discriminateunduly in the terms of interconnection offered to other networks, to provide theDirector with price notifications 24 hours in advance of them taking effect and toforward copies of all interconnect agreements

15

Chapter 3

Market definition

Mobile voice call termination

3.1 There are two aspects to the definition of a relevant market: the relevantproducts to be included in the same market and the geographic extent of themarket. Oftel’s approach to market definition follows that used by UK competitionauthorities and is in line with those used by European and US competitionauthorities. Market boundaries are determined by identifying constraints on theprice-setting behaviour of firms. There are two main competitive constraints toconsider: how far it is possible for customers to substitute other services for thosein question (demand-side substitution); and how far suppliers could switch, orincrease, production to supply the relevant products or services (supply-sidesubstitution) following a price increase.

3.2 The concept of the ‘hypothetical monopolist test’ is a tool used to help identifydemand-side and supply-side substitutes. A product is considered to constitute aseparate market if a hypothetical monopoly supplier could impose a small butsignificant, non-transitory price increase (SSNIP) above the competitive levelwithout losing sales to such a degree as to make this unprofitable. If such a pricerise would be unprofitable, because consumers would switch to other products, orbecause suppliers of other products would begin to compete with the monopolist,then the market definition should be expanded to include the substitute products.However, the relevant market is not necessarily the smallest that it is possible todefine using the hypothetical monopolist test. It may be appropriate to includeproducts (or areas) over which there are common pricing constraints such thatthey should be included within the same relevant market even if demand andsupply-side substitution are not present.

3.3 As set out in Chapter 2, the service considered in this review is wholesalemobile voice call termination. Call termination is the service necessary for anetwork operator to connect a caller with the intended recipient of the call on adifferent network. As discussed in Chapter 2, if call termination was not availablea network operator could only terminate calls to other customers on its network.This service is referred to as wholesale because it is sold and purchased bynetwork operators rather than retail customers.

3.4 This review only looks at voice call termination on public mobile networks (henceprivate mobile networks are excluded). Fixed voice call termination is dealt with in aseparate review, Review of fixed geographic call termination markets - 17 March 2003,(available at: www.oftel.gov.uk/publications/eu_directives/2003/eu_geo_term/index).Data termination is not considered. The analysis, hence, only covers GSM, 3G andTetra networks, but does not cover public wireless local access networks (WLANs)which are devoted to data transmission. SMS termination is also excluded.

16

Calling party pays

3.5 Before considering the market definition, it is important to consider the “callingparty pays” (‘CPP’) arrangement adopted in the UK telephony market because ithas a notable impact on the boundaries of this market. Under the CPParrangement, the calling party (and not the called party) pays the total price of aretail call. This means that the voice call termination charge is included in theoriginating network provider’s (either fixed or mobile) cost base and is reflected inthe retail price it sets for calls originating on its network. CPP, thus, leads to adisconnection between the person paying for the calls and so, indirectly, for thetermination charge (i.e. the calling party) and the person who makes the choice ofthe terminating network and could thereby influence the level of the terminationcharge (i.e. the called party).

3.6 The overall effect of this arrangement in the retail market (i.e. the market forcalls to mobiles) is that, while MNOs have an incentive to keep the price of thoseservices required and paid for by the subscriber at a level to attract and retaincustomers, they have less incentive to keep the price of calls to mobiles low. Thisis because those calling a mobile subscriber cannot take their business elsewhereif dissatisfied, as they have no alternatives to terminating the calls on the networkto which the called party subscribes.

3.7 In the wholesale market, the effect of the CPP arrangement is similar. For callsfrom fixed-to-mobiles, it means that the operators running fixed public electroniccommunications networks (fixed PECNs) pay the MNOs to terminate calls on theMNOs’ networks. The MNO has little incentive to keep voice call terminationcharges low, because the fixed PECN providers will pay a high charge as theyhave a commercial interest in ensuring that all calls made by their subscribers areterminated. For off-net mobile-to-mobile calls (i.e. from one MNO’s network toanother), the MNOs pay each other for termination of calls. Again, there is littleincentive to keep termination charges low, not least since cutting them would ineffect give the MNO’s competitors an advantage in the retail market by reducingtheir costs. Hence, CPP means that an MNO is likely to be able to raise voice calltermination charges above the competitive level without suffering sufficientadverse effects to make the rise unprofitable.

3.8 Overall, the Director considers that the CPP arrangement provides the MNOswith the freedom and the incentives to set their voice call termination chargesabove the competitive level.

Product market definition

3.9 As discussed above, the relevant product market is arrived at by starting fromthe smallest potential definition. In this case, the smallest possible marketdefinition is wholesale voice call termination to a specific mobile subscriber (ornumber), as a call to another individual it is unlikely to be a sufficiently goodsubstitute for a call to a specific recipient.

17

3.10 The following sections scrutinise that definition using the ‘hypotheticalmonopolist test’ to see whether it should be expanded to include other productseither because of demand-side substitution, supply-side substitution or theexistence of common pricing constraints.

Demand-side substitution

3.11 To assess whether there are any demand-side substitutes that should beincluded in the relevant market, it is necessary to examine the effect on customers’behaviour of an increase in termination charges by the hypothetical monopolistand whether such behaviour could make the rise unprofitable.

Retail demand-side substitution

3.12 The Director considers that an increase in termination charges is likely to leadto a rise in retail call prices. Termination charges form the largest part of themarginal cost incurred by a fixed PECN provider when providing fixed-to-mobilecalls. BT’s retail prices for calls to mobile are the sum of the charge it pays to theterminating MNO and BT’s regulated retention rate (discussed in Oftel’s Review ofthe fixed narrowband wholesale exchange line, call origination, conveyance andtransit markets, consultation – published on 17 March 2003). Therefore, BTtransfers any increase in the costs of providing calls to mobiles via its retail prices.Fixed PECN providers other than BT do not have market power in the retailmarket; it can thus be expected that they would follow BT’s behaviour. As forMNOs, Oftel considers that none of them has SMP in the retail market (discussedin Oftel’s Review of competition: mobile access and call origination – April 2003).Hence, when faced with an increase in the marginal cost of providing off-net calls,it is likely that MNOs will pass at least some of this cost into their off-net retailprices.

3.13 It is, therefore, relevant to consider how retail customers (the calling and thecalled parties) would respond to a price increase in calls to mobiles engendered bya rise in voice call termination charges and whether their reaction could be asource of competitive pressure on termination charges.

Behaviour of the calling party in response to an increase in the price of callsto mobiles

3.14 If callers reacted to an increase in the retail price for calling mobiles byemploying others means of communication to reach mobile subscribers, this formof substitution could act as a competitive constraint on voice call terminationcharges, although whether it would act as constraint depends on the amount ofsubstitution that takes place. It would affect the MNOs’ behaviour only if it wasenough to make the increase in the wholesale charges unprofitable. However, forcallers to react to an increase in the price of calls to mobiles, it is the Director’sview that three conditions need to be satisfied:

18

• callers must be sufficiently aware that they are calling a mobile and that theyare calling a specific network;

• callers must be sufficiently aware of the price of calling that particular network;and

• callers must be sensitive to changes in the prices of calling the network theywant to reach, i.e. an increase in the termination charge above the competitivelevel must cause consumers to adapt their behaviour to find an alternativesatisfactory way of contacting the person they want to call.

Awareness of calling a mobile and awareness of calling a specific mobile network

3.15 The lack of consumer awareness over the identification of mobile numbersappears to be fairly high. An Oftel residential consumer survey from August 2002(Consumers’ use of fixed telecoms services – published 24 October 2002) showsthat only 46% of all residential fixed consumers are “usually” aware of calling amobile phone (there has been little change in this figure over the previous 18months). This figure is higher (52%) among consumers who own a mobile phone,but is only 32% among consumers from households who do not have mobiles(Consumers’ use of mobile telephony - published 27 January 2003). These figureshave not changed much in the last two years, even though mobile penetration hasincreased.

3.16 In compiling the CC December 2002 report, the CC collected evidence onconsumer awareness of the identity of the particular mobile network they arecalling. The CC’s own market research (paragraph 2.136 in the December 2002report) indicated that on average only 28% mobile users knew whether they werecalling a mobile phone on the same network as themselves. The results of asurvey carried out by NOP for O2, which are quoted in the CC December 2002report (paragraph 2.136 in the December 2002 report ), found that 57% of fixedphone users who also owned a mobile claimed to know the mobile network theywere calling when using their fixed line, whereas only about 30% of fixed userswithout a mobile were aware of it. When NOP put the same question to mobile-only customers, a similar level of awareness was claimed as that of the owners ofboth fixed and mobile phones.

3.17 Hence, callers appear to the Director to have limited knowledge of specificnetworks they are connecting to when making a call to a mobile.

Awareness of relative and actual prices

3.18 Despite the high penetration of mobiles, there is still evidence of lowconsumer awareness of the costs of calling mobile numbers (in general and oneach specific networks). An Oftel qualitative study of telephone users’ behaviour(Oftel Price of Calls to Mobiles Qualitative Research Findings - published March2001) shows that most fixed line users do not generally check their phone bills;and, thus, that most were unable to quantify the costs of calling either a fixed or amobile number. Specific differences in the prices of calling different mobile

19

networks from a fixed line were also unknown and respondents assumed that theywould be essentially similar. A more recent Oftel residential consumer survey(Consumers’ use of fixed telecoms services – published 24 October 2002) showsthat just 17% know approximately how much it costs to call a mobile from theirfixed phone.

3.19 As part of their 2002 investigation on calls to mobile, the CC also assessedhow knowledgeable mobile owners and fixed-line users are about the actual costsof calling mobiles and the relative costs of different call types. The CC's surveyonly covered awareness of the price of a fixed-to-mobile call because the largenumber of different tariffs available would have made it difficult to elicit reliableresponses about mobile-to-mobile prices. The main result was that only 21% ofcustomers had an approximate idea of the true cost of such a call. Evidencecollected by the CC (paragraph 2.136 to 2.141 in the December 2002 CC report)also suggests that a large number of callers to mobiles have little knowledge eitherof the actual or relative levels of prices for calling each mobile network.

3.20 An NOP survey, conducted on behalf of O2 and referred to in the December2002 report (paragraph 2.137 to 2.141 in the CC report ), tested consumers’knowledge of relative prices. This revealed that about 48% of respondentsthought that fixed-to-mobile calls were a lot more expensive than fixed-to-fixedcalls (findings which correspond to actual price relationships) and 36% said thatoff-net calls were a lot more expensive than fixed-to-fixed calls (which alsocorresponds to the actual position). Given that the difference between fixed-to-fixed and fixed-to-mobile is actually smaller than the difference between fixed-to-fixed and off-net prices, the fact that fewer respondents thought that off-net callswere a lot more expensive tends to confirm the finding that there is a lack ofwidespread awareness of the relative prices of different types of call. This surveyalso found that 50% of the consumers appear to be aware that on-net prices arelower than off-net prices and that the price of fixed-to-mobile calls varied by time ofday. In addition, 44% knew that the price of calling mobiles from fixed lines variedby the network being called.

3.21 The Director believes that even though there is some consumer awareness ofrelative prices for different types of calls, knowledge of actual prices is low, inparticular of the price of calling a specific network. The Director also notes thatmobile number portability makes it more difficult for called parties to know/identifywhich network they are calling (and thus what is the relevant price), unless theyare calling a person with whom they are in a repeated calling relationship.

Consumers’ sensitivity to changes in the prices of calling a specific network -adapting behaviour

3.22 Even if awareness of prices was high, competitive pressure will only beexerted if consumers are willing to adapt their behaviour through substitution, sothat MNOs lose profits on mobile termination if they attempt to raise terminationcharges.

20

3.23 The Director has considered a range of different possible types ofsubstitution:

• mobile to fixed calls as a substitute for off-net calls;• mobile-to-mobile calls as a substitute for fixed-to-mobile calls;• on-net mobile-to-mobile calls as a substitute for off-net calls;• SMS as a substitute for mobile-to-mobile calls;• unified messaging and message clips’;• voice over Internet Protocol calls; and• call-back arrangement.

3.24 However, as explained below, the Director considers that none of the aboveis likely to have a constraining effect on voice call termination charges.

Mobile to fixed calls as a substitute for off-net calls

3.25 Following an increase in termination charges and, thus, in the cost of callinganother mobile, the calling party may switch to calling the intended party on a fixednumber. However, it is unlikely that a call to a fixed line can represent asatisfactory substitute for calling someone on a mobile network in a sufficientnumber of instances to act as a constraint on the charges for mobile voice calltermination. In particular, a call to a fixed line is not a viable alternative if thecalled party does is on the move, since immediacy of contact is an importantfeature of calls to mobiles.

Mobile to mobile calls as a substitute for fixed-to-mobile calls

3.26 After an increase in termination charges and, thus, in the cost of callinganother mobile, the calling party may keep calling the desired party’s mobilenumber, but from a mobile phone rather than from a fixed one. The ability of thisform of substitution to constrain voice call termination charges depends on itseffect on the profits from termination services for the network operator. The priceof on-net calls does not include the termination charge and the service is designedto attract subscribers who care about inbound call prices (so that if they can be onsame network as their callers). Hence, on-net call retail prices are generally set atlow levels and generate lower revenues for the MNOs. In contrast, off-net calls aresubject to the same termination charge as calls from fixed lines. Hence,substitution may act as a constraint only if callers were to switch from fixed-to-mobile calls to on-net calls.

3.27 In the Director’s view, for on-net calls to be a viable alternative to fixed-to-mobile calls:• the caller must know the mobile network (s)he is calling; and• the caller must be on the same network as the call recipient.

3.28 It is not clear if this happens sufficiently in practice for it to act as a constraint(the only evidence available suggests that 13% of consumers with both fixed and

21

mobile phones use their mobile to make ‘on-net’ calls because it is cheaper). Asdiscussed above, awareness of the specific network called is limited (around 60%among mobile users). In addition mobile subscribers are split between four mainoperators, hence, the probability of the caller being able to reach the desired callrecipient with an on-net call is substantially less than 100%. The probability that acall remains on-net will reflect the market share of the mobile operator in theoutgoing retail market, currently between 24% and 27% on the basis of totalsubscribers. However this probability is likely to be higher, because there isevidence that some customers with repeated calling patterns tend to congregateon the same network (see section on closed user groups). On this basis, theavailability of on-net calls as a substitute for calls from fixed phones to mobiles islimited.

3.29 However, if a caller knows that (s)he makes calls to one network more oftenthan to the others, this could influence the choice of network so as to benefit fromthe lower prices charged for on-net calls. In that case, calling mobile-to-mobile on-net might be an effective substitute for a call from a fixed phone. However, thissubstitution is unlikely to constrain termination charges because MNOs designtariffs so as to separate the more price sensitive subscribers from the others (e.g.by attracting them with low on-net prices). They can, thus, put up terminationcharges for the less price sensitive subscribers. This issue is examined more indetail under the heading ‘Closed user groups’.

On-net mobile to mobile calls as a substitute for off-net calls

3.30 Termination charges for off-net calls could be constrained by substitution toon-net calls. This requires the calling party to be on the same network as thecalled party or to use more than one network to originate their calls (by havingmultiple SIM cards).

3.31 The use of multiple SIM cards appears unlikely because the process ofswitching cards (to make different calls) is laborious and time-consuming. Inaddition, MNOs often lock handsets to the SIM that is originally sold with thehandset so that it can only be used on their network. Those factors, and limitedconsumer awareness of the higher cost of off-net, (an Oftel survey, Consumers’use of mobile telephony – February 2001, found that 37% of users thought that off-net calls were ‘a lot more’ expensive, 29% thinking that they were ‘a little more’expensive and 30% considering they were ‘roughly’ the same), means that it maybe some considerable time before multiple SIM devices generate any significantcompetitive pressure on voice call termination charges. Oftel has found(Consumers' use of mobile telephony – November 2001) that, at present, while63% of mobile users were aware that it was possible to use different SIM cards tolog onto another network, only 6% of these (4% of total mobile users) usedmultiple SIM cards. Further details on multiple SIM cards can be found in Annex Calternative solutions.

3.32 A constraint may, nevertheless, be posed by groups of mobile owners whoare willing to switch networks to be on the same network and thus pay on-net

22

prices to call each other (for more details on these groups see the section onclosed user groups below). If enough subscribers switch MNOs to be on the samenetwork - the one with the lowest on-net charges - MNOs may not just loserevenues from off-net calls, but may even lose subscribers.

3.33 However, the Director considers that substitution of calls by on-net calls couldreduce, rather than increase, the constraint on the general level of voice calltermination charges. That is because MNOs, by offering lower on-net call prices,can segment the market by type of customer and separate the more price-sensitive customers from the others who are less price-sensitive. They can thenset high termination charges for others (i.e. off-net termination charges). Thus, theDirector is of the view that the nature and extent of this type of call substitution isnot sufficient to act as a competitive constraint on termination charges.

SMS as a substitute for mobile to mobile calls

3.34 Short message services (SMS) enable parties to exchange text messagesbetween mobile phones. SMS cannot act as a constraint on voice call terminationcharges, because SMS termination is offered by the same MNO which providesvoice termination. Thus, the MNO could set charges for SMS termination in sucha way as to avoid any competitive pressure on its charges for voice termination.

3.35 In addition, SMS can only be relatively short because the number ofcharacters allowed in a text message is limited to 160 characters. Furthermore,SMS are not real time because, unlike mobile voice calls, SMS are transferredbetween networks on a store and forward basis, rather than on a ‘real time’ basis.Therefore, SMS do not afford the opportunity for immediate conversation andinteraction offered by voice calls.

3.36 Evidence collected by Oftel on mobile users’ behaviour supports the aboveconclusion that SMS are unlikely to be good substitutes for voice calls. Qualitativeevidence (Oftel Price of Calls to Mobiles Qualitative Research Findings - March2001) shows that text messaging is regarded, especially by the young (who sendthe highest proportion of messages), as an activity separate from voice calls andthat SMS are largely additional to voice calls to mobile phones, rather than asubstitute. Further, a more recent Oftel survey (Consumers’ use of mobiletelephony - 27 January 2003) found that 30% of mobile users never use an SMSinstead of a mobile voice call, 20% rarely do so, 14% do so sometimes and 36%do so frequently.

Unified messaging services and ‘message clips’ as a substitute for mobile tomobile calls

3.37 Unified messaging services and ‘message clips’ (short voice or videomessage sent over the IP network) could also develop as a potential substitute forvoice calls. However, these technologies are not developed enough to bedeployed and it is difficult to assess when they will become commerciallyavailable. Moreover, the Director is of the view that, as with SMS, these services,

23

once available, are more unlikely to put any competitive pressure on voice calltermination charges, as the same MNO sets the termination charge for all theseservices.

Voice over Internet Protocol (VOIP) calls as a substitute for calls to mobiles

3.38 Voice calls to mobile phones could also potentially be delivered as VOIP callsusing general packet radio system (GPRS). GPRS is the enhancement to GSMthat allows a mobile subscriber to transmit and receive data in a packet mode.Packet mode data has the advantage that the resources of the network are onlyused when data is transmitted or received. This important enhancement allowsMNOs to charge users for the amount of data they send or receive, as opposed tothe length of time they are connected; hence it is ideal for services such as webbrowsing and connecting to the Internet. When a mobile subscriber is connectedto the Internet (is “on-line”) they could be contacted by another person, who is also"on-line", through a VOIP call. An example is the service provided by MicrosoftMessenger, which allows two parties who are on-line to set up a VOIP call bylogging on to a dedicated website (for example, an individual can go to the websitethat offers the VOIP service and check if the person they want to call is on-line. Ifthey are on-line, the server sends an alert saying that an incoming call is beingmade, they can then decide whether to take the incoming call).

3.39 VOIP calls could represent an effective substitute if the price for standardvoice calls to mobile were to rise (due to an increase in termination charges).This is due to the fact that the payment arrangement is different because on-linemobile users pay for the data they send or receive (at a per bit rate). Therefore inthe example given above, the individual making the on-line call only pays to be on-line (this does not include a mobile termination charge) and the target of the call (ifthey answer the call) similarly only pays to be connected to the website/server.Thus, in a VOIP call both the called and the calling parties pay forreceiving/making it. Such a partial Receiving Party Pays (RPP) arrangementchanges the incentives on the called party and is likely to affect their behaviour,although it is unclear in what specific manner. For example, it is possible that thecalled party does not accept VOIP calls because they would have to in part pay forthem, thus forcing the calling party to reach them via a standard voice call to theirmobile. In this case, VOIP calls would not impose competitive pressure on thelevel of the termination charges.

3.40 The constraining effect of VOIP calls may also be undermined by the MNOs'behaviour. The reason for that is that it is the MNO to which the called partysubscribes that sets both the voice termination charges and the price for theInternet connection. That MNO also determines the quality of service (i.e. datadelay and bit errors). Hence, it could adjust the GPRS quality parameters suchthat it was still acceptable for web browsing and e-mail, but not for voice calls.

3.41 A further reason why VOIP is at present unlikely to be an effective substitutefor standard voice calls to mobiles is that the called party must be on-line to be

24

contactable. Furthermore, the mobile handset must be GPRS-enabled and havethe hardware and software necessary to convert speech into data.

3.42 In conclusion, the Director considers that at present it unlikely that theavailability of VOIP calls could put competitive pressure on the level of mobilevoice call termination.

Call-back

3.43 Call-back refers to a situation where the direction of a call is ‘reversed’ andthe calling party is called back by the called party, either in an ad hoc manner orthrough a commercial scheme. Call-back could render an increase in terminationcharges unprofitable only if the profitability of outgoing calls is lower than that ofincoming calls, and call-back is carried out in sufficient volume.

3.44The Director has no evidence of any commercial operators currently offeringcall-back on calls to mobiles within the UK and found no evidence that the practiceof ad-hoc call-back is having a constraining effect on voice call terminationcharges.

3.45 It is possible that MNOs could introduce a call-back service to offer analternative to callers to their subscribers. The Director believes that this form ofcall-back could not be relied upon in the immediate future to act as a viableconstraint on mobile voice termination charges. This is partly because MNOs haveno incentive to introduce a service of a price and a quality such that it could act asan effective substitute for their own monopoly service. More details on theDirector’s view on this form of call-back are contained in Annex C on alternativesolutions.

3.46 Consequently, the Director is of the view that currently no competitivepressure on termination charges is exerted by or is likely to come from this source.

Initial conclusions on the behaviour of callers in response to an increase inthe price for calls to mobiles

3.47 On the basis of the evidence and the arguments discussed above, theDirector considers that the behaviour of callers in response to a rise in terminationcharges and, thus, in the price of calls to mobiles, is unlikely to render thisincrease unprofitable. Consumers’ low awareness of the prices of calls to mobilenetworks, the limited availability of effective substitutes and the use by the MNOsof on-net prices to separate the most price sensitive consumers suggests thatcallers are unlikely to react to a price increase. The Director has thereforereached the initial conclusion that no constraint on the level of mobile voice calltermination charges is currently likely to arise from the behaviour of callers tomobiles.

25

Behaviour of the called party in response to an increase in the price of callsto mobiles

Choice of network

3.48 There would be constraints on termination charges if mobile subscriberschose their network on the basis of the prices of incoming calls and, thus, switchednetwork as a result of an increase in these prices. However, as mentioned above,the CPP arrangement implies that the calling party, and not the called party, paysthe total price of a retail call. Therefore, the called party, who makes the choice ofthe terminating network, is not affected by the level of the prices of calls to her/him(and thus by the level of mobile termination charge of her/his network).

3.49 However, it is still possible that mobile subscribers could react to a rise in thetermination charges of the MNO they subscribe to by switching to a network withlower termination charges, if they expected that the higher price of calling themwould have an impact on their callers. For this to be true, it is the Director’s viewthat a number of conditions would have to be met:• mobile subscribers would have to value incoming calls to such an extent that a

change in the price of these calls could make them change network;• callers to mobiles should be sensitive to the price of outgoing calls;• callers to mobile should be aware of the mobile network they are calling and of

the price of calling it; and• mobile subscribers should be aware that callers to mobile are sensitive to and

have knowledge of the prices of outgoing calls.

3.50 The Director considers that only if all four of the above conditions have beenmet, would mobile subscribers consider that the level of the termination charges oftheir network affects the cost for others to call them, thereby influencing thenumber of calls they would receive. This could then act as a competitive constrainton mobile termination charges, since an MNO that increased its charges wouldsuffer a loss of users on its network.

3.51 However, the evidence collected by the Director leads to two mainconclusions:• that consumer awareness of the price of calls to mobile phones is low

especially in respect of the price of calls to each specific network; and• that the price of incoming calls is not considered by consumers to be an

important factor in their choice of a mobile network and other factors are moreinfluential.

3.52 In Oftel’s consumer survey from August 2002 (Consumers’ use of mobiletelephony - published 24 October 2002), just 46% of consumers said they “usuallyknew” they were calling a mobile phone from their fixed phone at home. AQualitative research carried out in March 2001, showed that few people had anyreal idea of the cost of calling mobiles, but it was perceived as ‘expensive’.

26

3.53 Evidence collected by the CC (paragraph 2.136 to 2.141 in the CC December2002 report) suggests that the majority of callers did not know the identity of theparticular mobile network they were calling and that a large number of callers hadlittle knowledge either of actual prices or the relative levels of charges for callingeach network. The CC concluded that these findings overall “reveal a degree ofawareness on the part of consumers which is insufficient to enable them to makean appreciable impact on prices or to drive termination charges down tocompetitive levels” (paragraph 2.141 in the CC report).

3.54 The introduction of mobile number portability has further complicated mattersby rendering it extremely difficult for callers to find out which network they arecalling (and thus what is the relevant cost), unless they have a repeated callingrelationship with the person they are calling.

3.55 Oftel survey data on residential mobile owners suggests that the choice ofwhich mobile network to subscribe to is mostly driven by the cost to subscribers(and not to those who call them) of being part of a network. A consumer surveyfrom August 2002 (Consumers’ use of mobile telephony - published 24 October2002) revealed that:• 18% of mobile customers found out how much it would cost others to call their

mobile before choosing their network (up from 15% in February 2002; and• 8% of customers considered the cost of other people calling them as a major

factor in their choice of the mobile network (it was 9% in February 2002.

3.56 For the majority of residential customers (59%), outgoing call costs were themost significant reasons for choosing a network, compared with coverage andreception (21%) (Consumers’ use of mobile telephony - February 2002).

3.57 Business users – in particular small and medium sized enterprises (‘SMEs’) -appear to be more concerned than residential users about the cost of calling theirmobiles. This might be because their business depends on calls from clients, orbecause a significant part of their own phone bill is from their employees callingmobile phones. Oftel’s research (Consumers’ use of mobile telephony - 24October 2002) shows that 27% of SMEs consider the cheapest network to callwhen choosing a mobile network (up from 19% a year before). Similar toresidential consumers, SMEs revealed that other factors, such as coverage andcustomer service, were however of greater importance to them than the cost ofincoming calls when choosing the mobile network. SMEs are a small proportion oftotal business mobile ownership.

3.58 The British Market Research Bureau (BMRB) survey commissioned by theCC also found that the cost of incoming calls was not an important factor forconsumers when choosing their mobile network. It ranked 10th out of the 14factors suggested, the most important being ‘the price you pay to call others’. Inaddition, 61% of mobile users expressed more concern about the cost to them ofcalling others than the cost to others of reaching them. Only 9% were moreconcerned about the cost to others (paragraph 2.133 to 2.135 in the CC report).

27

3.59 These findings are consistent with those of the surveys of residentialcustomers commissioned by two of the MNOs. O2’s NOP survey found that fornearly 75% of respondents the cost to other people of calling them on their mobilephone was an unimportant factor when they decided which mobile network to join.Fewer than 20% said that it was important. A high proportion (85%) of bothcategories were unable to say why they took the view they did (paragraph 2.133 to2.135 in the CC report). The NOP survey commissioned by Vodafone showedthat the price of outgoing calls was much more important to mobile users than thecosts that others incurred to reach them (paragraph 2.133 to 2.135 in the CCreport.).

3.60 The Director considers that the available evidence supports the conclusionthat an insufficient number of consumers consider the prices of incoming callswhen choosing their network for it to provide a competitive constraint on the priceof calls to mobiles (see also section on closed user groups below).

Closed user groups

3.61 The existence of ‘closed user groups’ (i.e. groups of people whose memberscare about the cost to the other members of calling their mobile number), couldmitigate the effect of the CPP arrangement and act as a constraint on voice calltermination charges. However, for this constraint to be effective, these groupsshould be numerous and not isolated through targeted tariffs that bypasstermination.

3.62 As discussed above, the evidence available shows that few groups of peopleare sensitive to the cost of incoming calls, and even those that are sensitiveconsider the cost of outgoing calls to be far more relevant. The NOP survey,conducted for Vodafone and whose results are in contained in the CC’s December2002 report (paragraph 2.116 in the CC report ), reveals that 19% of mobile usersare on the same network as members of their family because it reduced callingcosts. Another NOP survey commissioned by O2, and reported in the CC report(paragraph 2.116 in the CC report ), also found that for around one-third ofrespondents the network used by people with whom they are likely to becommunicating is an important factor when deciding which network to join.However, it also found that for over half of respondents it was unimportant and ahigh percentage of these groups could not say why this factor was, or was not,important to them.

3.63 The BMRB survey commissioned by the CC (paragraph 2.116 in the CCreport ) confirms these findings. It found that 81% of consumers had neverchosen, or even changed, mobile network in order to be on the same network asthe people to whom they often spoke, while 15% had chosen their network on thisbasis and 4% had changed it for this purpose.

3.64 The Director considers that this evidence further reinforces his initialconclusion that most residential consumers are not affected in their choice of

28

mobile network by the cost to others of calling their mobile number. Only a smallminority of residential mobile users are interested in keeping the cost of callcharges among their circle of family and friends low.

3.65 Business users seem to be more concerned about the cost of calling theirmobiles both for their employees and their customers. Oftel’s August 2002 surveyof SMEs (Business use of mobile telephony - published 24 October 2002) showsthat a fifth of the SMEs with mobile phones had taken steps to reduce the cost tocall their own mobile phones (ranging from 19% of small businesses up to 38% ofmedium businesses), such as using mobile-to-mobile adaptations or private wireservices (see below).

3.66 In addition, the survey reports that 70% of SMEs with mobile phones usesome method to reduce the cost of customers to call their mobile phones. Toachieve that, 27% claimed to have chosen the cheapest network to call whenselecting a mobile network (up from 19% a year ago). However, the survey showsthat the most popular measure to keep cost down is keeping the call to the mobileshort (47%).

3.67 Overall, SMEs seem to be more concerned than residential consumers aboutthe costs of calling their mobiles and a significant proportion appeared to beadapting their behaviour to try to minimise this cost. However, SMEs are a smallproportion of total business mobile ownership and do not provide a completepicture of business users’ behaviour. In addition, those firms also stated that thecost of outgoing calls and the coverage were their most important concerns whenchoosing the network to which their employees subscribe. So even for thesecustomers, features other than the price of incoming calls have more influence ontheir choice of the network.

3.68 In addition, the effect of closed user groups in constraining voice calltermination charges is lessened by the ability of the MNOs to separate thesecustomers from others through special self-selecting arrangements that by-passhigh termination charges (e.g. on-net calls or private wire services). In this way,those customers who are most concerned about the price of calling mobiles andwho are most able to bring pressure to bear on the MNOs are prevented fromplacing any pressure on termination charges. The MNOs are then able to imposehigh termination charges on the less price-sensitive users.

Private wire services and mobile-to-mobile adaptations

3.69 As mentioned above, closed user groups do not generate sufficientcompetitive pressure to constrain the level of termination charges because theMNOs can separate them from less sensitive customers by offering specialarrangements that bypass termination. By separating out the more price-sensitivecustomers, the MNOs face less competitive pressure in setting charges for theother customers. Private wire services and mobile-to-mobile adaptations areexamples of such arrangements that segment the market in this way.

29

3.70 Private wire services are provided as leased lines that connect a corporateprivate exchange to a mobile network. They allow corporate customers to makeand receive calls to and from mobile phones without paying the retail price forfixed-to-mobile calls. Under private wire arrangements, the price for calls fromfixed-to-mobiles is lower than the price paid by the generality of fixed linecustomers. For SMEs that try to reduce the costs of calls to their own mobiles,private wire services are a relatively popular solution (Oftel’s Business use ofmobile telephony - published 24 October 2002 noted that 7% of small businessesand 21% of medium businesses used private wire services in this way).

3.71 Mobile-to-mobile adaptations consist of devices which give callers a radio linkfrom their private exchanges to allow fixed-to-mobile calls to be converted to on-net mobile-to-mobile calls. Using this facility, calls can be made from the fixedphones in the office to the company’s mobiles but are carried and charged in thesame way as calls from mobiles. This allows the subscriber to take advantage oflower prices for on-net calls rather than paying the standard price for calls fromfixed-to-mobiles.

3.72 Private wire services and mobile-to-mobile adaptations are likely to beintroduced only where the savings from lower prices for call to mobiles outweighthe costs of installing them. This, in turn, is likely to occur only where a sufficientproportion of the fixed line phone calls of the customer is directed to a singlemobile network. Thus, they are unlikely to be an effective substitute to standardfixed-to-mobile calls for residential consumers. However, as mentioned above, themain reason why their presence is unlikely to constrain termination charges forfixed-to-mobile calls generally is that they constitute a targeted tariff aimed atseparating out the most price-sensitive customers. They do not impose acompetitive constraint on the prices which operators can charge for calls to lessprice sensitive customers.

Multiple SIM cards

3.73 If mobile users could receive their incoming calls on mobile networks otherthan the one to which they subscribe for making outbound calls, this could putsome pressure on mobile voice termination charges. For that form of substitutionto take place, the called party must be able to switch their handset betweendifferent networks. This is possible through the use of multiple SIM cards.

3.74 A subscriber can have a mobile phone with an internal dual SIM cardholderthat allows him to switch from one network to another by turning the phone on andoff. There are already devices available in the UK market which allow customersto use different SIM cards in the same handset and, thus, switch betweennetworks (Oftel is aware of the existence of SIM holders that can hold up to fourSIMs). However, to place some pressure on the MNO with high terminationcharges the subscriber should be, by default, on the network with cheap voice calltermination charges and only switch to the other network to make cheap outboundcalls. This is laborious and time-consuming (because to switch networks is anoperation that requires time). In addition, it relies on the called party having the

30

incentive to change network every time (s)he needs to make a call and to switchback again at the end of the call, so that the next inbound call will use the networkwith lower termination charges. It is doubtful that such an incentive currentlyexists given the CPP arrangement and customer behaviour under thatarrangement. Hence, it is more likely that subscribers currently exploit the dualSIM card opportunity purely to take advantage of differences in the prices ofoutgoing calls.

3.75 An automatic mechanism to re-route calls can also be considered. In thatcase, a mechanism would be implemented that instructs the called party’s mobilephone to switch network automatically when a call is arriving. No such mechanismcurrently exists and, in the Director’s view, it is unlikely to develop in theforeseeable future, due to significant technological difficulties and to the lack ofincentives on the part of the called party to make use of a facility that reduces thecost of incoming calls. In addition, a further hurdle is posed by the need for MNOsto allow access to their handsets/SIM cards to install the necessary software (aswell as allowing any necessary signalling to pass across the mobile network tocontrol network selection) since the MNOs have little incentive to give this co-operation.

3.76 The main limitation of these two scenarios is that they rely on the called partyhaving an interest in reducing the cost to other persons of calling his/her mobile(as for example (s)he needs to acquire a multiple SIM handset). At present, thisincentive is absent (see above paragraph 3.7); though this could change overtime,as discussed in Annex C.

Call- divert

3.77 To have calls automatically forwarded, mobile users could subscribe to apersonal numbering service (PNS) which allows them to give out a single numberand have all their calls directed to any number they specify (e.g. their fixed line).Calls to PNS are more expensive than calls to mobile numbers because:• to divert calls an additional leg is added to their route (Calls to a personal

number go to the PNS provider first, which then forwards it to the appropriateterminating network). PNS providers have to ensure that they can cover thecost of terminating calls to any number specified by the subscriber (i.e. also tomobile numbers); and

• PNS providers need to keep a database of the numbers to which it mustforward the calls.

3.78 Calls to PNS can be either CPP or RPP, but in either case their higher priceimplies that it seems unlikely that they could represent effective substitutes to callsto mobile phones for called and calling parties. This service may, thus, beattractive only for those subscribers who are sensitive to the price for others to callthem on their mobile. However, the MNOs will typically have already separatedthese subscribers from the generality of subscribers by offering them speciallytargeted tariffs.

31

3.79 An alternative to PNS is for the called party to use a single terminal that isboth a cordless phone and a GSM phone. This allows subscribers to make andreceive calls using their "fixed line" (via the cordless element of the phone), whenwithin the range of the cordless base station and the GSM element, when outsideof it. Thus, while the called party is within the range of the cordless base station,there is a substitute for the termination service.

3.80 However, the cost of such dual phones is currently high and, since the rangeof the cordless base station is limited, the savings that accrue to the called party(from lower outgoing charges) are small. It is the calling party who derives most ofthe benefits from the use of these phones, but, as discussed above, due to CPP,most mobile subscribers would not factor into their decision any saving apart fromtheir own (see section on closed user groups for consideration of the minority ofmobile subscribers that do consider the cost to others). Hence, these phones arenot widely used. The situation may, however, become different if over time there isa change in the incentives on the called party change.

3.82 Therefore, for the reasons set out above the Director believes that automaticcall forwarding services do not currently generate significant pressure on the levelof mobile voice termination charges.

Initial conclusions on the behaviour of called party in response to anincrease in the price for calls to mobiles

3.83 On the basis of the evidence and the arguments discussed above, theDirector considers that the behaviour of mobile subscribers in response to a rise intermination charges above the competitive level and, thus, in the price of calls totheir mobiles is unlikely to render this increase unprofitable. The majority ofsubscribers are unlikely to react to a price increase and the presence of someusers who choose their mobile network also on the basis of the cost of incomingcalls is not sufficient to constrain voice call termination charges (because of theability of the MNOs to separate these customers from the other through specialself-selecting arrangements). Hence, at present, no significant constraints on theMNOs’ ability to set termination charges above the competitive level appear toarise from the behaviour of the called party.

Initial conclusions on retail demand-side substitution

3.84 In conclusion, the Director considers that there are no effective retaildemand-side substitutes that could constrain mobile termination charges to thecompetitive level.

Wholesale demand-side substitution

3.85 Substitution of wholesale voice call termination on an MNO’s network withwholesale voice call termination on a different MNO’s network cannot provide anydirect constraint on termination charges, since an operator wishing to offer calls to

32

a customer of a specific MNO must purchase termination from that MNO or it willnot be able to terminate the calls (on this issue see also the section on multipleSIM above).

Initial conclusions on demand-side substitution

3.86 For the reasons listed above, the Director considers that at present, there areno effective demand-side substitutes for voice call termination to specificsubscribers of a particular MNO. However, before reaching an initial conclusion onthe appropriate market definition, it is necessary to examine if there is anypotential for supply-side substitution or if products (or areas) over which there arecommon pricing constraints should be included within the market definition.

Supply-side substitution

3.87 Supply-side substitution occurs when, in response to a rise in the price of aproduct, suppliers of other products switch into supplying the product whose pricehas risen and render the price increase unprofitable for the firm whichimplemented it. This entry has to be fast enough (i.e. happen within one year) andat sufficiently low cost, otherwise it would be classified as new entry and its impactcould be considered as part of the SMP assessment rather than in the marketdefinition.

Retail supply-side substitution

3.88 For retail supply-side substitution to impose a constraint on the level ofmobile voice termination charges, there would have to be operators which do notcurrently provide calls to mobiles that can switch into such provision and thusundermine a price set above the competitive level. In order to have such an effect,the new provider(s) would have to be able to provide a service which did not relyon the provision of termination from the MNO to which the called party subscribes.At present, the Director cannot identify any such provider that would not dependon the MNO to which the called party subscribes to terminate the calls.

Wholesale supply-side substitution

3.89 For supply-side substitution to be an effective constraint on mobile voicetermination charges, there have to be other firms who could switch into theprovision of wholesale voice call termination to a specific subscriber of an MNO’snetwork with relative ease in response to an increase in termination charges.

MNOs other than the one to which the called party subscribes

3.90 Supply-side substitution in the wholesale market for mobile voice terminationcould come most easily from other MNOs, which have the necessary networkinfrastructure and expertise to terminate mobile calls. However, having a mobilenetwork is not sufficient for an MNO to be able to terminate calls to a subscriber ofa rival network. For this to happen, the mobile phone should automatically move

33

from its home network, on to that of the alternative MNO on which the call wouldthen be terminated.

3.91 It is possible for GSM handsets to operate on more than one network. Thiswas an original design requirement of GSM, and is relied upon for internationalroaming. However, it is not currently possible for the originating network of a callto a mobile to select which MNO terminates a mobile call.

3.92 Hence, it is the Director’s view that at present the lack of access tohandsets/SIM details and the technical difficulties in taking control of the handsetconstitute an effective barrier to an MNO providing voice termination tosubscribers of another MNO.

Local Area Networks over short-range radio technologies or Wireless Local AreaNetworks

3.93 Operators running WLANs could possibly enter the market for calls tomobiles competing with MNOs and, thus, put pressure on the level of mobile voicetermination charges. However, the Director believes that, at present, there aresignificant technical obstacles that would have to be overcome before such aservice could become viable for mobile users. WLAN operators cannot currentlyoffer the same coverage of GSM networks because of the limited range ofreception allowed by their equipment.

3.94 The main limitation of these two scenarios is that they rely on the called partybeing responsive to the price of inbound calls so they would be prepared to incursome cost to reduce the cost to the person calling his/her mobile (as, for example,(s)he needs to acquire a multiple SIM handset). At present, mobile subscribers donot take into consideration to any great extent the price of inbound calls whenmaking their purchasing decisions.

3.95 The Director has therefore reached the preliminary conclusion that all theseobstacles would prevent WLAN operators from being able to supply voice calltermination in competition with MNOs.

Mobile Virtual Network Operators

3.96 An MVNO is a firm that provides mobile telephony services to its customers,but does not have an allocation of spectrum and uses part of an MNO’s network.Currently, Oftel is not aware of there being any MVNO in the UK that providestermination services to its subscribers. To Oftel’s knowledge calls to the MVNO’ssubscribers are directly routed to the host MNO’s network and the originatingoperators pay this MNO for termination (directly). However, if an MVNO were ableto offer termination by having incoming calls to its customers routed to itself ratherthan to the host MNO and if it could select the mobile network on which toterminate each call on the basis of the charges, this may place some pressure onthe level of the termination charges.

34

3.97 For this form of arbitrage to happen, the MVNO should be able to control onwhich mobile network its subscribers were logged on at any time. Hence, theMVNO should be able to signal to its subscribers’ handsets, or SIM cards, tochange network, and then receive an acknowledgement back to say this had beendone successfully. However, even if this technological hurdle were overcome,MNOs would have little incentive to conclude the necessary agreements to allowMVNOs to operate in this way. Oftel is not aware of any MVNO in the UK whichhas such a multiple roaming agreement.

3.98 Furthermore, even if MVNOs could manage to negotiate lower terminationcharges, they would have the same weak incentive to reduce prices for calls tomobiles as the MNOs. As discussed above, retail subscribers select their mobileservice provider (MNO or MVNO) on the basis of the price of a number of serviceswhich do not often include the price of calls to their mobiles (an exception ismembers of closed user groups, see section above). Therefore, it seems unlikelythat the MVNOs, even if they managed to negotiate lower mobile terminationcharges, would reflect this in lower retail prices for calls to their subscribers; ratherthey would be much more likely to set lower prices for retail services purchased bytheir subscribers.

3.99 MVNOs, therefore, are unlikely to increase the competitive pressure on voicecall termination charges because a number of significant technologicaldevelopments would have to take place before they could exploit an arbitrageopportunity in termination charges. It is the Director’s view that the necessarychanges are unlikely to occur in the foreseeable future (i.e. in the next threeyears). In addition, they would lack the incentives to pass through any reduction intermination charge they may be able to obtain from MNOs into the prices thatcallers pay to reach their subscribers. Therefore, the Director considers thatMVNOs do not currently provide an effective constraint on mobile voicetermination charges.

Initial conclusion on supply-side substitution

3.100 For the reasons listed above, the Director considers that there are noeffective supply-side substitutes for voice call termination to the subscribers of aspecific MNO.

Common pricing constraint

3.101 On the basis of the conclusions reached above, there are no demand-sideor supply-side substitutes that should be included in the relevant market.Accordingly, the appropriate market definition would appear to be wholesale voicecall termination to a specified subscriber or number. However, the Directorconsiders that it would be wrong to narrow the market definition to this extent,because an MNO has limited ability to price discriminate between terminationcharges to different subscribers on its network. MNOs can separate the moreprice-sensitive customers by offering them arrangements that bypass thetermination charge and so takes such sales outside the scope of the market, e.g.

35

through a private wire service or use of on-net calls (these issues were discussedat length in the section on demand-side substitution). However, the MNOs are notable to price discriminate between customers when termination charges are paid.Thus a common pricing constraint applies to voice call termination to allsubscribers on one network. This implies that, if an MNO wishes to lowertermination charges for calls to one subscriber, it must in practice lowertermination charges for calls to all its subscribers, effectively equalising thecompetitive pressures placed on all the network’s termination charges. As aresult, the Director is of the view that the relevant market includes voicetermination to all subscribers of one MNO.

3G networks

3.102 With the advent of 3G services the question arises of how termination ofvoice calls on 3G network fits in the market definition discussed above.

Operators that have both a 2G and a 3G network

3.103 The four incumbent MNOs have acquired the spectrum necessary to offer3G services and are rolling out the necessary networks. For these operators thatrun both a 3G and a 2G mobile network, it is necessary to consider whether voicetermination on the 2G network and termination on the 3G network of the sameoperator are in the same economic market. The hypothetical monopolist testneeds to be repeated to check whether one service could act as a constraint onthe charge levied for the other service.

Demand-side substitution

3.104 The Director has considered whether termination on the 2G network andtermination on the 3G network of the same operator are demand-side substitutes.Subscribers to the 3G services of these operators will be given a dual modehandset that works on both 2G and 3G networks and will receive voice callsthrough both networks. However, the network on which the call is terminated willbe chosen by the terminating operator and neither the originating operator nor thecalling party will be able to affect this choice or even be aware of whether the 2Gor the 3G network has been chosen for termination.

3.105 Since callers are unable to choose the network on which calls terminate,voice call termination on the 2G network does not appear to be an effectivedemand side substitute if charges for termination of calls to 3G subscribers wereraised above the competitive level. Equally, voice call termination on the 3Gnetwork does not appear to be an effective demand side substitute if charges fortermination of calls to 2G subscribers were raised above the competitive level.

Supply-side substitution

3.106 The fact that the two networks are run by the same operator also impliesthat termination on the 2G network will never be a supply-side substitute for

36

termination of calls to 3G subscribers. The fact that an MNO can offer terminationon its 2G network for calls to its 3G subscribers will not impose any additionalconstraint on the level of the charges for termination on its 3G network. An MNOwill not undercut its own charges. For the same reason, termination on the 3Gnetwork cannot be a supply-side substitute for termination of calls to 2Gsubscribers.

Common pricing to originating operators

3.107 However, the Director believes that the further issue of common pricing alsoneeds to be considered. Although it is not yet clear how prices for 3G voice calltermination will be set, the Director has defined the market on the assumption thatan MNO with both a 2G and a 3G network will wish, in practice, to present asingle, averaged price to purchasers of voice call termination to its 2G subscribersand purchasers of voice call termination to its 3G subscribers. MNOs will use bothnetworks to terminate calls to 3G subscribers. It would be feasible for MNOs to setdifferent charges for each of 2G and 3G termination and to levy such chargesdepending on the network used for termination. There would, however, be somepractical problems to be addressed, such as whether upgrades to billing systemswould be needed and that originating operators may not know which network hadbeen used for termination. For the purposes of the market definition, the Directorhas assumed that MNOs will levy a single charge, which will be a weightedaverage of termination charges on the 2G and on the 3G network. This does notmean that the MNOs will set the same underlying charge for terminating calls ontheir 2G and 3G networks, but that the price paid by an originating operator toterminate a call to a subscriber of an MNO is likely, through the choice of theMNO, to be a weighted average of the underlying charges.

3.108 This pricing policy would imply that, from the point of view of originatingoperators, a common price would be paid for voice call termination on the 2Gnetwork and voice call termination on the 3G network. Therefore, it would bereasonable to include them in the same economic market.

3.109 However, if originating operators were to pay different charges on the basisof whether (or in what proportions) the call terminated on each of the 2G and 3Gnetworks, there would be a case for defining separate markets for 2G and 3Gtermination. This would not reflect any substantive change in the constraints onthe MNOs’ behaviour in setting the terms and conditions for 2G and 3Gtermination. Therefore, in the view of the Director, even if he were to defineseparate markets, the services in which the MNOs had SMP and the proportionateremedies would be no different.

’3’

3.110 With the grant of the licences for 3G spectrum, a new operator, '3', hasobtained spectrum to run a 3G mobile network. For the reasons alreadydiscussed above in relation to market definition, the Director believes that voicecall termination on 3's mobile network represents a separate market. The

37

difference in technology between termination on 2G and 3G networks does notalter the analysis and the conclusions on the potential demand and supply-sidesubstitutes for voice call termination.

3.111 It should be pointed out that, for some time, 3 will also be providingtermination services on a 2G network. Until it has rolled out its national network, 3will rely, to some extent, on the 2G network of another operator to terminate callsto its subscribers. 3 will not be directly providing these 2G termination services (asit has no 2G network), but simply reselling the services of the other MNO.However, from the perspective of originating operators, 3 will be the terminatingoperator for these calls and will set the charges for these 2G termination servicesas if it were providing them directly. For the same reason, discussed above, the2G and 3G termination services provided by 3 to originating operators are likely tobe subject to common pricing. For this reason the Director considers that these 2Gand 3G termination services to 3’s customers are part of the same economicmarket. The implication of this is that the 2G service provided to 3’s customers ispart of the market for termination services to 3’s customers, and the underlyingconveyance on a 2G network which underpins this service is part of the market fortermination on that 2G operator’s network.

Tetra network

3.112 For the reasons discussed above in relation to market definition generally,the Director believes that voice call termination on the tetra mobile network run byInquam represents a separate market. The difference in technology betweentermination on 2G and Tetra networks does not alter the analysis and the initialconclusions on the potential demand and supply-side substitutes for voice calltermination.

SMS

3.113 Another question related to the product market definition is whether SMStermination and voice termination are in the same market. Voice termination andSMS termination should be regarded as part of the same market if they aredemand-side substitutes or supply-side substitutes or if they are subject to acommon pricing constraint, for example because they are part of a “clustermarket”.

Demand side substitution

3.114 As discussed above (see paragraph 3.34 to 3.36 above), the Director is ofthe view that SMS does not provide an adequate alternative for voice calls tomobile phones on the demand side. This is due to the fact that consumersperceive SMS as a limited substitute for a voice call and, mostly, as an activitylargely additional to voice calls to mobile phones. The main reasons for this arethat SMS enables parties to exchange only relatively short messages (the numberof characters in a message is limited) and that SMS can be delayed, because,

38

unlike a mobile voice call, an SMS is transferred between networks on a “storeand forward” basis.

Supply side substitution

3.115 In relation to supply side substitution, as for voice termination, currentlyeach MNO can only supply SMS termination to the subscribers of its own network.Hence, SMS termination to a specific mobile customer is offered by the sameMNO which provides voice termination to that customer. Although an MNO canoffer SMS termination to its subscribers, it is unlikely to do so in such a way as toimpose any additional constraint on the level of the charges for voice terminationon its network. It is very unlikely that an MNO will undercut its own charges. Thisimplies that supply side substitution from SMS providers would not constrain voicetermination charges.

3.116 Furthermore, MNOs are able to price-discriminate between purchasers ofSMS termination and purchasers of voice termination, and currently set differentcharges for these two services.

Cluster market

3.117 In both the fixed and mobile retail markets, the Director considers a range ofcall origination services (such as access and outgoing calls) to be in the samemarket on the basis of a cluster market analysis. This is because consumerspurchase these services as a bundle. Operators, thus, compete for retailsubscribers, not on the prices of each single outgoing service, but on the overallprice of the bundle. This implies that these services, even though no service in thebundle is a demand or supply side substitute for any of the others, are subject to acommon pricing constraint and thus can be considered to be part of linkedmarkets.

3.118 In the case of mobile voice and SMS termination, the Director considers thatthis argument does not apply because of the consequences of the CPParrangement. The CPP arrangement implies that, even though it is the calledparty that chooses the terminating network on which it can be reached, it is thecalling party who ultimately bears the cost of termination. Since the choices ofwhether to make a call to a particular recipient, and of what type (voice or SMS),are made by a number of different callers at different times, termination servicesare subject to separate purchasing decisions of different retail customers and arenot part of a bundle. The calling parties do not make a single decision to purchaseall their voice and SMS call termination services from a particular supplier in thesame way that fixed and mobile customers choose their supplier of originationservices. Therefore, given the way retail choices are made, the competitiveconditions in the wholesale provision of SMS termination and voice termination arenot linked by a cluster market analysis in the same way as the provision oforigination services. For more details about the issue of linked national marketsfor mobile services see below (paragraphs 3.121 to 3.122).

39

3.119 In conclusion, the Director takes the initial view that SMS termination andmobile voice termination are not part of the same market. Accordingly, the marketsthat the Director has identified are for termination of voice calls on individual publicmobile networks, or in the case of 3’s 2G services voice call termination providedto its subscribers.

3.120 On the basis of the analysis and of the evidence discussed above, theDirector currently considers that any adequate wholesale supply or demand sidesubstitutes for termination of calls to the subscribers of a specific MNO do notcurrently appear to exist. Current technology does not allow the termination of acall to a mobile other than on the network of the MNO to which the called partysubscribes. This appears unlikely to change in the near future. At the retail level,the Director is of the view that, at present, there are no effective alternatives forcallers that could act as a constraint on termination charges. In addition, callersappear to have limited awareness of the cost of calling mobiles. There is aminority of mobile users that show a higher elasticity to the price of incoming calls.The MNOs have, however, separated these users by offering them special tariffs,thus preventing this group from putting any effective pressure on terminationcharges. Technological conditions and the behaviour of called and calling partiesmay change over time, but the Director believes that this is extremely unlikely tohappen in the next three years. Hence, the Director believes that, at present,there are separate markets for voice termination to the subscribers of each MNO.

Alternative product market definitions

National market for wholesale mobile voice call termination on all networks

3.121 A possible alternative definition could be a national market for mobile voicecall termination. Such a definition would require that the suppliers of voice calltermination (the MNOs) compete with each other on a national scale for thewholesale business of carrying calls to mobiles. Such competition would need tobe distinct from retail competition for mobile customers, since otherwise the moreappropriate market definition would be linked national markets for mobile services(see below).

3.122 This could happen if one MNO could offer termination of calls to customersof another MNO. However, currently MNOs cannot do so (see section onwholesale supply substitution above) and, therefore, they cannot compete in theprovision of voice call termination. Hence, the Director considers this marketdefinition to be inappropriate given the current and expected conditions.

Cluster national market for mobile services

3.123 Even if the hypothetical monopolist test led to distinct markets for differentmobile services, these services might be subject to a common pricing constraintand thus be part of a cluster market.

40

3.124 If mobile owners purchased a bundle of services from MNOs that includedvoice call termination together with other retail services, even if each element ofthe bundle were not a demand-side substitute for any of the others, the fact thatthey are all consumed and supplied together would link them. This is becauseunder these circumstances, MNOs would be competing for customers not on theprice of each single service, but on the overall price of the bundle and, thus, thevarious services would be subject to a common pricing constraint. This impliesthat an MNO would not be able to raise voice call termination charges, whilekeeping prices for the other services in the bundle at the same level, withoutseeing its customers switch to another network in response to an increase in theoverall price of the bundle. The MNO would, therefore, be able to raisetermination charges only if, at the same time, it reduced prices for other servicesso as to maintain at the same level the overall price of the bundle. If this weretrue, an MNO could be constrained in its ability to increase charges for voice calltermination (though the extent of such constraint would depend upon the level ofcompetition in relation to the provision of the overall bundle).

3.125 However, as discussed above, there is currently no evidence that mobileowners usually consider the prices of incoming calls in addition to the prices ofoutgoing calls, when choosing their mobile network. At present, the evidencesuggests that calls to mobiles are not part of the bundle of services on whichMNOs compete for subscribers. Therefore, the Director is presently of the viewthat the appropriate market definition is not that of a cluster national market formobile services.

Geographic market definition

3.126 As termination of voice calls to the subscribers of each MNO is a separatemarket, it is reasonable to conclude that the geographic extent of each marketmatches the scope of each mobile termination provider’s network(s) or in the caseof 3 the network that it uses to terminate 2G calls to its subscribers. This isjustified by the fact that termination charges are uniform across each network (orgroups of networks, i.e. 2G and 3G).

Initial conclusions

3.127 For the reasons outlined in this chapter, the Director believes that, atpresent, mobile voice termination to each MNO’s subscribers constitutes aseparate market (with voice call termination to 2G and 3G subscribers in the samemarket if the MNO has both types of networks).

3.128 Accordingly, the Director considers that at present there are six separatemarkets:

a) wholesale voice call termination provided to the subscribers of 3b) wholesale voice call termination provided by Inquamc) wholesale voice call termination provided by O2d) wholesale voice call termination provided by Orange

41

e) wholesale voice call termination provided by T Mobilef) wholesale voice call termination provided by Vodafone

Forward look

3.129 The Director considers that at present there are no adequate supply ordemand side substitutes for termination of calls to the subscribers of a specificMNO. However, the Director considers that, as technology evolves, effectivewholesale supply and/or demand side substitutes may arise and marketboundaries may thus vary. In addition, changes in the behaviour of callers and ofcalled parties may take place over time, as awareness of the cost of callingmobiles and of methods for keeping down the cost of making and receiving callson mobiles increases. The Director, nevertheless, believes that it is extremelyunlikely that these developments, which could lead to changes to the marketdefinition, will happen in the next three years. Under Clause 81 of theCommunications Bill there is a duty to review these markets at appropriateintervals.

Relationship between this market definition and the Commission’sRecommendation

3.130 The Director’s initial conclusion that mobile voice termination on eachMNO’s networks, or in the case of 3’s 2G services voice call termination providedits subscribers, constitutes a separate market is broadly consistent with theCommission’s recommendation.

Question 1

Do respondents agree with the market definition? If not respondents shouldplease provide evidence that might support an alternative view

Question 2

Will future developments in the mobile market, including changingconsumer behaviour, increase competition in voice call termination? Ifrespondents think they will evidence should be provided explaining how thiswill occur.

42

Chapter 4

Assessment of market power

Introduction

4.1 Under the new Directives significant market power (SMP) has been newlydefined so that it is equivalent to the competition law concept of dominance. Article14 of the Framework Directive states:

"An undertaking shall be deemed to have significant market power if, eitherindividually or jointly with others, it enjoys a position equivalent todominance, that is to say a position of economic strength affording it thepower to behave to an appreciable extent independently of competitors,customers and ultimately consumers."

4.2 SMP may be held by one company in a given market (single dominance) or bymore than one company (collective dominance). In assessing whether anyprovider has SMP, this review takes account of the Commission’s Guidelines onSMP as well as Oftel’s equivalent guidelines.

4.3 Clause 43 of the Communications Bill provides that a person to whom an SMPservice condition is applied must be a communications provider or person whomakes associated facilities available and a person whom Ofcom (or the Director inthe interim) have determined to have SMP in a specific market for electroniccommunications networks, electronic communications services or associatedfacilities.

Relevant market

4.4 As discussed in Chapter 3, the relevant markets for the purpose of this revieware:

a) wholesale voice call termination provided to the subscribers of 3b) wholesale voice call termination provided by Inquamc) wholesale voice call termination provided by O2d) wholesale voice call termination provided by Orangee) wholesale voice call termination provided by T Mobilef) wholesale voice call termination provided by Vodafone

Criteria used in assessing significant market power

4.5 This market power assessment analysis focuses on single firm dominance.The Director considers that in the market for wholesale mobile voice calltermination SMP cannot be held by more than one company, due to the specificcharacteristics of this market (see below and Chapter 3). As a consequence,

43

there is no need to assess and apply the criteria relating to collective dominance inthe assessment of market power set out in this chapter.

4.6 This chapter begins with a brief discussion of the CPP arrangement, which theDirector believes is the key factor that shapes the competitive conditions prevailingin the wholesale mobile voice call termination markets. The rest of the Chapter isdevoted to a review of the single dominance criteria defined by the EUCommission and by Oftel in their respective SMP guidelines, which the Directorbelieves are the most significant in this market. The remaining criteria, which areless relevant, are discussed in Annex B.

4.7 To inform his SMP analysis, in addition to drawing on internal expertise(especially for technical aspects), the Director has gathered evidence in variousways. He has employed statistics and figures collected by Oftel as part of itsgeneral data gathering function. He has also used comparative tables containedin the EU 8th Implementation Report.

Calling party pays

4.8 As mentioned in Chapter 3, the CPP arrangement plays a fundamental role inshaping the boundaries as well as the competitive conditions of this market. TheCPP implies that demand for termination is generated and the charges for it areborne by the calling party, but that the level of these charges is strongly affectedby the action of the called party who chooses the termination network.

4.9 Hence, the CPP arrangement determines that termination on each MNO’snetworks represents a separate economic market in which each MNO is amonopolist.

Discussion of the single dominance criteria

4.10 This section is devoted to examination of the single dominance criteria set bythe EU Commission and by Oftel in their SMP guidelines most relevant to thismarket.

Market shares

4.11 Each MNO has (since launch of its services) a 100% share of the market forterminating voice calls on its own respective 2G network, both when measured byvolume of calls and by revenues. As services on 3G networks are launched, eachMNO running a 3G network will also have a 100% share of the market forterminating voice calls on that network. This means that each MNO is, in effect, amonopolist in the supply of termination for voice calls to its own network.

4.12 As mentioned in Chapter 3, until it has rolled out its national network, 3 willrely on the 2G network of another operator to terminate calls to its subscribers.Since calling parties and originating operators have no choice but to use 3 to

44

terminate those calls (even if 3 uses another MNO’s network), 3 has 100% of themarket for 2G mobile voice termination to its subscribers.

4.13 Market shares do not represent a conclusive criterion on their own indetermining whether a firm has SMP in the relevant market. However, accordingto Community case law, there is a presumption that firms with market sharespersistently above 50% are dominant (AKZO Chemie v Commission - Case C-62/86 [1993] 5 CLMR 215), unless contrary evidence is provided.

Ease of market entry and absence of potential competition

4.14 The threat of potential entry can prevent incumbent firms from raising pricesabove competitive levels. However, this threat becomes weak when there arebarriers to entry.

4.15 In this market, the CPP arrangement and the current level of technologicaldevelopment, which does not allow any other provider to provide termination on aspecific network apart from the operator running that network, generate anabsolute barrier to entry. No technological means of terminating a call elsewhereseems likely to become available and commercially viable in the near future. Asdiscussed in Chapter 3, a number of significant technological developments and/orchanges in consumers’ behaviour would have to occur before competition couldarise in the market for mobile voice call termination (see, in particular, the reasonsfor the absence of supply-side substitution).

Excessive prices and profitability

4.16 The ability to keep prices persistently and profitably above the competitivelevel is an important indicator of market power. In a competitive market, individualfirms should not be able to raise prices above costs and sustain excess profits forprolonged periods of time.

4.17 The Director has examined the MNOs’ behaviour in setting 2G voicetermination charges to verify whether this is constrained by competitive forces. Ifthe market for 2G voice termination were competitive, charges would be expectedto reflect costs. However, voice call termination charges appear to have beensubstantially above a reasonable estimate of each MNO’s costs for a number ofyears (despite formal and informal regulation).

4.18 The Director’s view is that the most appropriate basis for assessing whethercharges are cost reflective is forward looking long run incremental costs (LRIC)plus a mark-up for common costs. LRIC-based charges most accurately reflectthe resources consumed by the provision of services and correspond most closelyto the level that would occur in a fully competitive market. Hence, the Director hascarried out a detailed modelling of the LRIC of UK 2G mobile networks and hasestimated the LRIC of voice call termination for an efficient 2G operator. TheDirector has then added a mark-up for common costs. More details about howthe LRIC and the mark-up have been arrived at can be found in Annex E on LRIC.

45

4.19 The total costs of termination on a 2G network resulting from the aboveexercise are well below the actual charges levied by each of the MNOs. Even inthe case of O2 and Vodafone, whose charges have been subject to a price cap ofRPI-9% since 1998 (following the 1998 MMC investigation, O2 and Vodafone wererequired to reduce their weighted average termination charge from 14.8 to 11.7ppm in 1999/2000 and by RPI-9% in the following two years), this regulatoryintervention has not been sufficient to drive charges down to costs. Vodafone’s2G charges exceeded the price cap level in the first year and O2’s have done so inthe first two years and have never fallen below the maximum permitted level.Oftel’s estimate of LRIC and the MNOs’ average termination charges in the lastfour years are shown below in Table 4.1.

Table 4.1: Mobile termination charges – weighted average over time (nominalpence per minute)

02 Vodafone Orange1 T-Mobile2

1999/00 Weightedaverage

11.7 11.7 () ()

2000/01 Weightedaverage

10.86 10.86 () ()

Weightedaverage

10.2 10.2 () ()2001/02

Oftel’s figure3 6.01 6.01 7.07 7.07

Weightedaverage

9.35 9.35 () ()2002/03

Oftel’s figure 5.71 5.71 6.67 7.67Source: Oftel1) The relevant figure cannot be published because it is commercially confidential2) The relevant figure cannot be published because it is commercially confidential3) Oftel’s assessment of the LRIC of 2G voice termination plus a mark-up for common costs.

4.20 The comparison made above is undertaken within the current regulatedenvironment (i.e. in which there is direct regulation of Vodafone’s and O2’scharges and significant informal regulatory pressure on Orange and T-Mobile). Inthe absence of any ex-ante regulation (and the threat of regulation), the Directorbelieves that MNOs would have an incentive to set termination charges at theprofit-maximising level. The Director has estimated that unregulated charges maythus be at 20 pence per minute or even higher (details of the calculation of theprofit-maximising termination charges can be found in Oftel’s paper TerminationCharges in the Absence of Regulation available at:www.oftel.gov.uk/publications/mobile/ctm_2002/termination0402.pdf).

Countervailing buying power

4.21 Countervailing buyer power exists when a particular purchaser (or group ofpurchasers) of a good or service is sufficiently important to its supplier to influencethe price charged for that good or service. In order to constrain the price

46

effectively, the purchaser must be able to bring some pressure to bear on thesupplier to prevent a price rise by exerting a credible threat, for example not topurchase or to self-provide.

4.22 In this case, the question of whether each MNO providing voice calltermination has SMP depends on the extent to which its monopolistic position maybe off-set by the buyer power of purchasers.

4.23 BT is the major buyer of voice call termination on mobile networks (see table4.2 below). In theory BT might credibly threaten not to purchase termination froman MNO and this would deprive that MNO of the pricing freedom that it derivesfrom its monopoly over termination. In practice, this issue is irrelevant since BT,even if it did have buyer power, has not been able to exert it because of itsobligation to complete all calls whatever the terminating network. The reasons forthis obligation will be set out in the document End to End Connectivity (to bepublished in May 2003). This regulatory requirement curbs any buyer power thatBT may have and leaves the MNOs free to set terminating charges above thecompetitive level.

Table 4.2: Share of minutes terminated on mobile networks by originatingoperatorsOriginating Operator ShareBT 26.4%Other Fixed PECNs 14.8%Total Fixed 41.3%Off-net 18.6%On-net 40.2%Total Mobile 58.7%

Source: Oftel Market Information

4.24 With regard to other fixed PECN providers, it is unclear whether any of themhas a level of buyer power sufficient to off-set each MNO’s monopoly in providingtermination and constrain their charges to a cost-reflective level. To understandfully if voice call termination charges would be constrained, the Director wouldneed to compare the market power of each MNO against the countervailing buyerpower of each of the fixed network providers.

4.25 The Director believes that this analysis would be very extensive and not veryinformative. He is of the view that it would be the exception rather than the rulethat the level of countervailing buyer power in these negotiations would be of theprecise magnitude to ensure that voice call termination charges were constrainedto the competitive level. Furthermore, the level of the charges set by MNOssuggests that other fixed PECN providers do not have countervailing buyer power.

4.26 The same would also be true for all the bilateral relationships between the sixMNOs. It is difficult to say what would be the outcome of bilateral negotiationsbetween them in the absence of any regulatory intervention. It is unclear if anyone MNO has the level of buyer power necessary to off-set the other’s monopoly

47

and ensure that voice call termination charges were set at the competitive level. Itis likely that the MNO with the balance of inbound traffic in its favour (the volumeof calls sent and received by each operator is not equal - source CC Report, table6.14) would have the incentive to keep the termination charge high and it is hard topredict which MNO’s interest would prevail in the negotiations (more details on theoutcome of bilateral negotiations between the MNOs can be found in Chapter 5).

4.27 Therefore, the Director believes that it is unlikely that any purchaser oftermination would possess sufficient countervailing buyer power to prevent theMNOs from setting termination charges above the competitive level.

Initial conclusions on assessment of market power

4.28 The definition of the relevant market has led the Director to conclude initiallythat there is a separate market for termination on each MNO’s network(s). Thismeans that each MNO is, in effect, a monopolist in the supply of termination to itsown networks. This initial conclusion combined with the analysis of the competitiveconstraints, presented above, indicates that each mobile operator enjoys SMP inthe provision of mobile voice call termination on its network(s), and in the case of3’s 2G services voice call termination provided to its subscribers. Such SMPallows each of the MNOs to behave to an appreciable extent independently ofcompetitors and consumers.

Forward look

4.29 At present, the Director does not believe that competition is likely to increasein the provision of mobile voice call termination services because no one otherthan the terminating MNO can terminate calls to its own subscribers. Currently,there is an absolute barrier to entry that impedes the development of competitionin the market. Hence, the Director considers that MNOs are likely to continue tohave SMP for at least the next three years. Depending on how technology andconsumers’ behaviour will develop there might be scope for competition to developin the future. However, the Director considers it unlikely that the necessarydevelopments will take place in the three years.

Impact of SMP

4.30 Each MNO faces an absence of effective constraints on its ability to raisetermination charges above the competitive level and has the incentive to exploitthis freedom. This implies that it is likely that MNOs, in the absence of any form ofregulation, would have the incentive to set termination charges at profit-maximising level well above the competitive level.

4.31 The Director considers that mobile termination charges significantly in excessof costs is not economically efficient and does not represent a fair distribution ofcosts and benefits between those who call mobiles and those who own them.More details on the effects generated by excessive mobile termination chargescan be found in Chapter 5.

48

Initial conclusions on significant market power

4.32 The Director considers that each mobile operator enjoys SMP in the provisionof mobile voice call termination on its network(s), and in the case of 3’s 2Gservices voice call termination provided its subscribers. The Director alsoconsiders that this SMP allows each of the MNOs to behave to an appreciableextent independently of competitors and consumers.

4.33 The main reasons that have led the Director to this conclusion are that:

• each MNO has a 100% of the relevant market and is thus a monopolist;• none of the purchasers of mobile termination have countervailing buyer power

to constrain charges at the competitive level;• termination charges on 2G networks have been persistently above costs; and• there might be scope for competition to develop in the future, however, this

depends on how technology and consumers’ behaviour develops and theDirector considers it unlikely that the necessary developments will take place inthe near future.

Question 3Do respondents agree with Oftel’s analysis of the effect that CPP has on thecompetitiveness of the market for mobile voice call termination?

Question 4Do respondents agree with the criteria used in assessing SMP? If not, howwould they suggest developing the analysis?

Question 5Do respondents have any evidence of the existence of buyer powersufficient to constrain the level of mobile voice termination charges? Is thereevidence that such countervailing buyer power will develop in the nearfuture?

Question 6Do respondents agree that all MNOs have SMP in the provision of mobilevoice call termination over their networks? If not, please provide anyevidence to support a different conclusion.

Question 7Do respondents believe that future developments are likely to change thisanalysis in the near future? If so, please provide any evidence of how thiswill happen.

Definition of the dominant provider

49

4.34 The Director proposes that each MNO has SMP in the relevant market, as setout in the draft notification at annex A.

4.35 The Director considers it appropriate to prevent a person to whom an SMPservice condition is applied (i.e. the dominant provider) which is part of a group ofcompanies, from exploiting the principle of corporate separation. That is to say, thedominant provider should not use another member of its group to carry outactivities or to fail to comply with a condition, which would otherwise render thedominant provider in breach of its obligations.

4.36 Article 16 of the Framework Directive requires that, where a nationalregulatory authority determines that a relevant market is not effectivelycompetitive, it shall identify "undertakings" with SMP on that market and shall onsuch "undertakings" impose appropriate specific regulatory obligations. For thepurposes of EC competition law (Vino vs Commission, Case C-73/95 (1996) ECRI-5447), "undertaking" includes companies within the same corporate group, forexample, where a company within that group is not independent in its decisionmaking.

4.37 Accordingly, the Director considers it appropriate that the obligations detailedin this consultation document (see Chapter 6) and draft notification shall apply toeach and every MNO and any subsidiaries or holding companies, or anysubsidiaries of such companies, all as defined by section 736 of the CompaniesAct 1985 as amended by the Companies Act 1989.

The relationship between the market reviews and Competition Act 1998 andEnterprise Act 2002 investigations

4.38 Economic analysis carried out in this consultation document is for thepurposes of determining whether an undertaking or undertakings have SMP inrelation to this market review. It is without prejudice to any economic analysis thatmay be carried out in relation to any investigation or decision pursuant to theCompetition Act 1998 or the Enterprise Act 2002.

4.39 The fact that economic analysis carried out for a market review is withoutprejudice to future competition law investigations and decisions is recognised inArticle 15(1) of the Framework Directive which provides that:

"…The recommendation shall identify …markets …the characteristics ofwhich may be such as to justify the imposition of regulatory obligations…without prejudice to markets that may be defined in specific casesunder competition law…"

4.40 This intention is further evidenced in the European Commission’s SMPguidelines which state:

Para 25 "… Article 15(1) of the Framework Directive makes clear that themarket to be defined by NRAs for the purpose of ex ante regulation are

50

without prejudice to those defined by NCAs and by the Commission inthe exercise of their respective powers under competition law in specificcases." (repeated in paragraph 37)

Para 27: "…Although NRAs and competition authorities, when examiningthe same issues in the same circumstances and with the same objectives,should in principle reach the same conclusions, it cannot be excluded that,given the differences outline above, and in particular the broader focus ofthe NRAs’ assessment, markets defined for the purposes of competition lawand markets defined for the purpose of sector-specific regulation may notalways be identical".

Para 28: "…market definitions under the new regulatory framework, even insimilar areas, may in some cases, be different from those marketsdefined by competition authorities."

4.41 In addition, it is up to all operators to ensure that they comply with their legalobligations under all the laws applicable to the carrying out of their businesses. Itis incumbent upon all operators to keep abreast of changes in the markets inwhich they operate, and in their position in such markets, which may result in legalobligations under the Competition Act 1998 or Enterprise Act 2002 applying totheir conduct.

51

Chapter 5

Detrimental effects arising from SMP in the terminationmarket

5.1 Chapter 4 discussed the Director’s initial view on SMP in the markets formobile voice call termination. Having looked at a number of indicators, heprovisionally concluded that each MNO has SMP in the market for the provision ofvoice termination services to its subscribers.

5.2 This Chapter discusses the impact on final consumers of this SMP and of theincentives thereby faced by the MNOs.

Voice termination charges in the absence of regulation

5.3 The Director believes that, in the absence of any ex-ante regulation (and of thethreat of regulation), MNOs would have the incentive to set voice terminationcharges at the profit-maximizing level. They might not immediately increase theircharges to this level, but they would do so over time. The Director accepts that,since the threat of future regulation would remain, it is possible that the behaviourof the MNOs would be constrained and that the termination charges would notquite reach profit-maximizing levels (see Chapter 4 for more details on this).However, he is of the view that MNOs would raise their termination charges abovecosts if no ex-ante regulation were placed on them.

The ‘swings and roundabouts’ argument

5.4 It could be argued that the Director’s concerns about the SMP held by theMNOs in the markets for mobile voice termination are unfounded, because anyexcess profits that MNOs earned in the voice termination market would becompeted away at the retail level. Hence, even if consumers lose from paying theresulting high prices for calls to mobiles, they would gain through lower prices forother mobile services, and there would be no detrimental effect overall. Thisargument is referred to here as ‘swings and roundabouts’; in the CC’s report it wascalled the ‘waterbed effect’.

5.5 The Director does not agree with this argument for the following four reasons:

1 the degree of competition in the retail market in which the MNOsoperate is not sufficient for all excess profits to be competed away;

2 economic efficiency is diminished because of the inefficient structureof prices;

3 high termination charges give rise to undesirable distributionaleffects; and

4 high termination charges have a distorting effect on consumers’choice.

52

1) The degree of competition retail market in which the MNOs operate;

5.6 In order that all excess profits earned in the mobile termination market to becompletely bid away for the benefit of consumers, the retail market in which MNOsoperate would need to be perfectly competitive. If this were the case, the overallprofitability of each MNO would be unaffected by the level of the terminationcharges. However, the Director considers that, although no MNO has SMP in theretail market for mobile services (See Review of competition: mobile access andcall origination – published 11 April 2003), conditions are not such as to allow theDirector to conclude that all the excess profits would always feed through intolower prices and that the MNOs would not be able to retain some of these excessprofits.

5.7 In any event, even if any excess profits earned by the MNOs in the mobiletermination markets were completely bid away by competition in the retail market,the Director would still be of the view that the resulting pricing structure wouldgenerate detrimental effects on consumers. This is further discussed below wherethe other three reasons for rejecting the “swings and roundabouts” argument areexamined.

2) The effects on economic efficiency

5.8 The pricing structure arising from monopoly in termination and competition inretail services is extremely unbalanced. In the absence of regulation, the MNOswould have the incentive to set termination charges at profit maximizing levels anduse the excess profits earned in the termination markets to reduce retail prices,and to engage in non-welfare enhancing activities aimed at retaining the mostprofitable subscribers or inducing them to switch network. Hence, the structure ofthe prices for mobile services would provide distorted price signals to consumers.The overall effect would be sub-optimal with the volume of calls to mobiles belowthe efficient level and the usage level of mobiles, as well as the rate ofreplacement of handsets and network switching above the optimal level. Forfurther details see Oftel’s paper Ramsey prices and the incentives of mobileoperators (available at:www.oftel.gov/publications/mobile/ctm_2002/ramsey120202).

3) The distributional effects

5.9 Since termination charges are passed on in the prices for retail calls, if theformer are set above cost, callers end up paying more for fixed-to-mobile and off-net calls more than the costs incurred by MNOs as a consequence of their calls.However, it can be argued that even if termination charges are set above costs,since most consumers have a mobile phone, what callers to mobiles lose in payinghigh termination charges they gain in low access and outgoing call charges.

53

5.10 The Director accepts that in some cases callers to mobiles and mobilesubscribers are the same people. However, there is a significant number ofcallers for whom this effect is likely to be important.

5.11 The people most adversely affected by such a transfer are those without amobile phone, as these people call mobile phones but do not themselves own one.This group includes a disproportionate number of elderly and low-income people.Latest estimates by Oftel (Consumers’ use of fixed telecoms services – publishedApril 2002) suggest that roughly 19 per cent of British households, representingover 8 million people, have a fixed-line telephone but do not have a mobile phoneand that 10 per cent of their total fixed calls are to mobile phones. The Oftelresearch (Consumers' use of fixed telecoms services– published April 2002) alsoshows that roughly two-thirds of fixed line-only households earn less than £11,500compared with just over half of mobile-only households (although the older, retiredfixed-line-only households might have more assets — for example, ownership oftheir home).

5.12 The Director is aware of the counter-argument which is that mobile-only usersmight be disadvantaged by the lowering of termination charges. However, theDirector believes that this effect is out-weighed by the number of fixed line-onlyconsumers who are adversely affected by high termination charges. An Oftelsurvey (Consumers' use of fixed telecoms services Oftel residential surveyNovember 2002– published January 2003) conducted in February 2003 showsthat the group of mobile-only customers represents just 7 per cent of households.

5.13 In addition, the Director considers that some of those who have both a fixedline and a mobile phone suffer a detriment from excessive mobile terminationcharges. This is true for all of those people who use their fixed line instead of theirmobile phone to call a mobile or make more off-net calls to mobile than theyreceive.

5.14 The Director also believes that payphone users who call mobiles, but do notthemselves own a mobile phone, are unfairly contributing to subsidising the MNOs’customers through high termination charges. An Oftel survey (Consumers’ use offixed telecoms services – published April 2003) into users of payphones showsthat at least 7 per cent of British adults use payphones at least once a month andthat 60 per cent (i.e. 4 per cent of British adults) of payphone users call mobiles.Of this 60 per cent, 30 per cent (i.e. around 2% of British adults) do not own amobile phone themselves, thus, in the Director’s view, end up paying an excessivetermination charge without any benefit from lower retail prices.

4) The effects on consumers’ choice

5.15 The Director believes that termination charges for fixed-to-mobile calls setabove costs distort consumers’ choice, thus generating an inefficient allocation ofresources. As both fixed-to-mobile and off-net calls include the same terminationcharges choice between the two is not distorted (however, it is possible that MNOsdo not pass all of this cost (i.e. the termination charges) into their off-net retail

54

prices, as the retail mobile market is not yet perfectly competitive. In this case alsothe choice between these two types of calls is distorted). However, the Directorbelieves that excessive termination charges are likely to distort the choice betweenmaking a fixed line call to a mobile or an off-net call and an on-net mobile-to-mobile call.

5.16 As mobile technology uses more resources than fixed technology, it is moreefficient to use a fixed phone to make a call when this is possible. Current pricesprovide callers with the wrong signal as they often direct them towards calling on-net rather than from a fixed phone. Hence, the Director considers that excessivemobile termination charges are detrimental to consumers because they lead to thesetting of retail prices that generate an inefficient allocation of resources bydirecting consumers towards choosing the higher cost, more resource-intensiveservice (i.e. the mobile-to-mobile call).

Initial conclusions on the ‘swings and roundabouts’ argument

5.17 For the four reasons discussed above, the Director is of the view that the‘swings and roundabouts’ argument does not provide a justification for settingtermination charges above costs.

Further consideration on the effect on economic efficiency

5.18 It could be claimed (paragraph 2.435 of the CC Report ) that setting hightermination charges should not generate any concern in terms of efficiency,because the MNOs would, absent regulation, set a Ramsey pricing structure for allmobile services. The Director rejects this argument for the reasons set out below.

Ramsey prices

5.19 'Ramsey Prices' refers to the set of prices for a group of services thatmaximize economic efficiency social welfare when the presence of common costsacross these services does not allow the adoption of marginal cost pricing (sincethe firm(s) would not break even if it priced all services at marginal cost).

5.20 The starting point for understanding Ramsey pricing is that that the ’first best’pricing structure (i.e. the structure that maximizes economic efficiency) is achievedwhen the prices of all services are set equal to their respective marginal costs (thefirst-best outcome would also reflect any externality present in the market, but inthe interest of simplifying the discussion, the analysis abstracts from this point).Such prices direct consumers to purchase a service (or product) if, and only if,their willingness to pay exceeds the resource cost to society of producing thatservice. However, in industries with economies of scale and scope, like the mobileindustry, setting prices equal to the marginal cost is not feasible as it wouldprevent firms from breaking even, as they would fail to recover the fixed andcommon costs incurred in the production process. However, the extent ofeconomies of scale and scope in the mobile industry should not be

55

overstated. Economies of scale and scope at the network level are limited, e.g.Oftel’s LRIC model finds that only 3-5% of total network costs are common costs -for further details, see Oftel’s paper, Network Common Costs).

5.21 If the constraint that the set of prices must allow the firms to break even (i.e.to recover all costs and earn an appropriate return on the capital invested) is thenimposed, prices need to include a mark-up over marginal cost. To limit thedistortion (i.e. the reduction in economic efficiency) introduced by this departurefrom marginal cost pricing, the proportionate mark-up for common costs should beinversely related to the price elasticities of the demand for each of the services inquestion (the Ramsey principle). The level of the mark-up on each service should,thus, depend:

• on the marginal cost of providing the service (because the mark-up should beproportional to this cost):

• on the common costs to be recovered (because this determines the amount tobe recovered through the mark-ups); and

• on the size of the elasticity of this service relative to the elasticity of the otherservices (because the mark-up should be proportionately larger on a servicewith a relatively inelastic demand).

5.22 A set of mark-ups thus generated would cause the smallest reduction indemand compared to the first-best level.

5.23 Both own-price and cross-elasticities of demand (the own-price elasticity ofservice A is the percentage change in demand caused by a small change in theprice of A, the cross-price elasticity of service B with respect to A is the percentagechange in demand of B caused by a small change in the price of A), should alsobe taken into account in deriving the set of optimal Ramsey prices to reduce thedistortion introduced by the mark-ups. For example, services that arecomplements have negative cross-price elasticities, i.e. an increase in the price ofone service causes a reduction in the demand for the other service. Hence themark-up should be reduced on a service with a relatively larger (negative) cross-price elasticity, because the lower its price, the smaller the reduction in thedemand for the related service (the optimal mark-up can be negative if themagnitude of the cross-elasticity is sufficiently large).

5.24 In addition, the mark-up should also be abated for services that give rise topositive externalities. This is because any reduction in demand for the service willresult in additional loss of economic welfare via the externality. This is true in thecase of the mobile market as all telephone users derive a benefit from being ableto reach people on their mobile phone and, thus, gain from there being as manymobile subscribers as possible. Hence, to the extent to which this externality isnot completely internalised, there is an argument for subsidising subscriptionprices through higher termination charges. The Director has considered thisbenefit when assessing what the appropriate level of the termination chargeshould be by adding an appropriate externality surcharge.. More details on the

56

optimal externality surcharge can be found in Chapter 7 and Annex F onexternality.

5.25 However, the Director disagrees with the argument that MNOs would set theprices for their mobile services at Ramsey level.

Market elasticities

5.26 First, the Director believes that the MNOs would not set Ramsey pricesbecause the demand elasticities they face in the retail market are different fromthe market elasticities. This is due to the fact that the competition present in theretail market drives a substantial ’wedge’ between the elasticity faced by eachMNO and the market elasticity. However, Ramsey prices (i.e. the set of prices thatmaximize economic efficiency) should reflect the elasticities of the market demandcurves and not those of the demand curves faced by the individual operator.

5.27 The reason why the market elasticities are the relevant factors is becauseeconomic efficiency depends on output levels and consumers’ demand. A firmoperating in a relatively competitive market will lose sales in two different ways if itraises its prices. The price increases will reduce the overall demand for theproduct, and will also result in some of the firm’s sales being transferred to itscompetitors. In terms of the economic efficiency calculation discussed above,transfers of sales from one firm to another are not relevant charges in output. If areduction in sales by Firm A is matched by an increase in the sales of Firm B, thenoutput could not be said to have been reduced as a consequence of the pricechange. What is relevant in terms of the reduction in output is the overall loss ofsales of the product as a whole, which will be given by the market demandelasticity rather than the elasticity of demand curve facing an individual firm.

5.28 Having established that Ramsey prices depend on the relative pattern ofmarket elasticities, the next question is to consider the incentives that the MNOswould face when setting their prices for mobile services in the absence ofregulation. Each MNO operates in two separate markets (albeit that there areinterrelationships between prices in the two markets): a market for voicetermination to its subscribers, in which the MNO is a monopolist, and a singlemarket for retail services where it competes with other MNOs .

5.29 The disparity in competition between the two markets means that the relativesizes of the elasticities that each operator faces are quite different from the relativepattern of the market elasticities. Whilst each MNO effectively faces the marketdemand in termination, its individual demand curve in the retail market is muchmore elastic than market demand because of competition. In other words, in theretail market when an MNO raises its price it loses sales in two ways: (a)consumers make fewer calls or give up their mobile subscription altogether and (b)sales are lost to competing MNOs as consumers decide to take advantage of therelatively more attractive offerings of competitors. By contrast, in the terminationmarket a rise in termination charges would only generate effect (a).

57

5.30 Therefore, if the MNOs set prices on the basis of the demand elasticitiesfacing them (i.e. the firm specific ones) they would not be setting Ramsey prices,since these depend on the market elasticities.

Incentives on MNOs

5.31 The Director also believes that the relative conditions in the two markets inwhich the MNOs operate, rather than inducing them to set Ramsey prices, makesit impossible for them to do so on an individual basis.

5.32 If one of the MNOs decided to set Ramsey prices in both the termination andretail markets, it would be unable to sustain that set of prices, because thecompetitors would be able to offer lower retail prices funded by a highertermination charge. The MNO would, thus, lose retail customers to its competitorsand would have to respond by reducing its retail prices. It would also find itprofitable to raise its termination charge regardless of the state of competition inthe retail market. Overall, if hypothetically the retail market were perfectlycompetitive, each MNO would have incentives to continue lowering its retail prices,including below cost, so long as this was more than offset by termination profits(which are earned on calls to the mobile subscribers attracted by the low retailprices). The equilibrium would, hence, involve termination charges above theeconomically efficient Ramsey level possibly set at the profit maximizing level, andretail prices below the Ramsey level (for the theoretical underpinnings of thisresult, see section 3.1 of The Theory of Access Pricing and Interconnection, MarkArmstrong, 2001).

Conclusion on Ramsey prices

5.33 For the reasons discussed above, the Director believes that, in the absenceof any price control, the MNOs would have the incentive to set termination chargesabove costs and at an inefficiently high level.

Risk of discrimination towards new entrants

5.34 The SMP, and thus the pricing freedom, enjoyed by the MNOs in the mobiletermination markets could also be used to distort and reduce competition in theretail mobile market. Each MNO is a monopolist in the provision of terminationservices to its subscribers, but it competes with other operators in the outgoingand access market. Hence, it may exploit its position in the termination market toimpair its rivals’ ability to compete for customers. This has not been a problemuntil now because each of the four incumbent MNOs was in a position to respondto any attempt by one or more of its rival to limit its ability to compete.

5.35 However, the situation is changing with the entry of 3. Since 3 has justlaunched its services, it still has a small customer base and, thus, most of itsmobile-to-mobile traffic is off-net. If the four incumbent MNOs were to set highcharges for terminating calls from 3’s subscribers, 3 would have to raise the levelof its retail prices for outgoing off-net calls, which would impair its ability to acquire

58

subscribers. Given its asymmetric position in the retail market with respect to theincumbent MNOs, 3 would not be able to respond. Hence the Director believesthat it would be appropriate to safeguard competition in the retail market from anydistortion that may arise from the freedom that MNOs enjoy in the wholesaletermination market.

Effects on investments

5.36 The Director has also considered other effects generated by high terminationcharges, which may offset the detrimental effects just discussed.

5.37 It could be argued that high termination charges are necessary to financenetwork investment. The Director does not accept this argument. Providedcharges include an adequate return on capital, he considers that reducingtermination charges to a cost-based level would not impair the financial ability ofthe MNOs to pursue their 2G investment.

5.38 With regard to the effect of 2G high termination charges on investments onthe roll-out of 3G services, the Director believes that these should not besubsidised by 2G termination charges set in excess of costs. 3G is an incrementalinvestment from which a degree of profitability is anticipated. If its profitability isexpected to be sufficient to cover its expected cost of capital, then 3G should becapable of attracting funding. Therefore, the Director is of the view that theinvestment case for 3G should stand or fall on the basis of its own merit. The 3Ginvestment should not depend on subsidies from users of another technology andits success will not be affected by bringing 2G termination charges down to theircosts.

5.39 Hence, the Director considers that the take-up of 3G services will not behampered by a reduction of termination charges to a cost-reflective level.

59

Chapter 6

Options and proposals for remedies

Introduction

6.1 This chapter assesses options for regulatory remedies in the markets forwholesale mobile voice call termination on the networks of Vodafone, Orange, T-Mobile, O2, 3 [Hutchison 3G UK Ltd] and Inquam.

6.2 As explained in Chapter 4, the Director’s initial view is that each mobileoperator enjoys SMP in the provision of mobile voice call termination on itsnetwork(s), and in the case of 3, SMP in the provision of 2G services voice calltermination provided to its subscribers. Chapter 5 looks in detail at the detrimentsto consumers that stem from the SMP of the mobile operators, specifically inrelation to the setting of excessive charges.

6.3 This Chapter builds on the assessment of SMP set out in Chapter 4 and thespecific detriments discussed in Chapter 5 and looks at the remedies that might beapplied.

6.4 In the preceding work that Oftel and the Competition Commission haveundertaken in this area, consideration has been given to the underlying problem ofa lack of competitive pressure on call termination charges.

6.5 In Oftel’s view, it is appropriate in this review to give consideration first to thepotential for measures to be taken which would remove the underlying problem oflack of competitive pressure. Only if these measures are found to be impractical orineffective should attention be given to remedies to remove the detrimental effectsof the lack of competitive pressure. In other words, Oftel considers remedies whichwould result in there being no SMP before addressing the question of whichremedies would address the detrimental effects of SMP.

6.6 However, the Director would not propose to mandate the introduction ofmeasures unless he was confident that they would, in the immediate future, beeffective in introducing competitive pressure on termination charges. Most of themeasures to introduce competitive pressure on termination charges could beadopted voluntarily or be imposed by regulation.

Potential alternative solutions

6.7 As discussed in Chapter 3, some of these solutions are not yet technicallyfeasible, whilst others are feasible and in theory might lead to an increase incompetitive pressure on voice call termination charges. However none of themeasures have yet been introduced to act as an effective constraint on terminationcharges, possibly due to the lack of incentives on MNOs to do so.

60

6.8 Table 6.1 below describes these solutions. A fuller discussion can be found inAnnex C.

Solution Explanation1 Receiving Party Pays This solution consists of switching to a system of RPP

in which the recipient pays for the calls. This could beachieved in a number of ways:(a) the originating operator could allow the calling

party to request a reverse charge by dialling aprefix;

(b) the originating operator could allow the callingparty to call a local rate number and initiate anautomatic call-back;

(c) the mobile subscriber could be given theopportunity to accept to pay for a call at thestart or at some point during the calls; or

(d) the mobile subscriber could be given theopportunity to establish a pattern of incomingcalls for which they are prepared to pay.

2 Call back This solution involves the MNOs introducing a servicethat allows the calling party to have a choice of how tomake a call to a mobile customer, either:(a) he calls directly (and pays a price which

includes the termination charge set by theMNO), as it happens now, or

(b) he sends a text message or leaves a voicemailin the message box of the mobile customer.

In either case, the message would inform the mobilecustomer that the caller has called. The mobilecustomer would then have the option of calling back,and would bear the costs of making the mobileoriginated call.

3 Call divert This solution involves mobile subscribers using asingle terminal that is both a cordless phone and aGSM phone, which allows them to make and receivecalls using their "fixed line" (via the cordless elementof the phone) when within the range of the cordlessbase station, thus bypassing their mobile network.

4 Mobile virtualnetwork operators withmultiple roamingagreements

This solution involves MVNOs providing not justorigination services but also termination services totheir subscribers using other MNOs’ networks.By setting up multiple termination agreements, theMVNOs would be able to select the mobile network onwhich to terminate each call on the basis of itscharges, thus pressurising MNOs to offer them lowertermination charges.

5 Multiple SIMs This solution involves a subscriber being able to

61

acquire multiple SIM handsets that would allow themto make their calls using a network with low outgoingprices and receive their incoming calls on a networkwith lower termination charges

6 GSM Gateways This remedy works by setting up a mobile call on thenetwork where the call is to be terminated, and then"connecting" this call to the original call. By so doing,the originating operator pays the on-net retail rate,rather than the wholesale termination charge.

7 Tying terminationcharges tocompetitively suppliedservices

This solution involves setting the mobile voicetermination charges equal to the lowest existingtermination charge. Thereafter, any change in eachMNO’s termination charges is benchmarked againstthe percentage change of its weighted average retailprices.

8 Delivering a callfurther into theterminating network

This solution involves the originating operatorshanding over calls to the mobile network at the basestation controller nearest to the receiving party, thusby-passing some network components. This isachieved by setting up a network of leased lines thatreaches all base stations.

6.9 As discussed throughout this document (see Chapter 3), a key reason why theMNOs have a monopoly in the provision of termination services to theirsubscribers is the CPP arrangement adopted in the UK telephony market. Underthe CPP arrangement the calling party (and not the called party) pays the totalprice of retail call. Hence, the ideal solution would be to allow the calling party tochoose the network on which to terminate the call. If this were possible, it wouldprovide the MNOs with the incentive to compete on inbound call charges.However, this type of arrangement is currently technically not feasible and it isunlikely that it will ever be.

6.10 Most of the solutions listed above try to make up for the incentives generatedby the CPP and are all aimed at introducing some competitive pressure on thelevel of the mobile termination charges. However, the Director is of the view thatnone of them, in the immediate future, will create the necessary conditions that willbring termination charges down to their competitive level.

6.11 Solution 1 advocates a departure from the CPP principle for an RPParrangement where the called party pays. In the Director’s view a move to RPPwould be disruptive and uncertain in its overall economic effect. While such amove should not be ruled out forever, the Director does not believe it would beappropriate to introduce RPP at this time.

6.12 Solution 2, 3, 4 and 5 to be effective need to be accompanied by changes inthe behaviours of the called and calling parties which are unlikely to happenovernight. In particular, it would be helpful if callers to mobiles were better

62

informed about the relative prices for calling different mobile networks, andresponded to this information by making more calls to mobile subscribers onnetworks with low termination charges and fewer calls to networks with hightermination charges. This would induce mobile customers to become moreresponsive to the price of inbound calls so they would be prepared to incur somecost to enable callers to call them more cheaply. It is interesting to note that, ifsuch a change of behaviour took place, the MNOs themselves would have anincentive to compete on inbound call charges. Thus, in order for these solutions tobe effective it is necessary for there to be a change in consumer behaviour alongthe line of the type of change that would of itself remove the problem and the needfor a remedy.

6.13 Hence, the relevant question is whether any of these solutions would playany part in changing consumers’ information set and behaviour, so that callerswere more aware of the prices for calls to specific networks and mobile customerswere more responsive to the prices of inbound calls in choosing their mobilenetwork. In the Director’s view, none of the proposed remedies would by itselfquickly change consumers’ behaviour in this way, but each might have a part toplay in focussing more attention on inbound call charges and thus contributing to agradual change in behaviour.

6.14 A point of particular relevance is that mobile customers currently have theopportunity to exercise choice in a way that might constrain termination chargesonly when they make a choice of network. Such choice, while not irreversible, isexercised infrequently by most customers. Some of the proposed remedies wouldenable mobile customers to take action to put pressure on termination chargeswithout switching their network operator. To that extent the Director believesthese remedies could have a part to play in changing consumer behaviour. Thus,while he is unpersuaded of the adequacy of these proposed remedies in the shortterm, he would encourage the fixed and mobile operators to develop these ideaswith a view to seeing whether they might bring about changes in consumers’behaviour over time.

6.15 Solution 6 and 7 “borrow” competitive pressure by linking termination chargesto the charges for more competitive services. The problem with these remedies isthat they also dilute competitive pressure in the services to which call terminationis linked. Since there is, in a sense, a finite amount of competitive pressure, thelinking of a competitive service to a monopoly service increases the competitivepressure on the monopoly service at the expense of reducing the competitivepressure on the previously competitive services. It is thus an unattractive solution,as its potential to solve one problem is mirrored by its tendency to create anotherproblem.

6.16 Solution 8 does nothing to address the problem. The problem is not the levelof the costs of mobile voice termination, but the charges levied for the service,which are considerably above cost. This solution would not increase the amountof competitive pressure on termination charges and, hence, would be ineffective.

63

6.17 In light of the above, the Director’s initial conclusion is that none of thesolutions discussed above are currently likely to be effective, although solutions2,3,4 and 5 have the potential to focus consumers’ attention on the different levelsof voice call termination charges on different networks and could, over time,contribute to the development of competitive pressure on voice call termination

Options for regulatory remedies

6.18 Having considered the various alternative solutions and concluded that, atpresent at least, these are not viable alternative to regulation, the rest of thisChapter considers possible regulatory remedies to deal with the problems thathave been identified. However, the preliminary conclusions set out in this chapterare relevant only if the Director finds, after consultation, that the mobile operatorsdo indeed have SMP in the provision of mobile voice call termination services ontheir own network, and in the case of 3 wholesale 2G voice call terminationprovided to its subscribers. The Director will reach his final conclusions in the lightof responses to this consultation document.

6.19 Clause 84(1) of the Communications Bill provides that where the Director hasmade a determination that a person is dominant in the market reviewed, he shallset such SMP conditions as he considers appropriate and as are authorised in theBill. This implements Article 8 of the Access Directive.

6.20 Paragraphs 21 and 114 of the European Commission’s Guidelines on marketanalysis and SMP state that this means that the Director must impose one or moreSMP conditions on a dominant provider. Furthermore, the European Commissionstates that the imposition of no SMP conditions on a dominant provider would beinconsistent with the new regime. Thus, the Director is under an obligation toimpose at least one appropriate SMP condition on each of the mobile operatorswhere SMP is confirmed.

6.21 The Communications Bill (clauses 42-47 and 75-86) sets out what obligationsthe Director can impose if he finds that any undertaking has SMP. Thoseobligations relevant to this review are:

• The provision of network access;• No undue discrimination;• Transparency;• Cost orientation;• Cost recovery, including charge controls; and• Cost accounting and accounting separation.

6.22 Oftel has set out its intention to consider the appropriateness of SMPconditions on their merits in its regulatory option appraisal guidelines. For this,please see www.oftel.gov.uk/publications/about_oftel/2002/roa0602.htm. Oftelalso notes that Recital 27 of the Framework Directive provides that ex-anteregulation should only be imposed where there is not effective competition and

64

where competition law remedies are not sufficient to address the problem. In orderto provide a full analysis, Oftel has therefore also considered the option of no ex-ante regulation, and whether it would be sufficient to rely on competition law alone,while noting the obligation referred to in paragraph 5.4.

Aims of regulation and conditions that can be imposed

6.23 Clause 4 of the Communications Bill sets out the Community requirementsfor regulation. The Director, in considering whether to propose any conditions, hasconsidered all of these requirements. In particular, the Director has considered therequirement to promote competition and to secure efficient and sustainablecompetition. Therefore, the Director has particularly considered whether conditionsare required to prevent SMP in this market being used to distort competition indownstream markets.

6.24 Any SMP conditions imposed must comply with the various tests set out inthe Communications Bill. The Director must also bear in mind the duties set out inclause 4 of the Communications Bill.

6.25 In particular, each Condition must be appropriate (see clause 84(1)) andsatisfy the tests set out in clause 44 of the Communications Bill, namely that eachcondition must be:

(a) objectively justifiable in relation to the networks, services or facilities towhich it relates;

(b) not such as to discriminate unduly against particular persons or aparticular description of persons;

(c) proportionate to what the condition is intended to achieve; and

(d) in relation to what it is intended to achieve, transparent.

6.26 It is the Director’s initial view that the options proposed in this chapter satisfythe relevant requirements specified in the Communications Bill, as discussed indetail in the following paragraphs.

Options for remedies

6.27 Oftel has considered four different options for levels of regulation on each ofthe MNOs in relation to the relevant markets. Each option has been considered inrelation to the four main 2G MNOs, 3 and Inquam.

65

What are the options for regulation on Vodafone, O2, Orange and T-Mobile?

6.28 The various options considered below relate to options for regulation in themobile voice call termination markets for Vodafone, O2, Orange and T-mobile(hereinafter referred to in this chapter as ’the MNOs’). The Director considers itappropriate to consider different regulatory remedies to be applied in respect of 2Gand 3G voice call termination services, because of the differing competitive andeconomic conditions that characterise their provision.

6.29 The options considered are:

• Option A

No ex-ante regulation;

• Option B

A requirement to secure transparency through publication of prices and priornotification of price changes

• Option C

Option B plus a requirement to make transparent the charges, terms andconditions through publication of a reference offer; a requirement to provide mobilevoice call termination on fair and reasonable terms; a requirement not to undulydiscriminate in the provision of the service; and imposition of charge controls.

• Option D

Option C plus a requirement to maintain cost-accounting systems; a requirementto set prices on the basis of forward looking long-run incremental costs; and arequirement for separate accounting systems.

6.30 The Director’s initial view is that:• The MNOs’ 3G voice call termination services should not be subject to ex-ante

regulatory controls as set out in Option A• The most appropriate obligations to apply on the MNOs in relation to the

provision of voice call termination on their 2G networks are those set out inOption C.

6.31 Fuller analysis of each of the options is provided below.

Option A

No ex-ante regulation

66

6.32 Where markets are effectively competitive, ex post competition law issufficient to deal with any abuses that may arise. However, where markets arecharacterised by monopoly or enduring market power, it is necessary to considerwhether the ex post application of competition law is likely to provide adequateprotection for consumers or to allow competition to develop in related markets.

6.33 In the case of mobile call termination on 2G networks, the Director hasconcluded that the MNOs have the freedom and incentives to set terminationcharges in excess of cost. Chapter 5 considers the impact of this on finalconsumers, concluding that in the absence of specific regulatory controls, theMNOs would set inefficient termination charges above costs to the detriment offinal customers. The Director believes that ex ante regulation is required to preventthe MNOs from setting excessive charges for these services or acting in a mannerthat is anti-competitive and/or of detriment to consumers. If ex post competitionlaw powers were relied on, the Director would expect the charges to be set atexcessive levels, to the detriment of consumers, until he was able to conclude aninvestigation under competition law. The Director believes that this would not bein the best interests of consumers, nor does he believe that the resulting volatilityin prices would be in the overall interests of the industry. The Director also notesthat the Competition Commission, in its report on mobile call termination charges,concluded that ex ante controls for these services were appropriate.

6.34 However, the position with respect to 3G services is rather different. Theseare new and innovative services, where take-up is uncertain, and costs areunclear. In general, and consistent with what has been stated in the past, theDirector does not normally apply regulatory controls to such new services asregulation could deter continued investment and development of new services,ultimately disadvantaging consumers. Furthermore, it is also presently unclearhow charges for 3G voice call termination services will develop and Oftel will beseeking further information on this during this consultation. In the absence of suchinformation and as 3G services are new and innovative services to whichregulatory controls might have a detrimental effect, the Director considers that exante regulation would be disproportionate.

6.35 The Director will however continue to look at how 3G voice call terminationcharges develop to ensure that the regulatory controls applied to 2G voice calltermination services are not circumvented. If this were to be the case the Directorwould consider whether some additional ex ante regulatory controls were requiredto prevent such circumvention.

6.36 Therefore, for the reasons, set out above, the Director’s initial view is thatapplication of ex ante regulation is disproportionate in relation to the MNOs 3Gvoice call termination services.

Option B

A requirement to secure transparency through publication of prices andprior notification of price changes

67

• Obligation to publish charges and proposed changes to those charges inadvance of them taking place

6.37 The requirement to publish charges and changes to those charges inadvance has two main purposes: (a) to assist transparency for the monitoring ofpotential anti-competitive behaviour and (b) to give advanced warning of chargechanges to providers purchasing wholesale termination services. The latter isimportant to ensure that other providers have sufficient time to plan for suchchanges, as they may need to restructure their retail prices in response to chargechanges at the wholesale level.

6.38 The Director considers that, consistent with the current contractual obligationbetween the MNOs and purchasers of voice call termination services, the period ofprior notification for charge changes should be 28 days. This period allowssufficient time for purchasers of the services to reflect changes in their retailprices.

6.39 This is particularly important in respect of this market as each of theoperators has SMP in the relevant market and because purchasers of terminationservices have no choice but to pay the mobile voice call termination charge whenterminating calls to each of the mobile operators.

6.40 However, such an obligation in respect of the MNOs 3G voice call terminationservices may be unwarranted. 3G retail services are new and innovative servicesand as such their take-up is uncertain and it remains unclear how they will developand at what speed. The Director does not seek to regulate new and innovativeservices in advance of their launch or soon after launch as these factors remainuncertain. This is so as not to adversely affect the continued development andinvestment in such services.

6.41 For these reasons, the Director’s initial view is that it is necessary to requirethe MNOs to publish their charges for 2G voice call termination services and topublish changes to those charges 28 days prior to those changes taking effect.However, for the reasons set out above, this requirement should not apply to theMNOs 3G voice call termination services.

Option C

Option B plus a requirement to make transparent the charges, terms andconditions through publication of a reference offer; a requirement to providemobile voice call termination on reasonable request and on fair andreasonable terms; a requirement not to unduly discriminate in the provisionof the service; and imposition of charge controls.

6.42 This option includes the transparency requirements set out in Option B, plusa requirement to publish a Reference Offer; a requirement to provide mobile

68

termination on reasonable request and on fair and reasonable terms, not to undulydiscriminate, and charge controls.

• Obligation to publish a reference offer

6.43 A requirement to publish a Reference Offer (RO) has two main purposes: (a)to assist transparency in monitoring potential anti-competitive behaviour (ifcharges are to be published); and (b) to provide visibility of the terms andconditions on which other providers can purchase services.

6.44 Publication of an RO will ensure stability and certainty to operatorspurchasing mobile voice call termination services. In addition it may enablenegotiations between parties to be completed more quickly, help to avoid disputes,and give confidence to purchasers of such services confidence that those servicesare provided on non-discriminatory terms.

6.45 Publication of a RO is particularly important where there is also a requirementto provide access for the purposes of voice call termination services. Where thereis no obligation to provide access the benefits of publishing an RO are significantlyless as there is no obligation to provide access in accordance with what is in theRO. Publication may however, still be beneficial for improving generaltransparency.

6.46 The Director’s initial view therefore is that the MNOs should be obliged topublish a RO in respect of any 2G voice call termination service to which anaccess obligation applies. In the absence of an access obligation there should beno requirement to publish a RO, as the cost of compliance would not be justifiedby the limited benefits.

• Obligation to meet reasonable requests for mobile voice call terminationservices on fair and reasonable terms

6.47 Without an obligation to provide access to mobile voice call terminationservices on reasonable request and on reasonable request and on fair andreasonable terms, competition in downstream markets could be severely distortedby the mobile operators refusing to supply services or by doing so on terms thatwere unfair and unreasonable. This behaviour would be of particular concern if itwere to be directed towards 3, which, as a new entrant with a small subscriberbase, would have a larger share of off-net traffic compared to the other MNOs.The inability to terminate calls on other mobile networks or or to have to meetunreasonable termination terms, such as high charges to do so, would reduce 3’sability to compete with the MNOs for subscribers in the retail market. Thisreduction in competition would reduce the choice available to consumers.

6.48 The ability, and incentives, on MNOs to act in such a way suggests that arequirement to provide access for the purposes of voice call termination on fairand reasonable terms is justified. However, the Director believes that this is a

69

burdensome regulatory requirement and may not be appropriate in relation to 3Gvoice call termination services on 3G networks. 3G services are new andinnovative services, take-up is limited and currently uncertain and it is not yet clearhow the services will develop or at what speed. The Director generally avoidsregulating new or emerging services, as he believes that to do so might have anegative impact on investment incentives and on the continued development ofsuch services.

6.49 Hence the Director’s initial view is that:• the MNOs should be subject to a requirement to provide access on reasonable

request and on reasonable terms and conditions to their 2G voice calltermination services; but

• they should not be subject to such a requirement in respect of their 3G voicecall termination services.

• Obligation not to unduly discriminate in the provision of access

6.50 As explained in Oftel’s guidelines Imposing access obligations under the newEU Directives – 13 September 2002, a non-discrimination obligation seeks toprevent a vertically integrated SMP operator acting in such a way as to have amaterial adverse effect on competition. In the absence of an obligation not tounduly discriminate the MNOs may offer different terms to different purchasers oftheir voice call termination services that have anti-competitive effects. The Directortherefore considers that an obligation not to unduly discriminate is appropriate.

6.51 However, while it appears to the Director that discriminating between differentPECN providers in the supply of mobile voice call termination services would havea material adverse effect on competition, he has not yet formed an opinion onwhether a MNO providing voice call termination services to itself on different termswould constitute undue discrimination, as it is not clear whether suchdiscrimination would have a material adverse effect on competition. Oftel will beconsidering this further during the consultation period, and welcomes the views ofstakeholders on this issue.

6.52 As explained above 3G services are new and evolving services and for thereasons set out above, the Director generally seeks not to regulate such services.A requirement not to unduly discriminate in relation to 3G voice call terminationservices is therefore considered to be disproportionate.

6.53 The Director’s initial view is that a requirement not to unduly discriminate inthe provision of access should apply to the MNOs to the extent that they areproviding 2G voice call termination services. The provision of 3G terminationservices should not, however, be subject to such a requirement for the reasons setout above.

6.54 For the reasons set out above, the Director’s initial view is that:

70

• the MNOs should be subject to a requirement not to unduly discriminatebetween different PECN providers in the provision of 2G voice call terminationservices; but

• the MNOs should not be subject to such a requirement in respect of their 3Gvoice call termination services.

6.55 The Director will consider the issue of applying the requirement not to undulydiscriminate in the provision of 2G voice call termination services by an MNO toitself during this period of consultation.

• Charge controls

6.56 In Chapter 4, the Director sets out his initial view that each MNO has amonopoly of voice call termination on its own network(s) and that the Directorconsiders that MNOs are likely to continue to have SMP for at least the next threeyears.

6.57 As discussed in Chapter 4, it appears to the Director that there are insufficientconstraints on an MNO’s ability to raise termination charges above the competitivelevel. In the absence of regulation, the MNOs would have the incentive to set aprofit-maximizing price for the provision of wholesale voice call termination. Thedetriments of this are discussed in greater detail in Chapter 5.

6.58 However, such a requirement on 3G voice call termination services may beconsidered disproportionate. 3G retail services are new and innovative and assuch are only available to very few subscribers at the moment, although this maybe expected to change over the period of this review. As stated elsewhere in thisdocument regulation of such services is likely to have an adverse impact oncontinued investment in to new 3G services, which may limit consumer choice inthe longer term. The Director currently therefore considers that it would bedisproportionate to make termination on 3G networks subject to charge controls.

6.59 However a charge control may be appropriate on the MNOs’ 2G terminationservices, given the risk that excessive prices will be set due to a lack ofcompetition in the market. To deal with this a charge control with transparent, easyto monitor compliance conditions can help ensure that firms do not abuse theirdominant position and that competition develops to the benefit of consumers.

6.60 The two main forms of charge regulation are an RPI-X control and a rate ofreturn control. Under the former, the firm is prevented from increasing prices bymore than inflation minus X percent per annum. Under the latter arrangement, thefirm is required to earn no more than the specified rate of return in each year. Thisis a form of cost plus regulation and requires the level of allowable costs to beassessed annually.

6.61 RPI-X regulation has a number of advantages over rate of return control. Inparticular, it avoids overly intrusive and bureaucratic regulation. Price controls are

71

not set every year according to the rate of return earned in the year, but for aperiod of years, and they are not revisited until the end of the period.

6.62 By contrast, rate of return control is burdensome, requiring detailedassessments of costs on an annual basis.

6.63 In the Director’s view, a price control if appropriately designed is aproportionate response where competitive pressures and other regulation is nototherwise sufficient to ensure that prices are cost reflective. In this case the chargecontrol would address SMP and the lack of competitive pressures to constraincharges to the competitive level and the material adverse effects this would haveon end-users.

6.64 For the reasons discussed, the Director’s initial view is that only the MNOs'charges for 2G voice call termination services should be subject to a form of RPI-Xcharge control. Details of this control are discussed in Chapter 7.

Bilateral agreement of mobile-to-mobile voice call termination charges

6.65 It has been argued that Oftel’s conclusion that, absent regulatory intervention,each MNO would unilaterally set its off-net voice call termination charges abovecosts depends on each MNO being able to set its own termination charge byunilateral decision and change its termination charge at will. It has been suggested(paragraph 2.479 to 2.485 of the CC report) that the outcome would be different ifpairs of competing MNOs entered into bilateral agreements setting their off-netvoice termination charges simultaneously and with no possibility, once theagreement is signed, of either of the MNOs changing its termination chargeunilaterally. Hence, an alternative to the charge control, albeit restricted to mobile-to-mobile off-net calls, might be to impose on the MNOs the obligation to engagein bilateral negotiations to set off-net termination charges. This remedy would notapply to the setting of fixed-to-mobile termination charges because the incentivesfaced by the MNOs in determining the two sets of charges are different. MNOscompete with each other for customers, hence the call termination charges set byeach MNO affect the costs of its competitors, and hence their ability to compete inthe retail market.

6.66 The reason why bilateral negotiations could be effective in constraining thelevel of the off-net voice termination charges is that they alter the incentives facedby the MNOs. If an MNO sets its voice call termination charges unilaterally, it hasthe incentive to set its charges at a high level because, by doing so, it can expect:

(a) to generate profits on voice call termination services to fund competition inthe retail market; and

(b) to increase the costs of its rival MNOs in providing off-net outbound calls.

6.67 In addition, the MNO will consider that, if it does not set its own off-net voicetermination charges at a high level, the other MNOs will set their charges at a highlevel, in which case it will be at a competitive disadvantage.

72

6.68 If reciprocal voice call termination charges are set through bilateralagreements, an MNO cannot gain an advantage over the other MNOs by alteringits termination charges. An MNO can only change its termination charges byagreeing a variation to the agreement with the other MNO. Hence, if an MNO triedto increase its voice call termination charges, the termination charges of itscompetitor would also increase simultaneously. Any change by one MNO to itsvoice call termination charges would, therefore, also affect its cost of terminatingcalls on its rivals’ networks. This removes the incentive for any MNO to raise itsvoice call termination charges. Economic literature on the issue has beenproduced to support this claim.

6.69 However, the Director considers it would be unsafe to rely on this argument,even though there it has a degree of support in the economic literature, todetermine how and whether to regulate off-net voice call termination charges. Theconclusions of the economic literature are not unambiguous and there are alsomodels that yield above-cost termination charges. In the Director’s view thecurrent economic literature is insufficiently robust to conclude that no directregulation of off-net termination charges is required.

6.70 Furthermore, the Director believes that this is a developing literaturecomposed of a small number of simplified and stylised models which abstractsfrom issues of strategic entry deterrence and relies on a number of simplifyingassumptions, such as symmetry of traffic flows between MNOs.

6.71 If some of assumptions are relaxed, the results of these models changemarkedly. For example, if the traffic is not in balance between each pair of MNOs(as it currently is the case), the actual outcome of a bilateral negotiation willdepend on the relative bargaining power of the two MNOs. The presence of trafficimbalances creates incentives for the MNO with a surplus of incoming calls overoutgoing calls to raise termination charges above cost and for the MNO with adeficit to reduce them. This tension between the operator in surplus and theoperator in deficit may create a degree of ‘stickiness’ in the movement of voice calltermination charges from any starting position, because of their opposing interests.The Director considers that any proposed remedy should be robust whether thereare traffic imbalances or not, and it is not clear that this proposal satisfies thiscondition.

6.72 In addition, bilateral agreements would not work when there is a new entrant(such as 3) because, each of the incumbent MNO may find it mutually beneficial tokeep off-net termination charges high to create an entry barrier for the newentrant.

6.73 The Director's views are discussed in more detail in the Oftel papersTermination of off-net mobile to mobile calls(www.oftel.gov.uk/publications/mobile/ctm_2002/offnet0602.pdf) and BilateralAgreement of Mobile-to-mobile Call Termination Charges.

73

6.74 Hence, the Director considers that the analysis that underlies this proposal isincomplete and not robust. He believes that the MNOs’ incentives are not alignedwith the interests of consumers, even when the MNOs are forced to enter intobilateral negotiations.

6.75 In conclusion, the Director is of the view that bilateral negotiations would notavoid the need to regulate mobile-to-mobile voice termination charges. This viewis supported by the CC (See paragraph 2.479 of the CC Report).

Option D

A requirement to secure transparency through publication of prices and areference offer; a requirement to provide mobile voice call termination onreasonable request and on fair and reasonable terms; a requirement not tounduly discriminate in the provision of such access; imposition of chargecontrols; a requirement to maintain cost-accounting systems; a requirementto set prices on the basis of forward looking long-run incremental costs; anda requirement for separate accounting systems.

6.76 This option includes the transparency requirements set out in Option B, arequirement to publish a Reference Offer; the access, no undue discriminationobligations and charge controls set out in Option C, plus a requirement to maintaincost-accounting systems; a requirement to set prices on the basis of forwardlooking long-run incremental costs; and a requirement for separate accountingsystems.

• Obligation to set charges on the basis of forward looking long-run incrementalcosts

6.77 In effectively competitive markets, the price of services would be driven downto competitive levels by commercial pressures and the entry and exit of firms, iethe competitive process. However, in markets where competition does not providesufficient pricing constraints, it can be necessary to apply ex-ante regulation inorder to ensure that prices reflect costs.

6.78 However, while these incentives exist, Option C, which includes chargecontrols, may sufficiently deal with these concerns, because the charge controlwould reduce the level of charges towards costs over a set period of time. Theneed to have an additional constraint on charges may therefore be consideredunnecessary in this instance.

6.79 Equally, such a requirement on 3G voice call termination services may beconsidered disproportionate. 3G retail services are new and innovative and asstated elsewhere in this document regulation of such services is likely to have anadverse impact on continued investment into the delivery of new 3G services,which may limit consumer choice in the longer term.

74

6.80 For these reasons, the Director’s initial view is that it is unnecessary torequire that the MNOs set charges for 2G or 3G voice call termination services onthe basis of forward looking long run incremental costs.

• Obligation to publish separate accounts for network and retail activities for theprovision of voice termination services and to maintain appropriate costaccounting systems

6.81 Obligations to maintain appropriate cost accounting systems and toseparately account for termination services are used to monitor compliance withother regulatory obligations. Appropriate cost accounting systems ensure that it ispossible to demonstrate compliance with obligations to set charges on forwardlooking long run incremental costs. In the absence of an obligation on the MNOs toset charges in this way, an obligation for cost accounting systems is unnecessaryand disproportionate.

6.82 An obligation to account separately for voice termination services may berequired to monitor compliance with an obligation to not unduly discriminate.However, as set out above, it is not clear to Oftel whether discrimination of thetype that would necessitate such an accounting obligation constitutes unduediscrimination. Without the need to explicitly monitor such discrimination anobligation to separately account for voice termination would be unjustified.

6.83 For the reasons set out above the Director’s initial view is that obligations topublish separate accounts for network and retail activities for the provision of voicetermination services and to maintain appropriate cost accounting systems areunnecessary.

Comparison of Options

6.84 Oftel has considered 4 different options for regulating mobile voicetermination services on each of the MNOs.

6.85 3G retail services are new and innovative services and as such their take-upis uncertain and it remains unclear how they will develop and at what speed. TheDirector does not seek to regulate new and innovative services in advance of theirlaunch or soon after launch, as these factors remain uncertain. Regulation at anearly stage may adversely affect the continued development and investment insuch new and innovative services. For these reasons the Director’s initial view isthat Option A, i.e. no ex-ante regulation, is appropriate in respect of the MNOs 3Gvoice call termination services.

6.86 The obligations set out in option B would ensure transparency for purchasersof wholesale termination services, and provide advance warning of chargechanges to providers purchasing wholesale termination services. While useful,these obligations are insufficient to deal with the problems identified. In particular,

75

they would not constrain the MNOs’ ability to set excessive charges for their 2Gvoice call termination services.

6.87 For these reasons the combination of obligations included in Option C wouldbe more appropriate for the supply of 2G termination services by the MNOs. Thecharge controls would address the problem of MNOs ability to set excessivecharges. The other elements of Option C would ensure that the MNOs providedthe 2G mobile voice termination services on reasonable request and on fair,reasonable, and transparent terms, and that they would be unable to undulydiscriminate between wholesale customers.

6.88 Option D would require the MNOs to provide mobile voice call terminationservices on fair, reasonable, and transparent terms, prevent undue discriminationin the provision of these services, require that prices be set on the basis of forwardlooking long-run incremental costs, require the publication of separate accounts fornetwork and retail activities for the provision of voice termination services andmaintenance of appropriate cost accounting systems. However, for the reasonsset out above, Oftel considers that these additional requirements are unnecessarygiven the options set out in Option C that, in the Director’s initial view adequatelydeal with the problems that have been identified.

6.89 For the reasons set out above the Director considers that:• No ex-ante regulation, as set out in Option A, should be applied to the MNOs

in relation to the provision of their 3G voice call termination services.• The appropriate obligations that should apply to the MNOs in relation to the

provision of their 2G voice call termination services are as set out in Option C.

6.90 Clause 44 requires conditions to be justifiable, non-discriminatory,proportionate and transparent. The proposed conditions are objectively justifiable,in that they relate to the need to ensure that competition is not distorted to thedetriment of consumers. They do not unduly discriminate, in that they are imposedon all providers of the same services where the competitive and economicconditions under which these services are provided are similar. They areproportionate, since they only apply the minimum regulation necessary to securethe desired outcome and they are transparent as to the effect that they areintended to achieve given the explanation in this document.

Question 8

Do you agree that Option A is the most appropriate and proportionateregulation of 3G mobile voice call termination services terminated on:• Vodafone’s 3G network?• T-Mobile’s 3G network?• O2’s 3G network?• Orange’s 3G network?

Question 9

76

Do you agree that Option C is the most appropriate and proportionateregulation of:• Vodafone’s provision of mobile voice termination services terminated on

its 2G network?• T-Mobile’s provision of mobile voice termination services terminated on

its 2G network?• O2’s provision of mobile voice termination services terminated on its 2G

network?• Orange’s provision of mobile voice termination services terminated on its

2G network?

Detail of the proposed remedies for the MNOs in respect of 2G mobile voicecall termination on their 2G networks

Requirement to notify charges (proposed Conditions MD6 and MC 6)

6.91 As explained in the option appraisal, the Director’s initial view is that theMNOs should be required to give 28 days notice before they are permitted tochange any charge for their 2G mobile voice call termination services. This isreflected in the proposed Conditions MD6 and MC6.

6.92 The Director has considered how much notice should be required to be givenbefore the MNOs can change their charges for 2G mobile voice call terminationservices and, as a result, believes that a notification period of 28 days isappropriate, in that it is broadly consistent with current contractual terms andallows sufficient time for changes in the wholesale charge to be reflected in otherproviders' retail charges while not unduly impacting on the ability of the MNOs tochange their wholesale charges.

6.93 The Director therefore proposes that the MNOs should be required to give 28days notice of changes in charges for their 2G mobile voice call terminationsservices.

6.94 The notice should include the following information:• a description of the Network Access in question;• a reference to the location in the Dominant Provider’s current Reference Offer

of the terms and conditions associated with the provision of that NetworkAccess;

• the date on which or the period for which any amendments to charges will takeeffect (the "effective date"); and

• the relevant network tariff gradient.

6.95 The Director considers that the proposed conditions meet the tests set out inthe Communications Bill. The Director has considered all the Community

77

requirements set out in Clause 4 and in particular the requirement to promotecompetition and to secure efficient and sustainable competition.

6.96 The Director considers that, in accordance with the tests set out in Clause44(2) of the Communications Bill, the proposed conditions are objectivelyjustifiable, non-discriminatory, proportionate, and transparent. They are objectivelyjustifiable, in that the benefits of publication and notification of charges outweighany possible disadvantages. They are proportionate, in that only information that isnecessary to other network providers needs to be notified. They do notdiscriminate unduly against the MNOs in that the same condition is imposed on allproviders of the same services where the competitive and economic conditionsunder which these services are provided are similar and they are transparentgiven the explanation in this consultation document.

Requirement to publish a reference offer (proposed Conditions MD5 andMC5)

6.97 As explained in the option appraisal, the Director’s initial view is that theMNOs should publish a reference offer (RO) for their 2G mobile voice calltermination services. This is reflected in proposed Conditions MD5 and MC5.

6.98 The published RO must set out such matters as:• a clear description of the services on offer;• terms and conditions including charges and ordering, provisioning, billing

and dispute resolution procedures;• information relating to technical interfaces and points of interconnection; and• conditions relating to maintenance and quality.

6.99 The Director considers that the proposed conditions meet the tests set out inthe Communications Bill. The Director has considered all the Communityrequirements set out in Clause 4 and in particular the requirement to promotecompetition and to secure efficient and sustainable competition.

6.100 The Director considers that the proposed conditions meet the tests set out inClause 44(2) of the Communications Bill. They are objectively justifiable in that itrequires that terms and condition are published in order to prevent unduediscrimination and provide stability in markets. They are proportionate, in that onlyinformation that is necessary to ensure that that there is no material adverse effecton competition is required to be provided. They do not discriminate in that they areimposed on all providers of the same services where the competitive andeconomic conditions under which these services are provided are similar and theyare transparent as to what it is intended to achieve given the explanation in thisdocument.

Requirement to provide network access (proposed Conditions MD1 andMC1)

78

6.101 As explained in the option appraisal, the Director’s initial view is that theMNOs should be required to meet all reasonable requests for network access, i.e.in this context 2G mobile voice call termination. Oftel also believes that such 2Gmobile voice call termination should be provided on fair and reasonable terms.These obligations are reflected in proposed Conditions MD1 and MC1.

6.102 The Director considers that the proposed conditions meet the tests set out inthe Communications Bill. The Director, in proposing this condition has consideredall the Community requirements set out in Clause 4 and in particular therequirement to promote competition, to further the interests of consumers, and topromote effective and sustainable competition.

6.103 The Director considers that, in accordance with the tests set out in Clause44(2) of the Communications Bill, the proposed conditions are objectivelyjustifiable, non-discriminatory, proportionate, and transparent. The proposedconditions are objectively justifiable, in that they relate to the need to ensure thatcompetition is not distorted to the detriment of consumers. They do not undulydiscriminate, in that they are imposed on all providers of the same services wherethe competitive and economic conditions under which these services are providedare similar. It is proportionate, since they only apply the minimum regulationnecessary to secure the desired outcome and they are transparent as to the effectthat they are intended to achieve given the explanation in this document.

6.104 The Director also believes that the proposals to require the MNOs to providenetwork access meets the tests set out in Clause 84(4) of the Communications Billin that, amongst other things, it is feasible for the MNOs to provide such networkaccess and the absence of such access would have a detrimental effect oncompetition.

Requirement not to unduly discriminate (proposed Conditions MD2 andMC2)

6.105 As explained in the option appraisal, the Director’s initial view is that theMNOs should be prevented from unduly discriminating in the provision of 2Gmobile voice call termination services. This is reflected in proposed ConditionsMD2 and MC2.

6.106 An obligation of non-discrimination might have disadvantages if it preventeddiscrimination that was economically efficient or justified. However, the proposedcondition provides that there should be no undue discrimination. Oftel hasconsidered how it might treat undue discrimination in its statement Imposingaccess obligations under the new EU Directives, September 2002. The statementnotes that any obligation with respect to undue discrimination has the objective ofpreventing behaviour that has a material adverse effect on competition. This doesnot mean that there should not be any differences in treatment betweenundertakings, rather that any differences should be objectively justifiable, forexample, by differences in underlying costs of supplying different undertakings.

79

6.107 The Director considers that the proposed condition meets the tests set out inthe Communications Bill. The Director in proposing this condition has consideredall the Community requirements set out in Clause 4 and in particular therequirement to promote competition and to secure efficient and sustainablecompetition.

6.108 The Director considers that, in accordance with the tests set out in Clause44(2) of the Communications Bill, the proposed conditions are objectivelyjustifiable, non-discriminatory, proportionate, and transparent. The Directorconsiders that the proposed conditions are objectively justifiable to prevent anydiscrimination that has a material adverse affect on competition. They do notunduly discriminate, in that they are imposed on all providers of the same serviceswhere the competitive and economic conditions under which these services areprovided are similar. They are proportionate, since they only apply the minimumregulation necessary to secure the desired outcome and they are transparent as tothe effect that they are intended to achieve given the explanation in this document.

Charge controls (proposed Conditions MD3, MD4 and MC3, MC4)

6.109 As explained in the option appraisal, the Director’s initial view is that pricesfor the MNOs' 2G mobile voice call termination services should be subject to acharge control designed to ensure that prices are reduced from current levels andbetter reflect costs. This is reflected in proposed Conditions MD3, MD4, and MC3and MC4.

6.110 The Director considers that the proposed conditions on the MNOs meet thetests set out in the Communications Bill. The Director has considered all theCommunity requirements set out in Clause 4 and in particular the requirement topromote competition, to secure efficient and sustainable competition and to securethe maximum benefits for persons who are customers of communicationsproviders.

6.111 The Director considers that, in accordance with the tests set out in Clause44(2) of the Communications Bill, the proposed conditions are objectivelyjustifiable, non-discriminatory, proportionate, and transparent. They areobjectively justifiable in that the benefits of RPI-X price controls are widelyacknowledged as an effective mechanism to reduce prices in a situation wherecompetition does not act to do so. Although charge controls are only proposed inrespect of the MNOs 2G voice call termination services, they are not discriminatoryfor the reasons set out in the option appraisal. The conditions are proportionatesince they are the most effective mechanism to reduce charges from their currentlevels, and they are transparent in that they are clear in their intention to controltermination charges.

6.112 The Director considers that the tests in Clause 85 have been met, namely topromote efficiency, to promote sustainable competition and to confer the greatestpossible benefit on end-users. As noted in paragraph 5.60, there is a risk that, insituations where SMP is persistent, pricing will be distorted and not at competitive

80

levels. The proposed condition is necessary in order to promote competition and toprovide benefits to end users in that it acts to reduce charges for wholesale inputsto retail prices in the absence of competition reducing prices.

6.113 Clause 85 requires the extent of the investment of the dominant provider tobe taken into account when setting a charge control condition. The extent of theMNOs’ investment is taken into account since the methodology employed allowsfor the cost of capital in relation to investments made in the past as well asefficient investments necessary in the future.

Question 10

Do you have any comments on the justification or detail of the proposedremedies on the MNOs as set out above?

81

What are the options for regulation on 3?

6.114 3 is a new entrant into the UK mobile market, with a small subscriber baseand consequently a much larger proportion of off-net traffic than the other MNOs.This places 3 in a different position from the MNOs and determines why theproposed regulatory obligations differ.

6.115 As 3 is currently rolling out its 3G network it does not currently have fullgeographic coverage for its 3G services. However, 3 does offer national coverageof services through a roaming agreement with one of the MNOs, which enables itto terminate mobile calls on the MNO’s 2G network when its subscribers areoutside of its 3G network coverage area. The roaming agreement allows 3 to re-sell the voice call termination services of the 2G operator to its subscribers. 3therefore, provides both 3G and 2G voice call termination services.

6.116 The various options considered below relate to options for regulation for 3’s2G and 3G mobile voice call termination services.

• Option AA

No ex-ante regulation;

• Option BB

A requirement to set prices on the basis of forward looking long-run incrementalcosts.

• Option CC

Option BB plus, a requirement to secure transparency through publication ofprices and prior notification of price changes, a requirement to make transparentthe charges, terms and conditions through publication of a reference offer; arequirement to provide mobile voice call termination on reasonable request and onfair and reasonable terms; and a requirement not to unduly discriminate in theprovision of the services.

• Option DD

Options BB and CC plus, a requirement to have cost-accounting systems in place;a requirement for separate accounting systems, and imposition of charge controls.

6.117 The Director's initial view, for the reasons given below, is that:• 3’s 3G voice call termination services should not be subject to ex-ante

regulatory controls as set out in Option AA; and• The appropriate obligations that should apply on 3 in relation to the provision

of 2G termination services to its subscribers are those set out in Option BB.

82

A fuller analysis of each of the options is provided below.

Option AA

No ex-ante regulation

6.118 Where markets are effectively competitive, ex post competition law issufficient to deal with any competition abuses that may arise.

6.119 Although 3 does not have a 2G mobile network, because of its roamingagreement it is able to re-sell 2G voice call termination services. In the absence ofregulation, 3 would have both the ability and incentive to add a profit maximisingmark-up to the charges for the 2G services that it re-sells. This is due to the lackof competitive constraints discussed in Chapter 4. The effect of this would be bothto disadvantage consumers and to distort competition in the mobile market, byallowing 3 to gain more net revenue from 2G termination services than would beavailable to its competitors. The Director does not believe that the application ofex post competition law powers would be sufficient to prevent 3 from seeking toapply profit maximising mark-ups for the 2G voice call termination services that itre-sells. He therefore considers that, in order to ensure that consumers are notdisadvantaged and that competition is not distorted, it would be appropriate andproportionate to impose restrictions, using ex ante powers, on the level of mark upthat 3 can impose when it provides 2G call termination services under its roamingagreement.

6.120 3G services are new and innovative services, where take-up is uncertain,and costs are unclear. In general, and consistent with what has been stated in thepast, the Director does not normally apply regulatory controls to such new servicesas regulation could deter continued investment and development of new services,ultimately disadvantaging consumers. Furthermore, it is also presently unclearhow 3G voice call termination charges will develop and Oftel will be seekingfurther information on this during this consultation. In the absence of such chargeinformation and as 3G services are new and innovative services to whichregulatory controls might have a detrimental affect, on balance the application ofregulation appears disproportionate.

6.121 Therefore, for the reasons, set out above, the Director’s initial view is thatapplication of ex ante regulation is disproportionate in relation to 3’s 3G voice calltermination services.

Option BB

A requirement to set prices on the basis of forward looking long-runincremental costs.

• Obligation to set charges on the basis of forward looking long-run incrementalcosts

83

6.122 In effectively competitive markets, the price of services would be drivendown to competitive levels by commercial pressures and the entry and exit offirms, ie the competitive process. However, in markets where competition does notprovide sufficient pricing constraints, it can be necessary to apply ex-anteregulation in order to ensure that prices reflect costs.

6.123 However, such a requirement on 3G termination services may beconsidered disproportionate. As discussed above 3G retail services are new andinnovative services and take-up is uncertain. In general the Director considers itinappropriate to regulate new and innovative services because of the negativeimpact that such regulation may have on the incentive to invest in new areaswhich, in the longer term would have an adverse affect on consumer choice.

6.124 Nevertheless, because of its roaming agreement 3 is able to re-sell 2Gvoice call termination services. This puts 3 in a different position to the otherMNOs who the Director proposes should be subject to regulatory controls thatreduce the charges for their 2G voice call termination services. The regulatedcharge for these 2G termination services would form the main input into the costsfor 3 in re-selling these 2G termination services. However, 3 will be free to add amark-up on to the regulated charge for the 2G termination services that it re-sells.In order that competition in the retail market should not be distorted, this mark-upshould reflect only efficiently incurred forward looking long run incremental costs(LRIC) plus an appropriate rate of return on the capital investment and anappropriate mark-up for the recovery of common costs.

6.125 A requirement to set charges on a LRIC basis would meet the need toconstrain charges while ensuring that 3 was able to cover the costs, including anappropriate rate of return, of re-selling 2G voice call termination services.

6.126 For these reasons the Director’s initial view is that a requirement to setcharges on a forward looking long-run incremental basis for the 2G terminationservices that 3 re-sells is proportionate and should be applied. However, for thereasons set out above, this requirement should not apply to 3’s 3G terminationservices.

Option CC

A requirement to secure transparency through publication of prices andprior notification of price changes, a requirement to secure transparencythrough publication of prices and prior notification of price changes and arequirement to set prices on the basis of forward looking long-runincremental costs, plus a requirement to make transparent the charges,terms and conditions through publication of a reference offer; a requirementto provide mobile voice call termination on reasonable request and on fairand reasonable terms; and a requirement not to unduly discriminate in theprovision of the service.

84

6.127 This option includes the requirement to set charges on the basis of forwardlooking long run incremental costs set out in Option BB plus, a requirement tosecure transparency through publication of prices and prior notification of pricechanges; a requirement to publish charges and changes to those charges inadvance; a requirement to make transparent the charges, terms and conditionsthrough publication of a reference offer; a requirement to provide mobile voice calltermination on fair and reasonable terms; and a requirement not to undulydiscriminate in the provision of the services.

• Obligation to publish charges and proposed changes to those charges inadvance of them taking place

6.128 The requirement to publish charges and changes to those charges inadvance has two main purposes: (a) to assist transparency for the monitoring ofpotential anti-competitive behaviour and (b) to give advanced warning of chargechanges to providers purchasing wholesale termination services.

6.129 Publication of charges is particularly important where there is a requirementto provide access for the purposes of supplying voice call termination services.Where there is no such obligation the benefits of price publication are less clearalthough the publication of charges does assist general transparency. However,the Director considers that as a new entrant, at present a requirement on 3 topublish prices would be disproportionate and that an obligation to set charges onthe basis of forward-looking LRIC should be sufficient regulatory control on 3’scharges.

6.130 In relation to 3’s 3G voice call termination services, price publication may beunwarranted. 3G retail services are new and innovative services and as such theirtake-up is uncertain and it remains unclear how they will develop and at whatspeed. The Director does not seek to regulate new and innovative services inadvance of their launch or soon after launch, as these factors remain uncertain.This is so as not to adversely affect the continued development and investment insuch services.

6.131 For these reasons, the Director’s initial view is that it is not necessary torequire 3 to publish its charges for 2G or 3G voice call termination services.

• Obligation to meet reasonable requests for 2G mobile voice call terminationservices on reasonable request and on fair and reasonable terms

6.132 Without an obligation on the mobile operators to provide access to mobilevoice call termination services on fair and reasonable terms, competition indownstream markets could be severely distorted by the mobile operators refusingto supply services or by doing so on terms that were unfair and unreasonable. Forexample, mobile operators could refuse to supply termination services or only doso on terms that were unfair and unreasonable to some fixed or mobile operators.This would be a particular problem if it were to be directed towards 3, which, as a

85

new entrant to the market lacks counter-veiling buyer power because of its smallpresence in the market. In the long term, this is likely to result in fewer competitorsand therefore less choice.

6.133 However, 3 has a small subscriber base and consequently a much largerproportion of off-net traffic than the other MNOs. As a consequence 3 doesn’t havethe same incentives as the MNOs to not offer end to end connectivity. In order tobe successful 3 needs to increase its market share and to do this it will need toensure end-to-end interoperability. The incentives described in the paragraphabove do not therefore apply to 3.

6.134 For the reasons described above the Director’s initial view is that 3 shouldnot be subject to a requirement to provide voice call termination services onreasonable request.

• Obligation to publish a reference offer

6.135 A requirement to publish a Reference Offer (RO) has two main purposes:(a) to assist transparency in monitoring potential anti-competitive behaviour; and(b) to provide visibility of the terms and conditions on which other providers canpurchase services.

6.136 Publication of a RO will ensure stability and certainty to operatorspurchasing mobile voice call termination services. In addition it may enablenegotiations between parties to be completed more quickly, help to avoid disputes,and give confidence to purchasers of such services confidence that those servicesare provided on non-discriminatory terms.

6.137 Publication of a RO is particularly important where there is also arequirement to provide access for the purposes of providing voice call terminationservices. Where there is no obligation to provide access the benefits of publishinga RO are significantly less as there is no obligation to provide access inaccordance with what is in the RO, although it may still be beneficial for improvinggeneral transparency.

6.138 The Director’s initial view therefore is that in the absence of a requirement toprovide network access an obligation to publish a RO is unnecessary as anybenefits would be unjustified.

• Obligation not to unduly discriminate in the provision of network access

6.139 An obligation not to unduly discriminate would be particularly important toprevent the mobile operators discriminating in the supply of mobile terminationservices between different purchasers of such services.

6.140 However, as outlined above, because of its position in the market theincentives on 3 are not the same as on the MNOs, and application of a

86

requirement not to unduly discriminate appears disproportionate. Furthermore, inthe absence of a requirement to provide network access a requirement not tounduly discriminate appears unnecessary.

6.141 For these reasons, the Director’s initial view is that a requirement not tounduly discriminate in the provision of network access should not apply to 3’svoice call termination services.

Option DD

A requirement to secure transparency through publication of prices and areference offer; a requirement to provide mobile voice call termination onreasonable request and on fair and reasonable terms; a requirement not tounduly discriminate in the provision of such access; a requirement to setcharges on the basis of forward looking long-run incremental costs, plusimposition of charge controls; a requirement to maintain cost-accountingsystems; and a requirement for separate accounting systems.

6.142 This option includes the requirement to set charges on the basis of forwardlooking long run incremental costs set out in Option BB, the network access, noundue discrimination obligations and requirement to publish prices and an RO setout in Option CC, plus a requirement to maintain cost-accounting systems; arequirement for separate accounting systems; and imposition of charge controls.

• Obligation to publish separate accounts for network and retail activities for theprovision of voice call termination services and to maintain appropriate costaccounting systems

6.143 Obligations to maintain appropriate cost accounting systems and toseparately account for termination services are used to monitor compliance withother regulatory obligations. Appropriate cost accounting systems ensure that it ispossible to demonstrate compliance with obligations to set charges on a LRICbasis.

6.144 However, in this case it may be disproportionate to make 3 subject to arequirement to maintain special cost accounting systems solely for the purposes ofmonitoring that its 2G termination services are charges on a LRIC plus basis. Sucha requirement is burdensome and as such it is appropriate to consider whetherthere are other ways to monitor compliance with the proposed LRIC plusobligation.

6.145 As discussed above, based on the proposals of the Director, the major inputto the 2G termination services would be the regulated 2G termination charge that3 pays to the MNO with whom it has a roaming agreement.

87

6.146 The Director considers that it should be possible to monitor the LRICobligation on the 2G termination service re-sold by 3 without the need for a fullcost accounting obligation.

6.147 An obligation to account separately for voice call termination services maybe required to monitor compliance with an obligation to not unduly discriminate.However, as set out above, an obligation not to unduly discriminate in theprovision of network access is considered unnecessary and disproportionate in thecase of 3. In the absence of such an accounting separation obligation is notrequired and therefore unjustified.

6.148 For the reasons set out above the Director’s initial view is that obligations topublish separate accounts for network and retail activities for the provision of 3’svoice call termination services and for 3 to maintain appropriate cost accountingsystems are unnecessary.

• Charge controls

6.149 The incentives on MNOs to set excessive prices and how the incentives on3 to do so differ are discussed above. For these reasons, the Director considersthat the proposed obligation to set charges on a LRIC plus basis are sufficient tocontrol the charges for the 2G voice call termination services that 3 re-sells. TheDirector also considers that regulation in respect of 3’s 3G termination services isdisproportionate given the new and innovative nature of 3G services.

6.150 For these reasons, the Director’s initial view is that charge controls on the3’s voice call termination services would be disproportionate.

Comparison of Options

6.151 Oftel has considered 4 different options for regulating mobile voice calltermination services on 3.

6.152 3G retail services are new and innovative services and as such their take-upis uncertain and it remains unclear how they will develop and at what speed. TheDirector does not seek to regulate new and innovative services in advance of theirlaunch or soon after launch, as these factors remain uncertain. Regulation at anearly stage may adversely affect the continued development and investment insuch new and innovative services. For these reasons the Director’s initial view isthat Option AA, no ex-ante regulation is appropriate in respect of 3’s 3G voice calltermination services.

6.153 In respect of voice call termination services provided by to 3 to itssubscribers, the LRIC plus obligation would ensure that the mark-up applied to the2G terminations services reflected only efficiently incurred costs plus anappropriate rate of return on capital investment and the recovery of common costs.

88

For these reasons, the Director’s initial opinion is that the obligation set out inOption BB adequately addresses the problems that have been identified.

6.154 As, for the reasons discussed above, the obligation set out in Option BB isconsidered in the Director’s initial opinion adequate to address the problemsidentified, the additional obligations set out in Options CC and DD areunnecessary and would be disproportionate.

6.155 The Director considers therefore that:• No ex-ante regulation, as set out in Option AA, should be applied to 3 in

relation to the provision of its 3G voice call termination services.• The appropriate obligations that should apply to 3 in relation to the provision of

its 2G voice call termination services is as set out in Option BB.

6.156 Clause 44 requires conditions to be justifiable, non-discriminatory,proportionate and transparent. The proposed condition is objectively justifiable, inthat it relates to the need to ensure that competition is not distorted to thedetriment of consumers. It does not unduly discriminate, in that it is imposed on allproviders of the same services where the competitive and economic conditionsunder which these services are provided are similar and reflects the uniqueposition of 3 re-selling 2G voice call termination services. It is proportionate, sinceit only applies the minimum regulation necessary to secure the desired outcomeand it is transparent as to the effect that it is intended to achieve given theexplanation in this document.

Question 11

Do you agree that Option AA is the most appropriate and proportionateresponse to the provision of 3G mobile voice call termination servicesterminated on 3’s 3G network?

Question 12

Do you agree that Option BB is the most appropriate and proportionateresponse to 3’s provision of 2G voice call termination?

Detail of the proposed remedies for 3 in respect of its 2G mobile voice calltermination services

Basis of charges (proposed Condition MA1)

6.157 As explained in the option appraisal, the Director believes that in manycommunications markets ex-ante regulation should require charges to be set onthe basis of LRIC plus an appropriate mark-up for costs which are common acrossproducts. Charges should also allow for an appropriate return on capital. For thesame reasons as those set out in the option appraisal, the Director believes that

89

3’s charges for the 2G mobile voice call termination services that it re-sells shouldbe based on its LRIC and allow for an appropriate mark-up for the recovery ofcommon costs and an appropriate return on capital employed. This is reflected inproposed Condition MA1.

6.158 The Director considers that the proposed condition meets the tests set out inthe Communications Bill. The Director has considered all the Communityrequirements set out in clause 4 including, in particular, the requirement topromote competition and to secure efficient and sustainable competition. TheDirector believes the proposals meet these requirements, as, under the proposals,3 will be permitted to recover its costs and earn a reasonable return in theprovision of its 2G voice call termination services.

6.159 The Director considers that, in accordance with the tests set out in Clause44(2) of the Communications Bill, the proposed condition is objectively justifiable,non-discriminatory, proportionate, and transparent. The proposed condition isobjectively justifiable, in that it relates to the need to ensure that competition is notdistorted to the detriment of consumers. It does not unduly discriminate, in that itreflects the unique position of 3 as described in detail in the option appraisal. It isproportionate since it only applies the minimum regulation necessary to secure thedesired outcome and it is transparent as to the effect that it is intended to achievegiven the explanation in this document.

6.160 The Director considers that the tests in Clause 85 have been met. As notedabove, in markets where competition does not provide sufficient pricingconstraints, there is a risk that charges will be distorted and not at competitivelevels, it may therefore be necessary to apply ex-ante regulation in order to ensurethat prices reflect costs. The proposed condition is necessary in order to ensurethat the mark-up applied by 3 to the (proposed) 2G termination charge are basedon long run incremental costs with an appropriate rate of return on investment andan appropriate mark-up for the recovery of common costs. This will benefit end-users as it will ensure that the mark-up is not excessive, reflecting the charges thatmight be found in a competitive market.

6.161 The extent of investment of the dominant operator has been taken intoaccount as set out in Clause 85(2) and the proposed conditions provide for amark-up for an appropriate return on capital employed.

Question 13

Do you have any comments on the justification or detail of the proposedremedies on 3 as set out above?

90

What are the options for regulation on Inquam?

• Option AAA

No ex-ante regulation;

• Option BBB

A requirement to secure transparency through publication of prices and priornotification of price changes

• Option CCC

Option BBB plus, a requirement to make transparent the charges, terms andconditions through publication of a reference offer; a requirement to provide mobilevoice call termination services on reasonable request and on fair and reasonableterms; a requirement not to unduly discriminate in the provision of such services;and the imposition of charge controls.

• Option DDD

Option CCC plus, a requirement to maintain cost-accounting systems; arequirement to set prices on the basis of forward looking long-run incrementalcosts; and a requirement for separate accounting systems.

6.162 The Director’s initial view, for the reasons given below, is that:• The appropriate obligations that should apply on Inquam in relation to the

provision of its voice call termination services are those set out in Option BBB

Option AAA

No ex-ante regulation

6.163 Where markets are effectively competitive, ex post competition law issufficient to deal with any competition abuses that may arise.

6.164 However, the Director does not consider that he can rely exclusively on expost competition powers in relation to the provision of mobile voice call terminationby Inquam. As discussed in Chapter 4, the Director’s initial view is that Inquam hasSMP in the market for its mobile voice call termination services. Purchasers ofInquam’s voice call termination services have no choice but to pay the terminationcharge when terminating calls on Inquam’s network. However, given that only 4%of Inquam’s termination relate to off-net calls, as set out below the Directorconsiders that it would be disproportionate to impose non discrimination, costorientation or charge control conditions on Inquam. Nevertheless, the Directortakes note of the fact that, for the small number of offnet calls terminating onInquam’s network, callers have no option but to pay Inquam’s charge. The

91

Director considers that it is therefore appropriate and justified to impose anobligation on Inquam to publish its charges and give prior notification of pricechanges so that customers and interconnecting operators are aware in advance ofhow much they will be charged.

6.165 The Director therefore rejects Option AAA in relation to Inquam’s voice calltermination services.

Option BBB

A requirement to secure transparency through publication of prices andprior notification of price changes

• Obligation to publish charges and proposed changes to those charges inadvance of them taking place

6.166 The requirement to publish charges and changes to those charges inadvance has two main purposes: (a) to assist transparency for the monitoring ofpotential anti-competitive behaviour and (b) to give advanced warning of chargechanges to providers purchasing wholesale termination services. The latter isimportant to ensure that other providers have sufficient time to plan for suchchanges, as they may need to restructure their retail prices in response to chargechanges at the wholesale level.

6.167 The Director considers that, consistent with the current contractualobligation between MNOs and purchasers of voice call termination services, theperiod of prior notification for charge changes should be 28 days. This periodallows sufficient time for purchasers of the services to reflect changes in their retailprices.

6.168 For these reasons, the Director’s initial view is that it is necessary to requireInquam to publish its charges for its voice call termination services and to publishchanges to those charges 28 days prior to those changes taking effect.

Option CCC

Option BBB plus, a requirement to make transparent the charges, terms andconditions through publication of a reference offer; a requirement to providemobile voice call termination on reasonable request and on fair andreasonable terms; a requirement not to unduly discriminate in the provisionof termination services; and the imposition of charge controls.

6.169 This option includes the transparency requirements set out in Option BBB; arequirement to make transparent the charges, terms and conditions throughpublication of a reference offer; a requirement to provide mobile termination onreasonable request and on fair and reasonable terms; a requirement not to undulydiscriminate; and imposition of charge controls.

92

• Obligation to publish a reference offer

6.170 A requirement to publish a Reference Offer (RO) has two main purposes:(a) to assist transparency in monitoring potential anti-competitive behaviour (ifcharges are to be published); and (b) to provide visibility to the terms andconditions on which other providers can purchase services.

6.171 Publication of a RO will ensure stability and certainty to operatorspurchasing mobile voice call termination services. In addition it may enablenegotiations between parties to be completed more quickly, help to avoid disputes,and give confidence to purchasers of such services confidence that those servicesare provided on non-discriminatory terms.

6.172 However, Inquam has a limited number of active subscribers (approximately32,000 according to latest Oftel data) and a very high proportion of call on itsnetwork are on-net calls (96% according to latest Oftel data). This implies thatwere Inquam to set charges above costs, the impact on purchasers andparticularly on end customers would be small. Furthermore the proposedobligation to publish prices and give 28 days prior notification to changes in thoseprices set out above will improve transparency for purchasers of Inquam’s voicecall termination services.

6.173 For the reasons set out above the Director’s initial view is that requiringInquam to publish an RO would be disproportionate.

• Obligation to meet reasonable requests for mobile voice call terminationservices on fair and reasonable terms

6.174 In the absence of a regulatory requirement to provide access to mobilevoice call termination services on reasonable request and on fair and reasonableterms, competition in downstream markets could be distorted by the mobileoperator refusing to supply services or by doing so on terms that were unfair andunreasonable. However, in the case of Inquam the majority of calls terminating onInquam’s network also originate on its network (96% according to latest Oftel data)and its subscriber base is small in comparison to that of the other MNOs (see table6.1 below) so a decision by Inquam to refuse to supply its termination serviceswould have a minimal impact on competition in other markets.

Table 6.1: Mobile Network Subscriber NumbersOperator Number of UK subscribers

O2 11,084,000Vodafone 11,754,000Orange 12,696,000T-Mobile 10,749,000Inquam 42,000

93

Source: Oftel Market Information.

6.175 For the reasons set out above it is the Director’s initial view that it would bedisproportionate to place a regulatory requirement on Inquam to provide access toits voice call termination services on reasonable request.

• Obligation not to unduly discriminate in the provision of network access

6.176 An obligation not to unduly discriminate would be particularly important toprevent mobile operators discriminating in the supply of mobile terminationservices between different purchasers of such services.

6.177 However, the majority of the calls terminating on Inquam’s network are on-net calls, therefore it provides voice call termination to operators other than itselfon very limited occasions. Furthermore, as Oftel is not proposing, for the reasonsoutlined above, to require Inquam to provide access to its voice call terminationservices and it is intended that the prices for its services are published in advance,as set out in Option BB, the application of a no undue discrimination conditionappears disproportionate.

6.178 For these reasons the Director’s initial view is that it unnecessary to placean undue discrimination condition on Inquam in relation to its mobile voice calltermination services.

• Charge controls

6.179 In Chapter 4, the Director sets out his initial view that each MNO has amonopoly of voice call termination on its own network(s) and that the Directorconsiders that MNOs are likely to continue to have SMP for at least the next threeyears.

6.180 As discussed in Chapters 4 and 5, it appears to the Director that there areinsufficient constraints on an MNO's ability to raise termination charges abovecosts and that this may generate detrimental effects on consumers.

6.181 In the absence of regulation, the MNOs have an incentive to set a profit-maximizing price for the provision of wholesale voice call termination. Thedetriment caused is discussed in more detail in Chapter 5. However, as discussedpreviously, Inquam has very few subscribers in comparison with the other MNOs(see table 6.1 above) and the majority of calls terminated on its network alsooriginate on its network. Given this, any detrimental effects that Inquam’s ability toset charges for its voice call termination services above costs may have on theend customer are limited.

6.182 The Director's initial view is that imposing a charge control on Inquam wouldbenefit few consumers and that the burden imposed by the regulation wouldoutweigh any benefits. For these reasons, the Director’s initial view is that it is

94

disproportionate to make Inquam’s mobile termination charges subject to chargecontrols.

Option DDD

A requirement to secure transparency through publication of prices and areference offer; a requirement to provide mobile voice call termination onreasonable request and on fair and reasonable terms; a requirement not tounduly discriminate in the provision of such access; imposition of chargecontrols; a requirement to maintain cost-accounting systems; a requirementto set prices on the basis of forward looking long-run incremental costs; anda requirement for separate accounting systems.

6.183 This option includes the transparency requirements set out in Option BBB,the network access, no undue discrimination obligations and charge controlobligations set out in Option CCC, plus a requirement to maintain cost-accountingsystems; a requirement to set prices on the basis of forward looking long-runincremental costs; and a requirement for separate accounting systems.

• Obligation to set charges on the basis of forward looking long-run incrementalcosts

6.184 For the reasons set out above, Oftel thinks that a constraint on Inquam’sability to set its prices is unwarranted and that the benefit of any such constraintwould be outweighed by the costs of the regulation.

6.185 The Director’s initial view therefore, is that such an obligation on Inquam isdisproportionate.

• Obligation to publish separate accounts for network and retail activities for theprovision of voice call termination services and to maintain appropriate costaccounting systems

6.186 The obligations to maintain appropriate cost accounting systems and toseparately account for termination services are used to monitor compliance withother regulatory obligations. Appropriate cost accounting systems ensure that it ispossible to demonstrate compliance with obligations to set charges on the basis offorward looking long run incremental costs. In the absence of an obligation of thiskind on Inquam the Director considers that it would be unnecessary to requireInquam to maintain such accounting systems.

6.187 An obligation to account separately for voice call termination services maybe required to monitor compliance with an obligation to not unduly discriminate.However, as set out above, the Director considers that in the case of Inquam theeffect of any form of discrimination would be small, and therefore is proposing notto impose a non-discrimination obligation. Without the need to explicitly monitor

95

such discrimination he considers that an obligation to separately account for voicecall termination would be unnecessary and unjustified.

6.188 For the reasons set out above the Director’s initial view is that obligations topublish separate accounts for network and retail activities for the provision of voicecall termination services and to maintain appropriate cost accounting systems areunnecessary.

Comparison of Options

6.189 Oftel has considered four different options for regulating Inquam’s mobilevoice call termination services. Of the options considered, the Director hasdiscounted Option AAA, as this would not deal with the presence of SMP in therelevant market.

6.190 The obligations set out in option BBB would ensure transparency forpurchasers of wholesale termination services, and provides advanced warning ofcharge changes to providers purchasing wholesale termination services. Thisoption appears most appropriate for Inquam, which has a limited number ofsubscribers making mainly on-net calls, meaning that there are few purchasers ofInquam’s voice call termination services.

6.191 Since Inquam has a limited number of subscribers making mainly on-netcalls, the combination of obligations included in Options CCC and DDD, in theDirector’s view, for the reasons set out above would be disproportionate to thefinding that Inquam has SMP in the relevant market.

6.192 The Director's initial view, for the reasons given below, is that:• the appropriate obligations that should apply on Inquam in relation to the

provision of its voice call termination services are those set out in Option BBB

6.193 Clause 44 requires conditions to be justifiable, non-discriminatory,proportionate and transparent. The proposed condition is objectively justifiable, inthat the benefits of publication and notification of charges outweigh any possibledisadvantages. It does not unduly discriminate, in that it is imposed on allproviders of the same services where the competitive and economic conditionsunder which these services are provided are similar. It is proportionate, since itonly applies the minimum regulation necessary to secure the desired outcome andit is transparent as to the effect that it is intended to achieve given the explanationin this document.

Question 14

Do you agree that Option BBB is the most appropriate and proportionateresponse to the provision of mobile voice call termination servicesterminated on Inquam’s network?

96

Detail of the proposed remedies for Inquam

Requirement to notify prices (proposed Condition MB1)

6.194 As explained in the option appraisal, the Director’s initial view is that Inquamshould be required to publish charges and give notice before it is permitted tochange any charge for its mobile termination services. This is reflected in theproposed Condition MB1.

6.195 The Director has considered how much notice Inquam should be required togive before changing charges for mobile termination services and, as a result,believes that a notification period of 28 days is appropriate, in that it is broadlyconsistent with current contractual terms and allows sufficient time for changes inthe wholesale charge to be reflected in other providers' retail charges while notunduly impacting on the ability for Inquam to change its wholesale charges.

6.196 The Director therefore proposes that Inquam should be required to give 28days' notice of charge changes in this market.

6.197 The notice should include the following information:• a description of the network access in question; and• the date on which or the period for which any amendments to charges will take

effect.

6.198 The Director considers that the proposed conditions meet the tests set out inthe Communications Bill. The Director has considered all the Communityrequirements set out in Clause 4 and in particular the requirement to secureefficient and sustainable competition and maximise the benefit for customers.

6.199 The Director considers that, in accordance with the tests set out in Clause44(2) of the Communications Bill, the proposed condition is objectively justifiable,non-discriminatory, proportionate, and transparent. It is objectively justifiable, inthat the benefits of publication and notification of charges outweigh any possibledisadvantages. It is proportionate, in that only information that is necessary toother network providers needs to be notified. It does not discriminate undulyagainst Inquam, in that it is imposed on all providers of the same services wherethe competitive and economic conditions under which these services are providedare similar and it is transparent given the explanation in this consultationdocument.

Question 15

Do you have any comments on the justification or detail of the proposedremedy on Inquam as it is set out above?

97

Chapter 7

Proposed Charge Controls

The proposed charge controls for 2G mobile voice termination

7.1 In Chapter 6, the Director has reached the provisional conclusion that, giventhe initial finding of SMP for each operator in the relevant market, directcontrols (through a cap) should be imposed on the charges for terminatingmobile voice calls on the 2G mobile networks of O2, Orange, T-Mobile andVodafone. This Chapter sets out in more detail the proposed structure andlevel for these controls.

7.2 In order to impose a cap on a set of charges, it is necessary to identify:

(a) the appropriate final target level (the ‘fair target charge’) these chargesshould be brought down to by the end of the control; and then

(b) how these charges should be brought down to this fair target charge .

The fair target charge

7.3 In the case of mobile termination, it has been argued1 that the appropriate fairtarget charge should be based on a Ramsey approach. This is on theargument that setting mark-ups for the recovery of fixed and common costs onthe basis of the relative demand own-price and cross-price elasticities and ofany externality is the most efficient pricing methodology when marginal costpricing cannot be used.

7.4 In paragraph 2.519 of its report, the CC considered the use of Ramsey pricingand rejected it for the following reasons:

(a) the absence of consensus regarding absolute and relative price-elasticities,determination of which is necessary for the computation of Ramsey prices;

(b) the extreme disagreement among the parties as to the nature and extent offixed and common costs, determination of which is also necessary for thecomputation of Ramsey prices;

(c) the implausible outputs of some of the models claiming to compute Ramseyprices;

(d) the inconsistency between actual retail price patterns observed and theretail price patterns claimed as naturally resulting, in accordance withRamsey principles, from competition in the absence of regulatoryconstraint;

(e) the uncertainty as to the actual level of prices, having regard to the difficultyof allocating subscription charges between handsets and calls;

1 See paragraphs 2.516 to 2.520 of the CC report.

98

(f) the CC’s doubts as to the complete effectiveness of retail competitionnecessary for the emergence of Ramsey pricing patterns even in theory;and

(g) the CC’s inability, in any event, to recommend Ramsey prices in the retailmarket, if our proposals for regulation relate to one element only, namelytermination charges.

7.5 The Director has taken into account the CC’s findings and in particularconsidered points a), c), e), f) and g). These are discussed in more detailbelow. The Director’s conclusion is that it would not be appropriate to attemptto set Ramsey prices for three reasons:

1. the difficulties inherent in calculating reliable Ramsey prices;2. the fact that it is unlikely that all the other prices for mobile services would be

set at Ramsey level; and3. the distributional inequities generated by a Ramsey pricing structure for mobile

services.

1) Calculating Ramsey prices

The Director believes that it is impractical to set Ramsey prices because thisexercise requires an accurate assessment of, if not the absolute, at least therelative, elasticities and cross-elasticities of demand for each mobile service.Estimating demand elasticities is a complex exercise, and in this case is furthercomplicated by the fact that the range of estimates necessary to set Ramseyprices in the mobile sector is quite large (because the number of servicesinvolved is high2). This consideration is further strengthened by thedisagreement concerning the absolute and relative values of these elasticitiesin the evidence submitted by each MNO and Oftel during the CC inquiry lastyear.

A number of models to calculate Ramsey prices were submitted to the CC.

• The particular difficulties which were identified with the models whichwere provided to the Competition Commission were: they modelled onlya sub-set of the mobile services (excluding e.g. text messaging,roaming, international calls, mobile internet access etc).3

• The services that are modelled are captured in a highly simplifiedmanner. For example, calls from mobiles are modelled as a single“composite” service, when in reality there are different services formobile-to-fixed, off-net mobile-to-mobile and on-net calls (only Dr Rohlfs

2 In addition to all voice services, also data services and text messaging services should be included.3 See Ramsey Prices and Network Externalities: Dr Rohlfs’ Analysis, Oftel: 23 May 2002

99

on behalf of Oftel attempted to split out off-net from other calls frommobile, but this still leaves the latter as a composite service).4

• To capture the full set of services a matrix of elasticities of at least 9 x 9needs to be populated with robust values for the elasticities.5 However,the models put forward used a matrix of only 3 x 36 (or in the case of DrRohlfs’ model, 4 x 4, since he split out off-net calls from other mobile-originated calls).

• The models do not allow for different consumers paying different pricesfor the same service, e.g. because they are on different tariff packages.Dr Rohlfs modelled this to some extent in his estimate of targetedsubsidies,7 but this is very simplified compared to the richness andvariety of the price discrimination practised by the MNOs.

Even despite these omissions, the range of Ramsey termination chargesestimated in the four pricing models was extremely large, as shown in thefollowing table.

Model Range of mark-up oftermination charge over LRICin fixed-to-mobile price

Rohlfs (Table 9.1 at § 9.6) 0.8 - 1.6Frontier Economics (Table 9 at §42 of Appendix 9.1)

7.2 - 13.4 (incl. Common costs)2.2 - 8.3 (externality only)

DotEcon (Table 14 at § 60 ofAppendix 9.1)

3.2 – 7.7

CRA (§ 73 of Appendix 9.1) 5.2 – 8.5

The Director considers that the Frontier Economics, DotEcon and CRA models areflawed because of implausible elasticity assumptions and/or methodologicaldeficiencies.8 In particular, the claimed values of the cross elasticities in thesemodels imply wholly implausible values for the externality, suggesting that calledparties benefit from a call being made to a much greater degree than callers. The

4 This is reflected in Dr Rohlfs’ analysis commissioned by Oftel. See A Model of prices andcosts of mobile network operators, Dr Rohlfs, 22 May 20025 The 9 services referred to are: fixed-to-mobile calls, mobile subscription, mobile-to-fixed,off-net mobile-to-mobile, on-net, text messaging, roaming, international, and mobile internet accesscalls. Each service has one “own-price” elasticity of demand, and a “cross-price” elasticity relatedto each other service (although some of the cross-price elasticities may be zero). The own-priceelasticity is the percentage change in demand caused by a small change in the price of the service.The cross-price elasticity is the percentage change in the demand for other products caused asmall change in the price of the service.6 Fixed-to-Mobile calls, mobile subscription and composite mobile-originated calls.7 See Estimates of Targeted Subsidies, Dr Rohlfs, 19 June 20028 Report §§ 2.346, 2.519(c), and Appendix 9.1.

100

wide range of estimates itself serves to illustrate the impracticality and lack ofwisdom in relying on a particular set of estimated Ramsey prices.

2) The prices for retail mobile services

7.6 The Ramsey approach to pricing is concerned with the structure of a set ofprices. This methodology specifies the relativities between the prices for agroup of services that share common costs. Hence, if only some of theprices are on the basis of this methodology, the efficient outcome is notobtained. Hence it would be efficient to set termination charges at Ramseylevel only if all the other mobile services were also priced at Ramsey level.

7.7 The Director believes that even if he determined the termination charge atRamsey level, it would not necessarily follow that the MNOs would setRamsey-based prices for the other mobile services. A regulated Ramsey-based termination charge would be likely to ensure a set of Ramsey pricesfor all mobile services only if the retail mobile market was perfectlycompetitive. Only if this condition was satisfied would the MNOs beconstrained to select an overall Ramsey-based pricing structure. However,the Director believes that the retail mobile market does not fulfil this condition(See the Review of competition: mobile access and call origination -published in April 2003).

7.8 In addition the Director is of the view that there is no evidence that the MNOsare currently setting their retail prices anywhere near Ramsey level. This isespecially evident in the case of the prices for outgoing on-net and off-netmobile-to-mobile calls, where the differential is so large that it cannot bejustified in terms of different elasticities of demand.

7.9 A solution could be to regulate all mobile prices so as to achieve a completeset of Ramsey prices. However, the Director considers that, apart from theaforementioned difficulties in setting such prices, this would be too intrusiveand disproportionate a form of regulation given the level of competitionpresent in the mobile retail market.

The issue to be addressed, therefore, is to set one of the MNOs’ prices, i.e.the termination charge, given that the other, retail prices are set by the MNOsthemselves in the retail market. In such circumstances, the economicallyefficient outcome In theory would depend upon a ‘principal-agent’ model,which considers the optimal termination charge to be set by the regulator (or‘principal’), taking account of the retail prices that would be set by the MNOs(or ‘agents’), which themselves depend on that termination charge. Thiscompounds the practical difficulties set out above, by adding the substantialfurther complication of needing to model accurately the way in which retailprices are set. Given the complexity of retail prices, this is an extremelydifficult task.

3) Distributional effects

101

7.10 As discussed at length in Chapter 5, the Director considers that a set ofprices characterised by high termination charges and low subscription andoutgoing call prices would lead to some consumers unfairly subsidisingothers.

Conclusions on setting the fair target charge at Ramsey level

7.11 In conclusion, though the Director accepts that in theory a set of Ramseyprices for mobile services would generate the most efficient outcome, heconsiders that there are sufficient reasons to believe that such a pricingstructure cannot be achieved.

The proposed approach for setting the fair target charge

7.12 The Director is proposing to set the target charges on the basis of LRIC plusa mark-up for common costs based on the equal proportionate mark-up(‘EPMU’) approach and an externality surcharge.

LRIC

7.13 The Director is of the view that the most appropriate and economicallyefficient basis for regulatory charge controls is forward-looking LRIC. TheLRIC of voice termination is the additional cost an MNO incurs to providetermination. This can also be seen as the cost that the firm would avoid if itdecided not to provide voice termination, taking a long run perspective. LRICbased charges correspond more closely to the charges that would prevail inan effectively competitive market than accounting-based measures of cost. Itis a fundamental goal of price regulation to mimic the effects of a competitivemarket and this consideration underpins the use of LRIC.

7.14 The adoption of LRIC as a regulatory costing technique is used widely forexample by other NRAs in Europe, and by the FCC in the US. It has alsobeen identified as the most appropriate methodology to use for settinginterconnection charges by the European Commission in its 1998Recommendation on Interconnection Recommendation 98/195/EC 8 January1998) Furthermore, the CC has agreed, with the use of LRIC as theappropriate costing methodology for setting termination charges (SeeParagraph 2.251 of the CC report ).

Economic depreciation

7.15 The depreciation approach selected by the Director for the LRIC model iseconomic depreciation (For further details of the conceptual underpinnings,see Calls to mobile: economic depreciation, September 2001. For adiscussion of the cost path over time using economic depreciation in theLRIC model for key assets, see Additional Information Concerning Oftel’sLRIC Model, 12 February 2002). This matches the cost of equipment to its

102

actual and forecast usage over the long term. As a consequence, there isrelatively little depreciation in years where utilisation is low and relatively highdepreciation in years of full equipment utilisation. By contrast, most forms ofaccounting depreciation are relatively simple. The usual method is to take theactual price paid for equipment (or its replacement cost) and divide by theexpected equipment life to reach a depreciation charge for the year. Thetiming of cost recovery under economic depreciation varies from that underaccounting depreciation. Between 2001 and 2006 the use of economicdepreciation results in a higher pence per minute cost of terminating callswhilst in years prior to 2001, economic depreciation would have resulted inlower costs compared to an equivalent calculation based on accountingstraight-line depreciation

EPMUs for recovery of common costs

7.16 The Director regards it as appropriate that regulated services contribute tothe recovery of relevant common costs and it is Oftel’s usual practice to add amark up on the top of LRIC to allow for full cost recovery. The Directorbelieves that it is appropriate that these costs should be recovered by anEPMU. The use of EPMU to allow for common cost recovery is reasonable,given the undesirability or impracticality of other approaches.

7.17 The Director also considers that in this case EPMU is the most appropriateterminology for the recovery of common costs. EPMU is equivalent to thespecial case of Ramsey pricing when the super-elasticities (The super-elasticities include the own price and cross-price elasticities for each serviceweighted by the relative revenues) of the services are equal. The Directorhas already concluded that the value of the elasticities cannot be reliablyestimated and therefore that it cannot be established whether the super-elasticities are different from each other. Hence, using EPMU is no morearbitrary than using unreliable empirical estimates of Ramsey mark-ups but itis less burdensome to implement. In addition the Director believes thatcommon costs in the provision of mobile services are limited and, thus, hebelieves that the departure from the efficient pricing structure (if this could beestablished, i.e. if full information were available) caused by the use of EPMUis relatively small.

7.18 Further details concerning the implementation of LRIC and the mark-up forcommon costs can be found in Annex E.

Externality surcharge

7.19 The decision by a person to be part of a mobile network generates a benefitfor fixed and mobile subscribers (the ‘network externality’), because they willbe able to call or be called by him or, at least have the option to do so. Sincecallers to mobiles derive a benefit from the called party’s decision tosubscribe to a mobile network, it seems appropriate to the Director that thecharge they pay for the call should reflect this benefit. This is done by adding

103

a further mark up (an ‘externality surcharge’) to the termination charge whichreflects the value of this externality. More details on the network externalityand on the relevant surcharge can be found below and in Annex F.

The structure of the proposed charge controls

7.20 The previous section dealt with how the fair target charge should be set, thissection now describes how the Director proposes that current terminationcharges should be brought down to this target.

7.21 For this market, the Director proposes to place a cap on the weightedaverage of the charges levied by each the four MNOs (i.e. daytime, eveningand weekend charges) for terminating voice calls on their 2G networksweighted by the relative call volumes. The cap shall bring the weightedaverage charge down to the fair target charge level by 2006.

7.22 In particular, the Director proposes that:

i. the controls should last until 31st March 2006;ii. in the first period (running up to 31st March 2004) the Target Average

Charge (TAC) should be a specified figure, whereas in the following twoperiods the TAC should be set on the basis of an RPI-X formula (The Xis the percentage by which the weighted average charge has to fall ineach period to reach the ‘target’ charge in the final year of the proposedcontrol);

iii. there should be two separate sets of caps for termination of fixed-to-mobile and off-net calls;

iv. the weights in each cap should be based on the volumes of minutes ofthe relevant traffic experienced by each MNO during the previousfinancial year;

v. calls to ported-in mobile numbers should be included in the controls; andvi. since the fair target charges for the combined 900/1800MHz and the

1800MHz MNOs are different, the controls on these two types ofoperators should be set at different levels.

The control periods

7.23 The Director considers that the charge control regime should last until March2006. He is of the view that this is the appropriate length of time for a controlas it covers the period over which he considers it unlikely that competitiveconditions will change materially. He also considers that this is theappropriate period over which to bring termination charges down to the fairtarget charge level. Reductions over such a period will allow this fair targetcharge to be achieved sufficiently quickly to deliver substantial benefits toconsumers but also sufficiently gradually to allow operators and customers toadjust to new levels and structures of mobile charges.

104

7.24 The controls require that, during each period of the control, the averagecharge set by the regulated MNO (the Average Interconnection Charge or‘AIC’), does not exceed the charge with which the operator is required tocomply (the Target Average Charge or ‘TAC’). (More details on how Oftelproposes to calculate the AIC and the TAC for this control are contained inAnnex H on the Charge cap mechanism.)

7.25 For the first period of the control, the Director is proposing that the relevantTAC should be an absolute target in pence per minute. The TAC shall bedifferent for the combined 900/1800MHz MNOs (i.e. Vodafone and O2) andfor the 1800MHz MNOs (i.e. Orange and T-Mobile). The reason for thisdifference is explained below.

7.26 In each of the following two periods, the Director is proposing that therelevant TAC should be specified in terms of the RPI-X formula. There shallbe a different X for the combined 900/1800MHz MNOs and for the 1800MHzMNOs.

7.27 It is desirable that the TAC in the first period should be set as an absolutetarget in pence per minute rather than specified in terms of the RPI-X formulato allow the charges of the four operators to be aligned in two ways:

1) the same TAC could then be set for Orange and T-Mobile, reflecting theidentical efficient costs that they incur as 1800MHz operators (CurrentlyOrange and T-Mobile have different average charges); and

2) although the fair charge differs in each year for combined 900/1800MHzand 1800MHz operators, the target charges of the two types of operatorsin this first period could be set so that the TAC for each is the sameamount above the fair charge. Hence the gap between the TAC for thetwo types of operators would equal the gap between the fair charge foreach type of operator on average during the first period. One type ofoperator would not have an advantage over the other, which ensures thata distortion in competition in the retail market is avoided.

Calls from fixed network and off-net calls

7.28 The Director is proposing to have two separate sets of controls:

a) one on the charges for terminating voice calls from fixed phones on 2Gnetworks, and

b) one on the charges for terminating off-net (mobile-to-mobile) voice callson 2G networks.

7.29 He also proposes to set these two sets of controls at the same level, sincethe LRIC of termination does not differ depending on where the calloriginates.

105

7.30 MNOs currently levy three different charges depending on the time of dayand day of week when the call is terminated: day, evening and weekend.The Director believes that the control on each MNO should be placed only onthe weighted average of these three charges. He considers that it would beunduly intrusive to regulate the level at which each of these charges shouldbe set and intends to leave the MNOs free to vary these charges provided theoverall cap is met.

7.31 However, the Director is of the view that regulation should not necessarilyrequire that charges for terminating fixed-to-mobile and charges forterminating off-net calls should be identical. Hence, he considers that havingtwo sets of caps avoids the potential problem of an MNO loading the majorityof charges onto one type of call whilst keeping the average charge acrossboth types of calls below the level required by a single cap. Since theproposed controls are set at the same level, the pressure posed on each setof charges would not be different. This arrangement would not prevent theMNOs from entering into bilateral agreements or setting lower off-nettermination charges, if they so wish (any such arrangements would, ofcourse, still need to be compliant with the Competition Act).

7.32 Since each set of caps refers to a specific type of traffic, Oftel also considersthat the weights should be traffic specific. This mean that the weightsemployed to set the TAC and the AIC in each period of the control shouldreflect the volume of minutes of the relevant type of traffic terminated by eachoperator during the previous financial year. It is possible that an MNO is notable to identify the origin of the calls that terminate on its network. In thatcase, the MNO can ask for the Director’s consent to change the weights toreflect total traffic volumes (i.e. the sum of the volumes of minutes of mobile-to-mobile and fixed-to-mobile calls). This consent shall only last for the periodof the control during which it is requested.

Treatment of ported numbers

7.33 Since January 1999 portability of mobile numbers has been available in theUK and until September 2002 1.7 million numbers (out of 34 million – Ofteldata) have been ported. Furthermore, the volume of ported numberscontinues to rise. Under the current technical and commercial arrangements,calls to a ported number are routed via the donor MNO to the recipient MNO,which then receives the termination charge of the donor MNO (less a transitfee). Hence, calls to ported numbers are ‘ported in’ from the perspective ofthe recipient network and ‘ported out’ from the perspective of the donornetwork.

7.34 The Director considers that the level of porting of mobile numbers isbecoming significant enough to warrant proper consideration of how theyshould be treated in the charge control. He, therefore, proposes that thecontrol on termination charges for calls of each MNO covers all calls tohandsets connected to its networks, including calls to ported-in numbers.

106

7.35 However, he accepts that for calls to ported-in numbers the recipient (orterminating) MNO receives an amount which is different from its owntermination charge, and over whose level it has no control, and that this mayrender it difficult for the MNOs to comply exactly with their control for tworeasons:• it would need to forecast accurately the charges that all other MNOs

would set for termination, since these would form part of its own AIC viaported-in minutes; and

• it would have to set its termination charges taking account of the weightsin other MNOs’ caps.

A more detailed explanation of this is included in Annex H. Hence, the Directorproposes to include in the control on termination charges the proviso that he mayconsent to a variation in the volumes used in the calculation of the AIC so as toexclude call minutes to ported-in numbers. This consent shall be requestedseparately for each period of the control. The Director can grant it in one periodand then withhold it in the following period, if he is concerned that there might bemanipulation of the charge control arrangements to set excessive terminationcharges for calls to ported-in numbers.

The controls for the combined 900/1800MHz and the1800MHz operators

7.36 The Director has considered whether there should be different controls ie,one for each operator or whether they should all be subject to the same cap.He has reached the initial conclusion that the fair target charges for combined900/1800MHz and 1800MHz MNOs should be different and, thus, that thecontrols on these two types of operators should be set at different levels.

7.37 The Director has noted and accepts the CC’s conclusion that, at currenttraffic levels of the MNOs, combined 900/1800MHz and 1800Mhz operatorsemploy a similar amount of network equipment and so have similar costswhen considered on an accounting depreciation basis (Paragraphs 2.301-2.307 of the CC’s report). However, in calculating the fair target charge basedon LRIC, the Director has used economic depreciation to obtain the path ofcosts over time. As described above, under economic depreciation, costrecovery is deferred from earlier years, in which utilisation was lower, to lateryears, in which higher levels of utilisation are experienced. Due solely to thecharacteristics of the 1800MHz spectrum, rather than operator-specificconsiderations, this effect is more pronounced for 1800MHz networks whichhave a smaller maximum cell radius and so 1800MHz operators experiencelower site utilisation in the earlier years of their operation than the combined900/1800MHz operators. Therefore, the average 1800MHz operator’s costscan be expected to be higher than a combined 900/1800MHz operator’scosts in 2005/06.

The specific proposals

107

7.38 This last section describes the specific proposals and how these have beenderived.

7.39 Table 7.1 and 7.2 below summarise the proposed target average chargesand charge caps for mobile-to-mobile 2G voice termination and fixed-to-mobile 2G voice termination respectively. The proposed charge controls forfixed-to-mobile are identical to those for mobile-to-mobile voice termination.

Table 7.1: Proposed charge caps and TAC for 2G mobile-to-mobile voicetermination charges

Combined 900/1800MHz MNOs 1800MHz MNOsYear O2 Vodafone T-Mobile Orange

Target average charge Target average charge2003/04 6.95 ppm 6.95ppm 7.70 ppm 7.70ppm

Cap Cap2004/05 RPI-15 RPI-15 RPI-14 RPI-14

Cap Cap2005/2006 RPI-15 RPI-15 RPI-14 RPI-14

Table 7.2: Proposed charge caps and TAC for 2G fixed-to-mobile voice termination charges

Combined 900/1800 Hz MNOs 1800 Hz MNOsYear O2 Vodafone T-Mobile Orange

Target average charge Target average charge2003/04 6.95 ppm 6.95 ppm 7.70 ppm 7.70 ppm

Cap Cap2004/05 RPI-15 RPI-15 RPI-14 RPI-14

Cap Cap2005/2006 RPI-15 RPI-15 RPI-14 RPI-14

Derivation of the figures

7.40 As described above, the charge controls are derived by first determining thefair target charge in 2005/06 and then calculating the appropriate percentageby which the current charge should fall each year from the average existingcharges in 2002/03, allowing for inflation. A further adjustment is made toalign the target charges for the two types of operators after the first period sothat the TAC for the two types of operators is the same amount above the faircharge. The approach is illustrated in the figure below.

108

Figure 7.1: Starting charges and values of X

2005/06

11.00

LRIC+EPMU+Externality

2003/042002/03

Price,Cost(ppm)

900/1800MHz

2004/05

9.35

900/1800MHz

1800MHz

1800MHz

7.41 The derivation of the fair target charge is described in the section belowfollowed by the calculation of the appropriate percentage reductions and firstperiod target charges.

Derivation of the fair target charge

7.42 The fair target charge in 2005/06 is composed of the LRIC for voicetermination with a two-stage mark-up: firstly an EPMU for common costs andsecondly a mark-up for the network externality.

LRIC+EPMU for common cost recovery

7.43 The LRIC for voice termination is calculated from a LRIC model developed byOftel and published in April 2002 (see http://www.analysys.com/ukmobilelric),based on the costs of a reasonably efficient 2G mobile operator in the UK. Inits review of the charges for calls to mobiles, the CC agreed with the generalprinciples underlying the model methodology (see Annex E for details) andagreed that the April 2002 LRIC model was a suitable starting point for theassessment of the costs of terminating calls on mobile networks (paragraph2.287 in the CC report ).

7.44 However, in the light of the CC’s investigation last year and its December2002 report, the Director believes it appropriate to consider a number ofissues and potential adjustments concerning the output of the April 2002model. These issues are as follows and are discussed briefly in turn belowwith further details provided in Annex E:

109

Cost of capital;

Comparison with MNO data;

Market share; and

Non-network common costs.

a) Cost of capital

7.45 The appropriate cost of capital in the context of this market review is the costof capital for a reasonably efficient 2G mobile operator in the UK. In formingits views about the cost of capital, the Director uses a number ofmethodologies, but the main emphasis is on the use of the Capital AssetPricing Model (CAPM). The Director has undertaken a fresh analysis of eachof the components of the CAPM used to derive an estimate for the cost ofcapital in the light of more recent information. On this basis, he estimates thepre-tax real cost of capital to be in the range of 9.2% to 14.7% with a mid-point of 12%. Further details of the derivation of this range are provided inAnnex D.

b) Comparison with MNO data

7.46 In order to address concerns over the accuracy of the LRIC model, theDirector has undertaken a comparison between the outputs of the model andactual cost accounting data from the mobile operators. The Director hasderived adjustments, as detailed in Annex E, to be applied to the output of theLRIC model following the methodology proposed by the CC in its recentinquiry. In summary, the Director finds that an upward adjustment of 29.8%should be applied to the capital costs and a downwards adjustment of 12.5%should be applied to the operating costs in the LRIC model, to reconcile themodel’s output with the actual costs incurred as reported by the MNOs.

c) Market share

7.47 The output of the LRIC model is based on the costs of an average operatorwith a 25% market share in 2001, declining to 20% by 2010 following theentrance of the fifth operator, “3”. There is a question whether the fair chargeshould be adjusted to take into account the position of an MNO with a marketshare of total traffic lower than the average. This question is open to morethan one reasonable answer. However, in the current context the key costfigure is the fair charge in 2005/06 and the Director concludes that the point isacademic as even if a market share adjustment was implemented asproposed by the CC (paragraph 2.280 of the CC report ), this adjustmentwould diminish to zero by 2005/06 by design. Thus whether this adjustment ismade or not, there would be no adjustment to the fair charge in 2005/06.

110

7.48 The April 2002 LRIC model incorporates an EPMU for network commoncosts. Following the CC’s investigation, the Director is mindful to include anincrease in this common cost mark-up for the recovery of non-networkadministrative costs that should be recovered across all areas of thebusiness, including both network and retail services. The Director hasdetermined this mark-up to be 0.30ppm (in real 2000/01 terms) based oninformation submitted by the MNOs to the CC regarding administrativeoverheads. Details of this calculation are provided in Annex E.

The economically efficient externality surcharge

7.49 As mentioned above, the Director considers it appropriate to add a furthermark up (an ‘externality surcharge’) to the termination charge which reflectsthe value of the network externality.

7.50 Deriving a precise value for the economically efficient externality surcharge intheory is a complex and multi-faceted issue. A wide variety of factors shouldbe considered when setting this surcharge. Specifying a model that capturesall of the relevant conceptual points would thus be a very difficult task. Inaddition, such a model would rely on many parameters whose variablescannot reliably be estimated in practice (e.g. the elasticities of demand).

7.51 Several different quantification exercises have been carried out (which aredescribed in Annex F). Given the difficulties discussed above, the Directorconsiders that he should not rely on any single estimate of the externalitysurcharge. Each of them is incomplete because it only captures a sub-set ofthe factors and/or because some of the assumptions used (e.g. elasticities)cannot in practice be robustly derived from empirical data. However, theDirector considers that each of them, in conjunction with the others, providesa useful insight into the reasonable value for the externality surcharge.Based on the available evidence, the Director considers that a reasonableexternality surcharge is 0.4 ppm (in real 2000/01 terms). More details on thiscan be found in Annex F.

The fair target charge

7.52 Taking account of the factors raised above, the Director has determined thefair target charges in 2005/06 to be 4.73ppm and 5.29ppm (in real 2000/01terms) for combined 900/1800MHz and 1800MHz operators respectively, asshown in the table below. These charges are very similar to those determinedby the CC and presented in Table 2.11 of their report.

Table 7.3: Derivation of the fair target charge in 2005/06

Pence per minute (real 2000/01) Combined900/1800M

Hz

1800MHz

111

Original April 02 model (12.5% WACC) 3.58 4.35Cost of capital adjust (12.0% WACC) -0.06 -0.08Capital / Operating cost adjustment 0.51 0.32Non-network common cost mark-up 0.30 0.30Network externality surcharge 0.40 0.40Fair target charge 4.73 5.29

Calculation of X for the charge control

7.53 The value of X for the charge control can be calculated from the startingcharge (at the beginning of the first period) and the fair target charge (in2005/06). The starting charge is derived from the nominal prices charged in2002/03 after implementation of the 15% reduction in real terms asrecommended by the CC, to take effect within the period from 1 April to 25July 2003. Information received from the MNOs in March 2003 confirmed theaverage charge for combined 900/1800MHz operators to be 9.35ppm (asrequired by the existing price controls on Vodafone and O2) and showed theaverage charge to be 11.0ppm for 1800MHz operators. This leads to astarting charge at the beginning of the first period of 7.67ppm and 9.02ppm(in real 2000/01 prices for direct comparison with the fair target charges) forcombined 900/1800MHz and 1800MHz operators respectively.

7.54 By considering the starting charges and target charges in real terms (2000/01prices), the real percentage reduction can be calculated which is required toreach the fair target charge in 2005/06 after three successive applications ofthat reduction. This results in a reduction of 15% for combined 900/1800MHzand 16% for 1800MHz operators in each of the three periods (see Annex Gfor details).

7.55 To determine the TAC at the end of the first period as an absolute targetcharge in pence per minute, this 15% real reduction is applied to thecombined 900/1800MHz charge at the beginning of the first period of7.67ppm. This leads to a combined 900/1800MHz target charge of 6.52ppm(in real terms 2000/01 prices). As shown in Table 1 of Annex G, the gap inthe fair charges for the two types of operators at the end of the first period is0.69ppm. When added to 6.52ppm, this gives a target charge of 7.21ppm (inreal terms 2000/01 prices) for the 1800MHz operators. Expressed in nominalterms, the target average charges at the end of the first period are 6.95ppmfor combined 900/1800MHz operators and 7.70ppm for 1800MHz operators.

7.56 To complete the calculation, the value of X for the RPI-X control in theremaining two periods can be derived from the real percentage reductionnecessary to reduce the TAC at the end of the first period to the fair targetcharge in 2005/06. The following table shows the results of this calculation forcombined 900/1800MHz and 1800MHz operators separately.

112

Table 7.4: First period TAC and derivation of value of X for 2nd and 3rdperiods

combined900/1800MHz operators

1800MHz operators

1st period target charge (real 2000/01 ppm) 6.52 7.211st period target charge (nominal ppm) 6.95 7.70

Final target fair charge (real 2000/01 ppm) 4.73 5.29Value of X for 2nd & 3rd periods (rounded) 15% 14%

7.57 In conclusion, the Director proposes that the target average charge for thefirst period should be set at 6.95ppm for combined 900/1800MHz operatorsand 7.70ppm for 1800MHz operators, followed by a charge control of RPI-15for combined 900/1800MHz operators and RPI-14 for 1800MHz operators inthe remaining two periods.

Cost benefit analysis

7.58 The Director recognises that regulatory intervention is appropriate only whenthere is a reasonable expectation that its benefits will exceed its costs. TheDirector’s approach to the assessment of the appropriate level of voicetermination charges involves an estimation of the set of charges thatmaximises the welfare of consumers, subject to ensuring that MNOs are ableto earn a reasonable return on their investment. This assessment takesaccount of the benefits to all customers, including those calling mobiles aswell as the mobile customers themselves.

7.59 Models of economically efficient pricing are in principle well-suited to derivingestimates of the welfare gains from regulation. The Director’s reservationsabout the relevance and practicality of deriving Ramsey prices means thatsuch estimates should not be regarded as precise. But they provide anindication of the direction and broad magnitude of the effect. In order toestimate the welfare gains from regulation of mobile termination charges, theDirector compared two scenarios:

i) a scenario in which termination charges are brought down via thecharge control to the Director’s fair target charge and where otherprices are assumed to be set on a Ramsey basis; and

ii) an unregulated scenario in which MNOs set high termination charges,but are assumed to make no supernormal profits (ie make sufficientrevenues to cover costs, including the cost of capital).

7.60 This comparison represents a conservative estimate of the likely gains fromregulation. Importantly, this is because the welfare comparison aboveassumes no change in profits between the two scenarios (in each scenarioMNOs just cover costs including the cost of capital), ie it assumes a complete‘swings and roundabouts’ (or ‘waterbed’) effect, which is conservative

113

because the retail market is less than perfectly competitive. The resultingtotal welfare gain will therefore reflect a pure change in consumer surplus andthe more complex consideration of transfers of surplus between MNOs andconsumers is avoided.

7.61 The Director performed the above comparison using the relevant versions ofthe model produced by Dr Rohlfs and described in his paper A model ofprices and costs of mobile network operators, 22 May 2002. Dr Rohlfs’ modelis based on four services: mobile subscription, mobile-originated usage otherthan off-net, fixed-to-mobile usage and off-net usage. Both the constrainedRamsey run and the zero-profit unregulated run used in the CBA describedherein were based on linear demand functions. These versions were updatedto reflect adjusted cost inputs consistent with those described in Annex E andthe Ramsey model was additionally constrained to yield a fixed-to-mobileprice consistent with the fair charge for termination (the fixed-to-mobile pricebeing the sum of mobile termination and the fixed network operatorretention).

7.62 On the basis of the above analysis, the estimated gain in total welfare fromregulation of mobile termination at the fair charge is approximately £240M perquarter in 2005/06 (2003/04 prices). Oftel estimates that over the period ofthe charge controls, this would translate to a present value stream of benefitsof approximately £2,000M. This present value stream was obtained byconverting the per quarter gain in welfare to an annual figure, applying simplelinear interpolation between the base year (2000/01 in Dr Rohlfs’ model) andthe end of the control period (2005/06), applying a volume growth adjustmentconsistent with the LRIC model, discounting all gains over the three periods:August 2003 to end March 2004 (with the annual gain scaled down to reflectthe 8-month period), April 2004 to end March 2005 and April 2005 to March2006.

7.63 The Director therefore considers that there are substantial welfare gainsassociated with regulation of termination charges at the proposed fair charge.

114

Chapter 8

Consultation details

8.1 The Director seeks the views of interested parties on the analysis andproposals contained in this consultation document. The Director will then publish astatement setting out his conclusions.

8.2 With reference to regulation 6(6) of the Regulations, a consultation period of10 weeks has been permitted for representations. Representations must arrive atOftel no later than close of business on 24 July 2003. Representations receivedafter this time will not be taken into account, and no extensions of the deadline willbe permitted.

8.3 Where possible, comments should be made in writing and sent by e-mail tolawrence.knight@oftel.gov.uk. However, copies may also be posted or faxed tothe address below. If any stakeholders are unable to supply their comments in oneof these ways, please use the contact details below to discuss alternatives.

Lawrence KnightOftel50 Ludgate HillLondon EC4M 7JJ

Tel: 020 7634 8747Fax: 020 7634 8738

8.4 Confidential responses should not be sent via the Internet. Responses will bepublished on Oftel’s website in the Publications section under Responses to Oftelconsultations except where respondents indicate that the response, or part of it, isconfidential. Appointments to view written comments in Oftel’s Research andInformation Unit must be made in advance (see contact details below).Respondents are therefore asked to separate out any confidential material into aclearly marked annex. In the interests of transparency, respondents are requestedto avoid confidential markings wherever possible.

8.5 On this occasion, the Director is not inviting stakeholders to comment on therepresentations made by others.

Further copies of this document

8.6 This document can be viewed on Oftel’s website, www.oftel.gov.uk. Papercopies and more accessible formats such as large print, Braille, disc and audiocassette can be made available on request. Please contact Oftel’s Research andInformation Unit by telephoning 020 7634 8761 or by sending an e-mail toinfocent@oftel.gov.uk.

115

E-mail notifications

8.7 Oftel has a free e-mail based mailing list to help people stay informed aboutthe work that Oftel is doing. Each time an Oftel document is published and placedon Oftel’s website at www.oftel.gov.uk, subscribers to the list receive an e-mailalert. To register, please go to the What’s New section of the website and accessthe electronic form.

Publication of representations made by stakeholders8.8 On this occasion, the Director is not inviting stakeholders to comment on therepresentations made by others. However, in the interests of transparency, allrepresentations will be published, except where respondents indicate that aresponse, or part of it, is confidential. Respondents are therefore asked toseparate out any confidential material into a confidential annex which is clearlyidentified as containing confidential material. Oftel will take steps to protect theconfidentiality of all such material from the moment that it is received at Oftel'soffices. In the interests of transparency, respondents should avoid applyingconfidential markings wherever possible.

116

Annex A

NOTIFICATION OF PROPOSALS UNDER REGULATION 6 OF THE

ELECTRONIC COMMUNICATIONS (MARKET ANALYSIS) REGULATIONS

2003

Proposal for identifying markets, making market power determinations and

the setting of market power conditions

1. The Director General of Telecommunications (‘the Director’), in accordance with

regulation 6 of the Electronic Communications (Market Analysis) Regulations 2003

(S.I. 2003/330) (‘the Regulations’), hereby makes the following proposals for

identifying markets, making market power determinations and the setting of market

power conditions.

2. The Director is proposing to identify the following markets for the purpose of

making market power determinations:

(a) wholesale voice call termination provided to the subscribers of Hutchison

3G (UK) Limited, whose registered company number is 3885486 and any

Hutchison 3G (UK) Limited subsidiary or holding company, or any subsidiary of

that holding company, all as defined by Section 736 of the Companies Act

1985 as amended by the Companies Act 1989 (‘3’) (such termination being

provided via 3’s mobile network or via roaming on another provider’s network);

(b) wholesale voice call termination provided by Inquam Telecom (Holdings)

Limited, whose registered company number is 4244115 and any Inquam

Telecom (Holdings) Limited subsidiary or holding company, or any

subsidiary of that holding company, all as defined by Section 736 of the

Companies Act 1985 as amended by the Companies Act 1989 (‘Inquam’)

(such termination being provided via Inquam’s mobile network);

(c) wholesale voice call termination provided by O2 Limited, whose registered

company number is 2604354 and any O2 Limited subsidiary or holding

117

company, or any subsidiary of that holding company, all as defined by Section

736 of the Companies Act 1985 as amended by the Companies Act 1989 (‘O2’)

(such termination being provided via O2’s 2G and 3G mobile network);

(d) wholesale voice call termination provided by Orange plc, whose registered

company number is 3110666 and any Orange plc subsidiary or holding

company, or any subsidiary of that holding company, all as defined by Section

736 of the Companies Act 1985 as amended by the Companies Act 1989

(‘Orange’) (such termination being provided via Orange’s 2G and 3G mobile

network);

(e) wholesale voice call termination provided by T Mobile Limited, whose

registered company number is 4347402 and any T Mobile subsidiary or holding

company, or any subsidiary of that holding company, all as defined by Section

736 of the Companies Act 1985 as amended by the Companies Act 1989 (‘T

Mobile’) (such termination being provided via T Mobile’s 2G and 3G mobile

network);

(f) wholesale voice call termination provided by Vodafone Limited, whose

registered company number is 1471587 and any Vodafone Limited subsidiary

or holding company, or any subsidiary of that holding company, all as defined

by Section 736 of the Companies Act 1985 as amended by the Companies Act

1989 (‘Vodafone) (such termination being provided via Vodafone’s 2G and 3G

mobile network);

3. The Director is proposing to make market power determinations that the

following persons have significant market power in relation to the markets referred

to in paragraph 2 above:

(a) in relation to the market in sub-paragraph (a), 3;

(b)in relation to the market in sub-paragraph (b), Inquam;

(c)in relation to the market in sub-paragraph (c), O2;

118

(d) in relation to the market in sub-paragraph (d), Orange;

(e) in relation to the market in sub-paragraph (e), T Mobile;

(f) in relation to the market in sub-paragraph (f), Vodafone;

4. The Director is proposing to set market power conditions on the persons

referred to in paragraph 3 above as set out in Schedules 1, 2, 3 and 4 to this

Notification.

5. The effect of, and the Director’s reasons for making, the proposals to identify

the markets set out in paragraph 2 above are contained in chapter 3 of the

consultation document published with this Notification.

6. The effect of, and the Director’s reasons for making, the proposals to make the

market power determinations set out in paragraph 3 above are contained in

chapter 4 of the consultation document published with this Notification.

7. The effect of, and the Director’s reasons for making, the proposals to set the

market power conditions set out in Schedules 1, 2, 3 and 4 to this Notification are

contained in Chapter 6 of the consultation document published with this

Notification.

8 In identifying and analysing the markets referred to in paragraph 2 above, and in

considering whether to make the proposals set out in this Notification, the Director

has taken due account of all applicable guidelines and recommendations which

have been issued or made by the European Commission in pursuance of a

Community instrument, and relate to market identification and analysis or the

determination of what constitutes significant market power, in accordance with

Regulation 5 (2) of the regulations.

9. The Director considers that the proposed market power conditions referred to in

paragraph 4 above comply with, and are necessary for satisfying requirements in,

119

the provisions of the Directives, as appropriate and relevant to each of such

market power conditions, referred to in regulation 6 (8) of the Regulations.

10. Representations may be made to the Director about any of the proposals set

out in this Notification and the accompanying consultation document by xx July

2003.

11. Copies of this Notification have been sent to the Secretary of State, the

European Commission, and such of the regulatory authorities of other Member

States as the Director thinks fit.

NEIL BUCKLEY

Oftel Project Policy Director

A person duly authorised by the Director General pursuant to paragraph 8 of

Schedule 1 to the Telecommunications Act 1984

15 May 2003

120

SCHEDULE 1

Part 1: Definitions and Interpretation of this condition

1. The SMP condition in Part 2 of this Schedule shall, except insofar as it is

otherwise stated therein, apply to the market set out in paragraph 2(a) of the

Notification.

2. For the purpose of interpreting this Schedule the following definitions shall

apply:

“2G Public Electronic Communications Network” means a mobile Public ElectronicCommunications Network which operates using spectrum within the bands 880 to915 MHz, 925 to 960 MHz, 1710 to 1785 MHz, or 1805 to 1880 MHz;

"Act" means the Communications Act 2003;

"Director" means the Director General of Telecommunications as appointed under

section 1 of the Telecommunications Act 1984;

"Dominant Provider" means Hutchison 3G (UK) Limited whose registeredcompany number is 3885486 and any Hutchison 3G (UK) Limited subsidiary orholding company, or any subsidiary of that holding company, all as defined bySection 736 of the Companies Act 1985 as amended by the Companies Act 1989;

The condition proposed to be imposed on 3 under regulation 6(4) of theElectronic Communications (Market Analysis) Regulations 2003 as a result ofthe analysis of the market for wholesale voice call termination provided to thesubscribers of 3 (such termination being provided via 3’s 3G mobile network orvia roaming on another provider’s network) in which 3 has been found to havesignificant market power

121

“Network Access” means the resale by the Dominant Provider of

Interconnection to the 2G Public Electronic Communications Network of a

Communications Provider, together with any services, facilities or

arrangements which are necessary for the provision of Electronic

Communications Services over that Interconnection;

“Wireless Telegraphy" has the same meaning as in section 54 of the WirelessTelegraphy Act 1949.

3. Except insofar as the context otherwise requires, words or expressions shall

have the meaning assigned to them and otherwise any word or expression shall

have the same meaning as it has in the Act.

4. The Interpretation Act 1978 shall apply as if each of the conditions were an Act

of Parliament.

5. Headings and titles shall be disregarded.

122

Part 2: The condition

Condition MA1 – Basis of charges

MA1.1 Unless the Director directs otherwise from time to time, the Dominant

Provider shall secure, and shall be able to demonstrate to the satisfaction of the

Director, that each and every charge offered, payable or proposed for Network

Access is reasonably derived from the costs of provision based on a forward

looking incremental cost approach and allowing an appropriate mark up for the

recovery of common costs and an appropriate return on capital employed.

MA1.2 The Dominant Provider shall comply with any direction the Director may

make from time to time under this Condition.

123

SCHEDULE 2

The condition proposed to be imposed on Inquam under regulation 6(4)

of the Electronic Communications (Market Analysis) Regulations 2003 as

a result of the analysis of the market for wholesale voice call termination

provided by Inquam (such termination being provided via Inquam’s

mobile network) in which Inquam has been found to have significant

market power

Part 1: Definitions and Interpretation of these conditions

1. The SMP condition in Part 2 of this Schedule shall, except insofar as it is

otherwise stated therein, apply to the market set out in paragraph 2(b) of the

Notification.

2. For the purpose of interpreting this Schedule the following definitions shall

apply:

"Act" means the Communications Act 2003;

"Access Charge Change Notice" has the meaning given to it in Condition

MB1.7;

“Charges List” has the meaning given to it in Condition MB1.2;

"Director" means the Director General of Telecommunications as appointed

under section 1 of the Telecommunications Act 1984; and

"Dominant Provider" means Inquam Telecom (Holdings) Limited, whose

registered company number is 4244115 and any Inquam Telecom (Holdings)

Limited subsidiary or holding company, or any subsidiary of that holding

124

company, all as defined by Section 736 of the Companies Act 1985 as

amended by the Companies Act 1989.

3. Except insofar as the context otherwise requires, words or expressions shall

have the meaning assigned to them and otherwise any word or expression shall

have the same meaning as it has in the Act.

4. The Interpretation Act 1978 shall apply as if each of the conditions were an Act

of Parliament.

5. Headings and titles shall be disregarded.

125

Part 2: The condition

Condition MB1 – Requirement to publish and notify charges

MB1.1 Except in so far as the Director may otherwise consent in writing, the

Dominant Provider shall publish charges and act in the manner set out below.

MB 1.2 The Dominant Provider shall, within 1 month of the date that this Condition

comes into force, publish a list of its charges (the “Charges List”) for any Network

Access that it is providing as at the date that this Condition comes into force.

MB1.3 The Dominant Provider shall update the Charges List in relation to any

amendments or in relation to any further Network Access provided after the date

on which this Condition comes into force.

MB1.4 Publication referred to above shall be effected by:

a. placing a copy of the Charges List on any relevant website operated or

controlled by the Dominant Provider; and

b. sending a copy of the Charges List to the Director.

MB1.5 The Dominant Provider shall send a copy of the current version of the

Charges List to any person at that person’s written request (or such parts which

have been requested).

MB1.6 The Dominant Provider shall provide Network Access at the charges in the

Charges List and shall not depart therefrom either directly or indirectly.

MB1.7 The Dominant Provider shall send to the Director and to every person with

which it has entered into an Access Contract a written notice of any amendment to

the charges on which it provides Network Access or in relation to any charges for

126

new Network Access (an "Access Charge Change Notice") not less than 28 days

before any such amendment comes into effect.

MB1.8 The Dominant Provider shall ensure that an Access Charge Change Notice

includes:

a. a description of the Network Access in question; and

b. the date on which or the period for which any amendments to charges will take

effect (the "effective date").

MB1.9 The Dominant Provider shall not apply any new charge identified in an

Access Charge Change Notice before the effective date.

127

SCHEDULE 3

Part 1: Definitions and Interpretation of these conditions

1. The SMP conditions in Part 2 of this Schedule shall, except insofar as it is

otherwise stated therein, apply to the markets set out in paragraph 2(c) and

paragraph 2(f) of the Notification.

2. For the purpose of interpreting this Schedule the following definitions shall

apply:

“2G Public Electronic Communications Network” means a mobile Public ElectronicCommunications Network which operates using spectrum within the bands 880 to915 MHz, 925 to 960 MHz, 1710 to 1785 MHz, or 1805 to 1880 MHz;

"Act" means the Communications Act 2003;

"Access Charge Change Notice" has the meaning given to it in Condition MC6.2;

“Base Year” means-

(a) in relation to the first Relevant Year, the period starting 12 months prior to

the day on which the first Relevant Year commences and ending on 31 March

2003;

The conditions proposed to be imposed on O2 and Vodafone underregulation 6(4) of the Electronic Communications (Market Analysis)Regulations 2003 as a result of the analysis of the markets for wholesalevoice call termination (such termination being provided via O2 andVodafone’s 2G and 3G mobile network respectively) in which O2 andVodafone have been found to have significant market power

128

(b) in relation to the other Relevant Years, the period of 12 months ending on

31 March immediately preceding the Relevant Year in question;

“Call” means a circuit switched conveyance of a speech teleservice only (asdefined in the relevant standards of the European Telecommunications StandardsInstitute) originating in a Public Electronic Communications Network (whether fixedor mobile) and which terminates on a GSM mobile handset using the GSM airinterface for the conveyance of that speech call, which is connected to the 2GPublic Electronic Communications Network of the Dominant Provider. For thepurposes of this definition-

(a) “the relevant standards of the European Telecommunications

Standards Institute” means the European Telecommunications

Standard (ETS) of ETS 300 905 (GSM 02.03 version 5.3.2), Third

Edition, January 1998, which has been produced by the Special

Mobile Group of the European Telecommunications Standards

Institute; and

(b) “GSM” means the Global System for Mobile communications, as

defined in the relevant standards of the European Telecommunications

Standards Institute;

“Charging Period” means any of the current charging periods published by the

Dominant Provider;

"Director" means the Director General of Telecommunications as appointed

under section 1 of the Telecommunications Act 1984;

"Dominant Provider" means-

(a) O2 Limited, whose registered company number is 2604354;

(b) Vodafone Limited, whose registered company number is 1471587;

129

and any subsidiary or holding company of the companies listed in (a) to (b)

above, or any subsidiary of that holding company, all as defined by Section 736

of the Companies Act 1985 as amended by the Companies Act 1989;

“Donor Provider” means a Communications Provider whose Subscriber Number

has been passed to, or ported to the Dominant Provider;

“Fixed-to-Mobile Call” means a Call originating in a fixed Public Electronic

Communications Network only. For the avoidance of doubt, save where the

Director otherwise consents in writing it shall include Fixed-to-Mobile Calls to

Ported-In Numbers;

“Fixed-to-Mobile Interconnection Charge” means the published charge made by

the Dominant Provider for the Interconnection of a Fixed-to-Mobile Call, including

any amount received by the Dominant Provider from the Donor Provider for the

termination of Fixed-to-Mobile Calls to a Ported-In Number (save where the

Director consents in writing that Fixed-to-Mobile Calls shall exclude such calls to

Ported-In Numbers). It shall exclude any discounts offered by the Dominant

Provider, whether in respect of any particular Customer or any category of

Customers or any category of Calls;

“Functional Specification” shall have the same meaning as in Condition 21 of

the General Conditions of Entitlement;

“General Conditions of Entitlement” means those general conditions set by the

Director from time to time pursuant to section 42 of the Act;

“Mobile-to-Mobile Call” means a Call originating in a mobile Public Electronic

Communications Network of another Communications Provider. For the avoidance

of doubt, save where the Director otherwise consents in writing it shall include

Mobile-to-Mobile Calls to Ported-In Numbers;

130

“Mobile-to-Mobile Interconnection Charge” means the published charge made by

the Dominant Provider for the Interconnection of a Mobile-to-Mobile Call, including

any amount received by the Dominant Provider from the Donor Provider for the

termination of Mobile-to-Mobile Calls to a Ported-In Number (save where the

Director consents in writing that Mobile-to-Mobile Calls shall exclude such calls to

Ported-In Numbers). It shall exclude any discounts offered by the Dominant

Provider, whether in respect of any particular Customer or any category of

Customers or any category of Calls;

“Network Access” means the provision of Interconnection to the 2G Public

Electronic Communications Network provided by the Dominant Provider, together

with any services, facilities or arrangements which are necessary for the provision

of Electronic Communications Services over that Interconnection;

“Ported-In Number” means a Subscriber Number which has been passed to or

ported to the Dominant Provider;

"Reference Offer" means the terms and conditions on which the Dominant

Provider is willing to enter into an Access Contract;

“Relevant Year” means-

(a) the period beginning with the day on which this Condition comes into force

and ending with 31 March 2004 (‘the first Relevant Year’); or

(b) either of the following-

(i) the period of 12 months beginning on 1 April 2004 and ending on 31

March 2005;or

(ii) the period of 12 months beginning on 1 April 2005 and ending on 31

March 2006 (‘the last Relevant Year’);

131

“Retail Prices Index” means the index of retail prices compiled by an agency or a

public body on behalf of Her Majesty’s Government or a governmental department

from time to time in respect of all items (which is the Office for National Statistics at

the time of publication of this Notification);

“Subscriber Number” shall have the same meaning as in the Functional

Specification;

"Third Party" means a person providing a Public Electronic Communications

Network; and

“Wireless Telegraphy" has the same meaning as in section 54 of the WirelessTelegraphy Act 1949.

3. Except insofar as the context otherwise requires, words or expressions shallhave the meaning assigned to them and otherwise any word or expression shallhave the same meaning as it has in the Act.

4. The Interpretation Act 1978 shall apply as if each of the conditions were an Act

of Parliament.

5. Headings and titles shall be disregarded.

132

Part 2: The conditions

Condition MC1 – Requirement to provide network access on reasonable

request

MC1.1 The Dominant Provider shall provide Network Access to every Third Party

who reasonably requests in writing such Network Access and such Network

Access as the Director may direct from time to time.

MC1.2 The provision of Network Access in accordance with paragraph MC1.1

shall occur as soon as reasonably practicable and shall be provided on fair and

reasonable terms, conditions and charges and on such terms, conditions and

charges as the Director may from time to time direct.

MC1.3 The Dominant Provider shall comply with any direction the Director may

make from time to time under this Condition.

133

Condition MC2 – Requirement not to unduly discriminate

MC2.1 The Dominant Provider shall not unduly discriminate against particular

persons or against a particular description of persons, in relation to matters

connected with Network Access.

134

Condition MC3 – Control of Fixed-to-Mobile Interconnection Charges

MC3.1 Except in so far as the Director otherwise consents under paragraph

MC3.9 below, the Dominant Provider shall take all reasonable steps to secure

that, during any Relevant Year, the Average Interconnection Charge does not

exceed the Target Average Charge for any such Year.

MC3.2 In this Condition, the Average Interconnection Charge means the average

of the Fixed-to-Mobile Interconnection Charges during the Relevant Year in

question, which shall be weighted according to-

(a) the profile by Charging Period of the Dominant Provider’s minutes of

Fixed-to-Mobile Calls; and

(b) the volumes by month or part-month of the Dominant Provider’s minutes

of Fixed-to-Mobile Calls (except in so far as the Director otherwise consents

in writing that the weighting shall be derived from the sum of minutes of

Fixed-to-Mobile Calls and Mobile-to-Mobile Calls)

in the Base Year.

MC3.3 For the purposes of calculating the Average Interconnection Charge where

any Fixed-to-Mobile Interconnection Charges are in force during a part only of the

Relevant Year (commencing or ending at a date in the course of the Relevant

Year), the weighting shall be derived from-

(a) the profile by Charging Period of the Dominant Provider’s minutes of Fixed-to-

Mobile Calls; and

(b) the volumes by month or part-month of the Dominant Provider’s minutes of

Fixed-to-Mobile Calls (except in so far as the Director otherwise consents in

writing that the weighting shall be derived from the sum of minutes of Fixed-

to-Mobile Calls and Mobile-to-Mobile Calls)

135

in the corresponding part of the Base Year.

MC3.4 For the purposes of this Condition, the Target Average Charge means-

(a) for the purpose of the first Relevant Year 6.95 pence per minute; and

(b) for the purpose of the other Relevant Years, the Adjusted Base Target Charge

multiplied by the sum of 100% and the Controlling Percentage.

MC3.5 In paragraph MC3.4-

(a) the Adjusted Base Target Charge means the Base Target Charge divided by

the sum of 100% and the Adjustment Percentage; and

(b) the Base Target Charge means the average of the Fixed-to-Mobile

Interconnection Charges during the Base Year, weighted according to-

(i) the profile by Charging Period of the Dominant Provider’s minutes of Fixed-to-

Mobile Calls; and

(ii) the volumes by month or part-month of the Dominant Provider’s minutes of

Fixed-to-Mobile Calls (except in so far as the Director otherwise consents in

writing that the weighting shall be derived from the sum of minutes of Fixed-

to-Mobile Calls and Mobile-to-Mobile Calls)

in the Base Year.

MC3.6 For the purposes of calculating the Base Target Charge where any Fixed-

to-Mobile Interconnection Charges are in force during a part only of the Base Year

(commencing or ending at a date in the course of the Base Year), the weighting

shall be derived from-

(a) the profile by Charging Period of the Dominant Provider’s minutes of Fixed-

to-Mobile Calls; and

136

(b) the volumes by month or part-month of the Dominant Provider’s minutes of

Fixed-to-Mobile Calls (except in so far as the Director otherwise consents in

writing that the weighting shall be derived from the sum of minutes of Fixed-

to-Mobile Calls and Mobile-to-Mobile Calls)

in that part of the Base Year.

MC3.7 The Dominant Provider shall not make any Fixed-to-Mobile Interconnection

Charge for-

(a) a Fixed-to-Mobile Call which terminates on a recorded announcement provided

by the Dominant Provider informing the caller of an inability to complete that

call so as to establish a two-way path where the mobile handset used by the

called party is switched off, or rings and remains unanswered, or where

coverage is not available from the Dominant Provider’s 2G Public Electronic

Communications Network; and

(b) an unanswered Fixed-to-Mobile Call which is diverted in respect of the period

before that call is answered.

MC3.8 Notwithstanding (and without prejudice to the generality of) the obligation

imposed on the Dominant Provider by paragraph MC3.1 above,

(a) if the Dominant Provider has failed to secure that the Average Interconnection

Charge has not exceeded the Target Average Charge for any Relevant Year,

the Dominant Provider shall make such adjustments to its Fixed-to-Mobile

Interconnection Charges and by such day in the following Relevant Year as the

Director may direct for the purpose of remedying that failure. Such adjustments

in the following Relevant Year shall not be relevant for the purpose of

establishing compliance with paragraph MC3.1 above in that Relevant Year;

and

137

(b) if it appears to the Director that the Dominant Provider is likely to fail to secure

that the Average Interconnection Charge for the last Relevant Year does not

exceed the Target Average Charge for that Year, the Dominant Provider shall

make such adjustments to its Fixed-to-Mobile Interconnection Charges and by

such day in that Year as the Director may direct for the purpose of avoiding

that failure.

MC3.9 Where the Average Interconnection Charge is less than the Target

Average Charge for any Relevant Year, the Dominant Provider shall not make

such adjustments to its Fixed-to-Mobile Interconnection Charges in the following

Relevant Year to recover the difference between the Average Interconnection

Charge and the Target Average Charge for the Relevant Year in question, unless

the Director has given his prior written consent to such adjustments. Such

adjustments in the following Relevant Year shall not be relevant for the purpose of

establishing compliance with paragraph MC3.1 in that Relevant Year.

MC3.10 In this Condition-

‘Adjusted Base Target Charge’ has the meaning given to it in paragraph MC3.5;

‘Adjustment Percentage’ means, in relation to any Relevant Year, the percentage

by which the Average Interconnection Charge in the Base Year was equal to,

above or below (as the case may be) the Target Average Charge in the Base

Year. For the avoidance of doubt, the Adjustment Percentage shall be-

(a) negative, if the Average Interconnection Charge in the Base Year was below

the Target Average Charge in the Base Year;

(b) zero, if the Average Interconnection Charge in the Base Year was equal to the

Target Average Charge in the Base Year; or

(c) positive, if the Average Interconnection Charge in the Base Year was above

the Target Average Charge in the Base Year;

138

‘Average Interconnection Charge’ has the meaning given to it in paragraph MC3.2;

‘Base Target Charge’ has the meaning given to it in paragraph MC3.5;

‘Controlling Percentage’ means, in relation to any Relevant Year, the amount of

change in the Retail Prices Index in the period of 12 months ending on 31

December immediately before the beginning of that Year, expressed as a

percentage (rounded to two decimal places) of that Index as at the beginning of

that period, reduced by 15%; and

‘Target Average Charge’ shall have the meaning given to it in paragraph MC3.4.

139

Condition MC4 - Control of Mobile to Mobile Interconnection Charges

MC4.1 Except in so far as the Director otherwise consents under paragraph

MC4.9 below, the Dominant Provider shall take all reasonable steps to secure

that, during any Relevant Year, the Average Interconnection Charge does not

exceed the Target Average Charge for any such Year.

MC4.2 In this Condition, the Average Interconnection Charge means the average

of the Mobile-to-Mobile Interconnection Charges during the Relevant Year in

question, which shall be weighted according to-

(a) the profile by Charging Period of the Dominant Provider’s minutes of Mobile-

to-Mobile Calls; and

(b) the volumes by month or part-month of the Dominant Provider’s minutes of

Mobile-to-Mobile Calls (except in so far as the Director otherwise consents in

writing that the weighting shall be derived from the sum of minutes of Fixed-to-

Mobile Calls and Mobile-to-Mobile Calls)

in the Base Year.

MC4.3 For the purposes of calculating the Average Interconnection Charge where

any Mobile-to-Mobile Interconnection Charges are in force during a part only of the

Relevant Year (commencing or ending at a date in the course of the Relevant

Year), the weighting shall be derived from-

(a) the profile by Charging Period of the Dominant Provider’s minutes of Mobile-to-

Mobile Calls; and

(b) the volumes by month or part-month of the Dominant Provider’s minutes of

Mobile-to-Mobile Calls (except in so far as the Director otherwise consents in

writing that the weighting shall be derived from the sum of minutes of Fixed-to-

Mobile Calls and Mobile-to-Mobile Calls)

140

in the corresponding part of the Base Year.

MC4.4 For the purposes of this Condition, the Target Average Charge means-

(a) for the purpose of the first Relevant Year, 6.95 pence per minute; and

(b) for the purpose of the other Relevant Years, the Adjusted Base Target Charge

multiplied by the sum of 100% and the Controlling Percentage.

MC4.5 In paragraph MC4.4-

(a) the Adjusted Base Target Charge means the Base Target Charge divided by

the sum of 100% and the Adjustment Percentage; and

(b) the Base Target Charge means the average of the Mobile-to-Mobile

Interconnection Charges during the Base Year, weighted according to-

(i) the profile by Charging Period of the Dominant Provider’s minutes of

Mobile-to-Mobile Calls; and

(ii) the volumes by month or part-month of the Dominant Provider’s minutes

of Mobile-to-Mobile Calls (except in so far as the Director otherwise

consents in writing that the weighting shall be derived from the sum of

minutes of Fixed-to-Mobile Calls and Mobile-to-Mobile Calls)

in the Base Year.

MC 4.6 For the purposes of calculating the Base Target Charge where any Mobile-

to-Mobile Interconnection Charges are in force during a part only of the Base Year

(commencing or ending at a date in the course of the Base Year), the weighting

shall be derived from-

(a) the profile by Charging Period of the Dominant Provider’s minutes of

Mobile-to-Mobile Calls; and

141

(b) the volumes by month or part-month of the Dominant Provider’s minutes of

Mobile-to-Mobile Calls (except in so far as the Director otherwise consents

in writing that the weighting shall be derived from the sum of minutes of

Fixed-to-Mobile Calls and Mobile-to-Mobile Calls)

in that part of the Base Year.

MC4.7 The Dominant Provider shall not make any Mobile-to-Mobile

Interconnection Charge for-

(a) a Mobile-to-Mobile Call which terminates on a recorded announcement

provided by the Dominant Provider informing the caller of an inability to

complete that call so as to establish a two-way path where the mobile handset

used by the called party is switched off, or rings and remains unanswered, or

where coverage is not available from the Dominant Provider’s 2G Public

Electronic Communications Network; and

(b) an unanswered Mobile-to-Mobile Call which is diverted in respect of the period

before that call is answered.

MC4.8 Notwithstanding (and without prejudice to the generality of) the obligation

imposed on the Dominant Provider by paragraph MC4.1 above,

(a) if the Dominant Provider has failed to secure that the Average Interconnection

Charge has not exceeded the Target Average Charge for any Relevant Year,

the Dominant Provider shall make such adjustments to its Mobile-to-Mobile

Interconnection Charges and by such day in the following Relevant Year as the

Director may direct for the purpose of remedying that failure. Such adjustments

in the following Relevant Year shall not be relevant for the purpose of

establishing compliance with paragraph MC4.1 above in that Relevant Year;

and

142

(b) if it appears to the Director that the Dominant Provider is likely to fail to secure

that the Average Interconnection Charge for the last Relevant Year does not

exceed the Target Average Charge for that Year, the Dominant Provider shall

make such adjustments to its Mobile-to-Mobile Interconnection Charges and by

such day in that Year as the Director may direct for the purpose of avoiding

that failure.

MC4.9 Where the Average Interconnection Charge is less than the Target

Average Charge for any Relevant Year, the Dominant Provider shall not make

such adjustments to its Mobile-to-Mobile Interconnection Charges in the following

Relevant Year to recover the difference between the Average Interconnection

Charge and the Target Average Charge for the Relevant Year in question, unless

the Director has given his prior written consent to such adjustments. Such

adjustments in the following Relevant Year shall not be relevant for the purpose of

establishing compliance with paragraph MC4.1 in that Relevant Year.

MC4.10 In this Condition-

‘Adjusted Base Target Charge’ has the meaning given to it in paragraph MC4.5;

‘Adjustment Percentage’ means, in relation to any Relevant Year, the percentage

by which the Average Interconnection Charge in the Base Year was equal to,

above or below (as the case may be) the Target Average Charge in the Base

Year. For the avoidance of doubt, the Adjustment Percentage shall be-

(a) negative, if the Average Interconnection Charge in the Base Year was

below the Target Average Charge in the Base Year;

(b) zero, if the Average Interconnection Charge in the Base Year was equal to

the Target Average Charge in the Base Year; or

(c) positive, if the Average Interconnection Charge in the Base Year was above

the Target Average Charge in the Base Year;

143

‘Average Interconnection Charge’ has the meaning given to it in paragraph MC4.2;

‘Base Target Charge’ has the meaning given to it in paragraph MC4.5;

‘Controlling Percentage’ means, in relation to any Relevant Year, the amount of

change in the Retail Prices Index in the period of 12 months ending on 31

December immediately before the beginning of that Year, expressed as a

percentage (rounded to two decimal places) of that Index as at the beginning of

that period, reduced by 15%; and

‘Target Average Charge’ shall have the meaning given to it in paragraph MC4.4.

144

Condition MC5 – Requirement to publish a reference offer

MC5.1 Except in so far as the Director may otherwise consent in writing, the

Dominant Provider shall publish a Reference Offer and act in the manner set out

below.

MC5.2 Subject to paragraph MC5.7 below, the Dominant Provider shall ensure

that a Reference Offer in relation to the provision of Network Access includes at

least the following:

a. a description of the Network Access to be provided, including technical

characteristics (which shall include information on network configuration where

necessary to make effective use of the network);

b. the locations of the points of Network Access;

c. the technical standards for Network Access (including any usage restrictions

and other security issues);

d. the conditions for access to ancillary, supplementary and advanced services

(including operational support systems, information systems or databases for pre-

ordering, provisioning, ordering, maintenance and repair requests and billing);

e. any ordering and provisioning procedures;

f. relevant charges, terms of payment and billing procedures;

g. details of interoperability tests;

h. details of traffic and network management;

i. details of maintenance and quality as follows:

145

(i) specific time scales for the acceptance or refusal of a request for

supply and for completion, testing and hand-over or delivery of

services and facilities, for provision of support services (such as fault

handling and repair);

(ii) service level commitments, namely the quality standards that

each party must meet when performing its contractual obligations;

(iii) the amount of compensation payable by one party to another for

failure to perform contractual commitments;

(iv) a definition and limitation of liability and indemnity; and

(v) procedures in the event of alterations being proposed to the

service offerings, for example, launch of new services, changes to

existing services or change to prices;

j. details of measures to ensure compliance with requirements for network

integrity;

k. details of any relevant intellectual property rights;

l. a dispute resolution procedure to be used between the parties;

m. details of duration and renegotiation of agreements;

n. provisions regarding confidentiality of non-public parts of the agreements;

o. rules of allocation between the parties when supply is limited (for example, for

the purpose of co-location or location of masts);

p. the standard terms and conditions for the provision of Network Access;

146

MC5.3 The Dominant Provider shall, within 1 month of the date that this Condition

comes into force, publish a Reference Offer in relation to any Network Access that

it is providing as at the date that this Condition comes into force.

MC5.4 The Dominant Provider shall update the Reference Offer in relation to any

amendments or in relation to any further Network Access provided after the date

on which this Condition comes into force.

MC5.5 Publication referred to above shall be effected by-

a. placing a copy of the Reference Offer on any relevant website operated or

controlled by the Dominant Provider; and

b. sending a copy of the Reference Offer to the Director.

MC5.6 The Dominant Provider shall send a copy of the current version of the

Reference Offer to any person at that person’s written request (or such parts

which have been requested).

MC5.7 The Dominant Provider shall make such modifications to the Reference

Offer as the Director may direct from time to time.

MC5.8 The Dominant Provider shall provide Network Access at the charges, terms

and conditions in the relevant Reference Offer and shall not depart therefrom

either directly or indirectly.

MC5.9 The Dominant Provider shall comply with any direction the Director may

make from time to time under this Condition.

147

Condition MC6 – Requirement to notify charges

MC6.1 Except in so far as the Director may otherwise consent in writing, the

Dominant Provider shall publish charges and act in the manner set out below.

MC6.2 The Dominant Provider shall send to the Director and to every person with

which it has entered into an Access Contract a written notice of any amendment to

the charges on which it provides Network Access or in relation to any charges for

new Network Access (an "Access Charge Change Notice") not less than 28 days

before any such amendment comes into effect.

MC6.3 The Dominant Provider shall ensure that an Access Charge Change Notice

includes:

a. a description of the Network Access in question;

b. a reference to the location in the Dominant Provider’s current Reference Offer of

the terms and conditions associated with the provision of that Network Access;

c. the date on which or the period for which any amendments to charges will take

effect (the "effective date");

d. the relevant network tariff gradient.

MC6.4 The Dominant Provider shall not apply any new charge identified in an

Access Charge Change Notice before the effective date.

148

SCHEDULE 4

Part 1: Definitions and Interpretation of these conditions

1. The SMP conditions in Part 2 of this Schedule shall, except insofar as it is

otherwise stated therein, apply to the markets set out in paragraphs 2(d) - (e) of

the Notification.

2. For the purpose of interpreting this Schedule the following definitions shall

apply:

“2G Public Electronic Communications Network” means a mobile Public Electronic

Communications Network which operates using spectrum within the bands 880 to

915 MHz, 925 to 960 MHz, 1710 to 1785 MHz, or 1805 to 1880 MHz;

"Act" means the Communications Act 2003;

"Access Charge Change Notice" has the meaning given to it in Condition

MD6.2;

“Base Year” means-

The conditions proposed to be imposed on Orange and T Mobile underregulation 6(4) of the Electronic Communications (Market Analysis)Regulations 2003 as a result of the analysis of the markets for wholesalevoice call termination (such termination being provided via Orange and TMobile’s 2G and 3G mobile network respectively) in which Orange and TMobile have been found to have significant market power

149

(a)in relation to the first Relevant Year, the period starting 12 months prior to

the day on which the first Relevant Year commences and ending on 31 March

2003;

(b) in relation to the other Relevant Years, the period of 12 months ending on

31 March immediately preceding the Relevant Year in question;

“Call” means a circuit switched conveyance of a speech teleservice only (as

defined in the relevant standards of the European Telecommunications Standards

Institute) originating in a Public Electronic Communications Network (whether fixed

or mobile) and which terminates on a GSM mobile handset using the GSM air

interface for the conveyance of that speech call, which is connected to the 2G

Public Electronic Communications Network of the Dominant Provider. For the

purposes of this definition-

(a) “the relevant standards of the European Telecommunications

Standards Institute” means the European Telecommunications

Standard (ETS) of ETS 300 905 (GSM 02.03 version 5.3.2), Third

Edition, January 1998, which has been produced by the Special

Mobile Group of the European Telecommunications Standards

Institute.

(b) “GSM” means the Global System for Mobile communications, as

defined in the relevant standards of the European Telecommunications

Standards Institute.

“Charging Period” means any of the current charging periods published by the

Dominant Provider;

"Director" means the Director General of Telecommunications as appointed

under section 1 of the Telecommunications Act 1984;

150

"Dominant Provider" means-

(a) Orange plc, whose registered company number is 3110666;

(b) T-Mobile Limited, whose registered company number is 4347402;

and any subsidiary or holding company of the companies listed in (a) to (b)

above, or any subsidiary of that holding company, all as defined by Section 736

of the Companies Act 1985 as amended by the Companies Act 1989;

“Donor Provider” means a Communications Provider whose Subscriber

Number has been passed to, or ported to the Dominant Provider;

“Fixed-to-Mobile Call” means a Call originating in a fixed Public Electronic

Communications Network only. For the avoidance of doubt, save where the

Director otherwise consents in writing it shall include Fixed-to-Mobile Calls to

Ported-In Numbers;

“Fixed-to-Mobile Interconnection Charge” means the published charge made by

the Dominant Provider for the Interconnection of a Fixed-to-Mobile Call, including

any amount received by the Dominant Provider from the Donor Provider for the

termination of Fixed-to-Mobile Calls to a Ported-In Number (save where the

Director consents in writing that Fixed-to-Mobile Calls shall exclude such calls to

Ported-In Numbers). It shall exclude any discounts offered by the Dominant

Provider, whether in respect of any particular Customer or any category of

Customers or any category of Calls;

“Functional Specification” shall have the same meaning as Condition 21 of the

General Conditions of Entitlement;

“General Conditions of Entitlement” means those general conditions set by the

Director from time to time pursuant to section 42 of the Act;

151

“Mobile-to-Mobile Call” means a Call originating in a mobile Public Electronic

Communications Network of another Communications Provider. For the avoidance

of doubt, save where the Director otherwise consents in writing it shall include

Mobile-to-Mobile Calls to Ported-In Numbers;

“Mobile-to-Mobile Interconnection Charge” means the published charge made by

the Dominant Provider for the Interconnection of a Mobile-to-Mobile Call, including

any amount received by the Dominant Provider from the Donor Provider for the

termination of Mobile-to-Mobile Calls to a Ported-In Number (save where the

Director consents in writing that Mobile-to-Mobile Calls shall exclude such calls to

Ported-In Numbers). It shall exclude any discounts offered by the Dominant

Provider, whether in respect of any particular Customer or any category of

Customers or any category of Calls;

“Network Access” means the provision of Interconnection to the 2G Public

Electronic Communications Network provided by the Dominant Provider,

together with any services, facilities or arrangements which are necessary for

the provision of Electronic Communications Services over that Interconnection;

“Ported-In Number” means a Subscriber Number which has been passed to or

ported to the Dominant Provider;

"Reference Offer" means the terms and conditions on which the Dominant

Provider is willing to enter into an Access Contract;

“Relevant Year” means-

(a) the period beginning with the day on which this Condition comes into force

and ending with 31 March 2004 (‘the first Relevant Year’); or

(b) either of the following-

152

(c) the period of 12 months beginning on 1 April 2004 and ending on 31

March 2005;or

(iii) the period of 12 months beginning on 1 April 2005 and ending on 31

March 2006 (‘the last Relevant Year’);

“Retail Prices Index” means the index of retail prices compiled by an agency or a

public body on behalf of Her Majesty’s Government or a governmental department

from time to time in respect of all items (which is the Office for National Statistics at

the time of publication of this Notification);

“Subscriber Number” shall have the same meaning as in the FunctionalSpecification;

"Third Party" means a person providing a Public Electronic Communications

Network; and

“Wireless Telegraphy" has the same meaning as in section 54 of the Wireless

Telegraphy Act 1949.

3. Except insofar as the context otherwise requires, words or expressions shall

have the meaning assigned to them and otherwise any word or expression shall

have the same meaning as it has in the Act.

4. The Interpretation Act 1978 shall apply as if each of the conditions were an Act

of Parliament.

5. Headings and titles shall be disregarded.

153

Part 2: The conditions

Condition MD1 – Requirement to provide network access on reasonable

request

MD1.1 The Dominant Provider shall provide Network Access to every Third Party

who reasonably requests in writing such Network Access and such Network

Access as the Director may direct from time to time.

MD1.2 The provision of Network Access in accordance with paragraph MD1.1

shall occur as soon as reasonably practicable and shall be provided on fair and

reasonable terms, conditions and charges and on such terms, conditions and

charges as the Director may from time to time direct.

MD1.3 The Dominant Provider shall comply with any direction the Director may

make from time to time under this Condition.

154

Condition MD2 – Requirement not to unduly discriminate

MD2.1 The Dominant Provider shall not unduly discriminate against particular

persons or against a particular description of persons, in relation to matters

connected with Network Access.

155

Condition MD3 – Control of Fixed-to-Mobile Interconnection Charges

MD3.1 Except in so far as the Director otherwise consents under paragraph

MD3.9 below, the Dominant Provider shall take all reasonable steps to secure

that, during any Relevant Year, the Average Interconnection Charge does not

exceed the Target Average Charge for any such Year.

MD3.2 In this Condition, the Average Interconnection Charge means the average

of the Fixed-to-Mobile Interconnection Charges during the Relevant Year in

question, which shall be weighted according to-

(a) the profile by Charging Period of the Dominant Provider’s minutes of Fixed-to-

Mobile Calls; and

(b) the volumes by month or part-month of the Dominant Provider’s minutes of

Fixed-to-Mobile Calls (except in so far as the Director otherwise consents in

writing that the weighting shall be derived from the sum of minutes of Fixed-to-

Mobile Calls and Mobile-to-Mobile Calls)

in the Base Year.

MD3.3 For the purposes of calculating the Average Interconnection Charge where

any Fixed-to-Mobile Interconnection Charges are in force during a part only of the

Relevant Year (commencing or ending at a date in the course of the Relevant

Year), the weighting shall be derived from-

(a) the profile by Charging Period of the Dominant Provider’s minutes of Fixed-to-

Mobile Calls; and

(b) the volumes by month or part-month of the Dominant Provider’s minutes of

Fixed-to-Mobile Calls (except in so far as the Director otherwise consents in

writing that the weighting shall be derived from the sum of minutes of Fixed-to-

Mobile Calls and Mobile-to-Mobile Calls)

156

in the corresponding part of the Base Year.

MD3.4 For the purposes of this Condition, the Target Average Charge means-

(a) for the purpose of the first Relevant Year, 7.70 pence per minute; and

(b) for the purpose of the other Relevant Years, the Adjusted Base Target Charge

multiplied by the sum of 100% and the Controlling Percentage.

MD3.5 In paragraph MD3.4-

(a) the Adjusted Base Target Charge means the Base Target Charge divided by

the sum of 100% and the Adjustment Percentage; and

(b) the Base Target Charge means the average of the Fixed-to-Mobile

Interconnection Charges during the Base Year, weighted according to-

(i) the profile by Charging Period of the Dominant Provider’s minutes of

Fixed-to-Mobile Calls; and

(ii) the volumes by month or part-month of the Dominant Provider’s

minutes of Fixed-to-Mobile Calls (except in so far as the Director

otherwise consents in writing that the weighting shall be derived from

the sum of minutes of Fixed-to-Mobile Calls and Mobile-to-Mobile

Calls)

in the Base Year.

MD3.6 For the purposes of calculating the Base Target Charge where any Fixed-

to-Mobile Interconnection Charges are in force during a part only of the Base Year

(commencing or ending at a date in the course of the Base Year), the weighting

shall be derived from-

157

(a) the profile by Charging Period of the Dominant Provider’s minutes of Fixed-

to-Mobile Calls; and

(b) the volumes by month or part-month of the Dominant Provider’s minutes of

Fixed-to-Mobile Calls (except in so far as the Director otherwise consents in

writing that the weighting shall be derived from the sum of minutes of Fixed-

to-Mobile Calls and Mobile-to-Mobile Calls)

in that part of the Base Year.

MD3.7 The Dominant Provider shall not make any Fixed-to-Mobile Interconnection

Charge for-

(a) a Fixed-to-Mobile Call which terminates on a recorded announcement provided

by the Dominant Provider informing the caller of an inability to complete that

call so as to establish a two-way path where the mobile handset used by the

called party is switched off, or rings and remains unanswered, or where

coverage is not available from the Dominant Provider’s 2G Public Electronic

Communications Network; and

(b) an unanswered Fixed-to-Mobile Call which is diverted in respect of the period

before that call is answered.

MD3.8 Notwithstanding (and without prejudice to the generality of) the obligation

imposed on the Dominant Provider by paragraph MD3.1 above,

(a) if the Dominant Provider has failed to secure that the Average Interconnection

Charge has not exceeded the Target Average Charge for any Relevant Year,

the Dominant Provider shall make such adjustments to its Fixed-to-Mobile

Interconnection Charges and by such day in the following Relevant Year as the

Director may direct for the purpose of remedying that failure. Such adjustments

in the following Relevant Year shall not be relevant for the purpose of

establishing compliance with paragraph MD3.1 above in that Relevant Year;

and

158

(b) if it appears to the Director that the Dominant Provider is likely to fail to secure

that the Average Interconnection Charge for the last Relevant Year does not

exceed the Target Average Charge for that Year, the Dominant Provider shall

make such adjustments to its Fixed-to-Mobile Interconnection Charges and by

such day in that Year as the Director may direct for the purpose of avoiding

that failure.

MD3.9 Where the Average Interconnection Charge is less than the Target

Average Charge for any Relevant Year, the Dominant Provider shall not make

such adjustments to its Fixed-to-Mobile Interconnection Charges in the following

Relevant Year to recover the difference between the Average Interconnection

Charge and the Target Average Charge for the Relevant Year in question, unless

the Director has given his prior written consent to such adjustments. Such

adjustments in the following Relevant Year shall not be relevant for the purpose of

establishing compliance with paragraph MD3.1 in that Relevant Year.

MD3.10 In this Condition-

‘Adjusted Base Target Charge’ has the meaning given to it in paragraph MD3.5;

‘Adjustment Percentage’ means, in relation to any Relevant Year, the percentage

by which the Average Interconnection Charge in the Base Year was equal to,

above or below (as the case may be) the Target Average Charge in the Base

Year. For the avoidance of doubt, the Adjustment Percentage shall be-

(a) negative, if the Average Interconnection Charge in the Base Year was below

the Target Average Charge in the Base Year;

(b) zero, if the Average Interconnection Charge in the Base Year was equal to the

Target Average Charge in the Base Year; or

(c) positive, if the Average Interconnection Charge in the Base Year was above

the Target Average Charge in the Base Year;

159

‘Average Interconnection Charge’ has the meaning given to it in paragraph MD3.2;

‘Base Target Charge’ has the meaning given to it in paragraph MD3.5;

‘Controlling Percentage’ means, in relation to any Relevant Year, the amount of

change in the Retail Prices Index in the period of 12 months ending on 31

December immediately before the beginning of that Year, expressed as a

percentage (rounded to two decimal places) of that Index as at the beginning of

that period, reduced by 14%; and

‘Target Average Charge’ shall have the meaning given to it in paragraph MD3.4.

160

Condition MD4 - Control of Mobile to Mobile Interconnection Charges

MD4.1 Except in so far as the Director otherwise consents under paragraph

MD4.8 below, the Dominant Provider shall take all reasonable steps to secure

that, during any Relevant Year, the Average Interconnection Charge does not

exceed the Target Average Charge for any such Year.

MD4.2 In this Condition, the Average Interconnection Charge means the average

of the Mobile-to-Mobile Interconnection Charges during the Relevant Year in

question, which shall be weighted according to-

(a) the profile by Charging Period of the Dominant Provider’s minutes of Mobile-to-

Mobile Calls; and

(b) the volumes by month or part-month of the Dominant Provider’s minutes of

Mobile-to-Mobile Calls (except in so far as the Director otherwise consents in

writing that the weighting shall be derived from the sum of minutes of Fixed-to-

Mobile Calls and Mobile-to-Mobile Calls)

in the Base Year.

MD4.3 For the purposes of calculating the Average Interconnection Charge where

any Mobile-to-Mobile Interconnection Charges are in force during a part only of the

Relevant Year (commencing or ending at a date in the course of the Relevant

Year), the weighting shall be derived from-

(a) the profile by Charging Period of the Dominant Provider’s minutes of Mobile-to-

Mobile Calls; and

(b) the volumes by month or part-month of the Dominant Provider’s minutes of

Mobile-to-Mobile Calls (except in so far as the Director otherwise consents in

writing that the weighting shall be derived from the sum of minutes of Fixed-to-

Mobile Calls and Mobile-to-Mobile Calls)

161

in the corresponding part of the Base Year.

MD4.4 For the purposes of this Condition, the Target Average Charge means-

(a) for the purpose of the first Relevant Year, 7.70 pence per minute; and

(b) for the purpose of the other Relevant Years, the Adjusted Base Target Charge

multiplied by the sum of 100% and the Controlling Percentage.

MD4.5 In paragraph MD4.4-

(a) the Adjusted Base Target Charge means the Base Target Charge divided by

the sum of 100% and the Adjustment Percentage; and

(b) the Base Target Charge means the average of the Mobile-to-Mobile

Interconnection Charges during the Base Year, weighted according to-

(i) the profile by Charging Period of the Dominant Provider’s minutes of

Mobile-to-Mobile Calls; and

(ii) the volumes by month or part-month of the Dominant Provider’s

minutes of Mobile-to-Mobile Calls (except in so far as the Director

otherwise consents in writing that the weighting shall be derived from

the sum of minutes of Fixed-to-Mobile Calls and Mobile-to-Mobile

Calls)

in the Base Year.

MD 4.6 For the purposes of calculating the Base Target Charge where any Mobile-

to-Mobile Interconnection Charges are in force during a part only of the Base Year

(commencing or ending at a date in the course of the Base Year), the weighting

shall be derived from-

(a) the profile by Charging Period of the Dominant Provider’s minutes of

Mobile-to-Mobile Calls; and

162

(b) the volumes by month or part-month of the Dominant Provider’s minutes of

Mobile-to-Mobile Calls (except in so far as the Director otherwise consents

in writing that the weighting shall be derived from the sum of minutes of

Fixed-to-Mobile Calls and Mobile-to-Mobile Calls)

in that part of the Base Year.

MD4.7 The Dominant Provider shall not make any Mobile-to-Mobile

Interconnection Charge for-

(a) a Mobile-to-Mobile Call which terminates on a recorded announcement

provided by the Dominant Provider informing the caller of an inability to

complete that call so as to establish a two-way path where the mobile handset

used by the called party is switched off, or rings and remains unanswered, or

where coverage is not available from the Dominant Provider’s 2G Public

Electronic Communications Network; and

(b) an unanswered Mobile-to-Mobile Call which is diverted in respect of the period

before that call is answered.

MD4.8 Notwithstanding (and without prejudice to the generality of) the obligation

imposed on the Dominant Provider by paragraph MD4.1 above,

(a) if the Dominant Provider has failed to secure that the Average Interconnection

Charge has not exceeded the Target Average Charge for any Relevant Year,

the Dominant Provider shall make such adjustments to its Mobile-to-Mobile

Interconnection Charges and by such day in the following Relevant Year as the

Director may direct for the purpose of remedying that failure. Such adjustments

in the following Relevant Year shall not be relevant for the purpose of

establishing compliance with paragraph MD4.1 above in that Relevant Year;

and

163

(b) if it appears to the Director that the Dominant Provider is likely to fail to secure

that the Average Interconnection Charge for the last Relevant Year does not

exceed the Target Average Charge for that Year, the Dominant Provider shall

make such adjustments to its Mobile-to-Mobile Interconnection Charges and by

such day in that Year as the Director may direct for the purpose of avoiding

that failure.

MD4.9 Where the Average Interconnection Charge is less than the Target

Average Charge for any Relevant Year, the Dominant Provider shall not make

such adjustments to its Mobile-to-Mobile Interconnection Charges in the following

Relevant Year to recover the difference between the Average Interconnection

Charge and the Target Average Charge for the Relevant Year in question, unless

the Director has given his prior written consent to such adjustments. Such

adjustments in the following Relevant Year shall not be relevant for the purpose of

establishing compliance with paragraph MD4.1 in that Relevant Year.

MD4.10 In this Condition-

‘Adjusted Base Target Charge’ has the meaning given to it in paragraph MD4.5;

‘Adjustment Percentage’ means, in relation to any Relevant Year, the percentage

by which the Average Interconnection Charge in the Base Year was equal to,

above or below (as the case may be) the Target Average Charge in the Base

Year. For the avoidance of doubt, the Adjustment Percentage shall be-

(a) negative, if the Average Interconnection Charge in the Base Year was below

the Target Average Charge in the Base Year;

(b) zero, if the Average Interconnection Charge in the Base Year was equal to the

Target Average Charge in the Base Year; or

(c) positive, if the Average Interconnection Charge in the Base Year was above

the Target Average Charge in the Base Year;

164

‘Average Interconnection Charge’ has the meaning given to it in paragraph MD4.2;

‘Base Target Charge’ has the meaning given to it in paragraph MD4.5;

‘Controlling Percentage’ means, in relation to any Relevant Year, the amount of

change in the Retail Prices Index in the period of 12 months ending on 31

December immediately before the beginning of that Year, expressed as a

percentage (rounded to two decimal places) of that Index as at the beginning of

that period, reduced by 14%; and

‘Target Average Charge’ shall have the meaning given to it in paragraph MD4.4.

165

Condition MD5 – Requirement to publish a reference offer

MD5.1 Except in so far as the Director may otherwise consent in writing, the

Dominant Provider shall publish a Reference Offer and act in the manner set out

below.

MD5.2 Subject to paragraph MD5.7 below, the Dominant Provider shall ensure

that a Reference Offer in relation to the provision of Network Access includes at

least the following:

a. a description of the Network Access to be provided, including technical

characteristics (which shall include information on network configuration where

necessary to make effective use of the network);

b. the locations of the points of Network Access;

c. the technical standards for Network Access (including any usage restrictions

and other security issues);

d. the conditions for access to ancillary, supplementary and advanced services

(including operational support systems, information systems or databases for pre-

ordering, provisioning, ordering, maintenance and repair requests and billing);

e. any ordering and provisioning procedures;

f. relevant charges, terms of payment and billing procedures;

g. details of interoperability tests;

h. details of traffic and network management;

i. details of maintenance and quality as follows:

166

(i) specific time scales for the acceptance or refusal of a request for

supply and for completion, testing and hand-over or delivery of

services and facilities, for provision of support services (such as fault

handling and repair);

(ii) service level commitments, namely the quality standards that

each party must meet when performing its contractual obligations;

(iii) the amount of compensation payable by one party to another for

failure to perform contractual commitments;

(iv) a definition and limitation of liability and indemnity; and

(v) procedures in the event of alterations being proposed to the

service offerings, for example, launch of new services, changes to

existing services or change to prices;

j. details of measures to ensure compliance with requirements for network

integrity;

k. details of any relevant intellectual property rights;

l. a dispute resolution procedure to be used between the parties;

m. details of duration and renegotiation of agreements;

n. provisions regarding confidentiality of non-public parts of the agreements;

o. rules of allocation between the parties when supply is limited (for example, for

the purpose of co-location or location of masts);

p. the standard terms and conditions for the provision of Network Access;

167

MD5.3 The Dominant Provider shall, within 1 month of the date that this Condition

comes into force, publish a Reference Offer in relation to any Network Access that

it is providing as at the date that this Condition comes into force;

MD5.4 The Dominant Provider shall update the Reference Offer in relation to any

amendments or in relation to any further Network Access provided after the date

on which this Condition comes into force.

MD5.5 Publication referred to above shall be effected by-

a. placing a copy of the Reference Offer on any relevant website operated or

controlled by the Dominant Provider; and

b. sending a copy of the Reference Offer to the Director.

MD5.6 The Dominant Provider shall send a copy of the current version of the

Reference Offer to any person at that person’s written request (or such parts

which have been requested).

MD5.7 The Dominant Provider shall make such modifications to the Reference

Offer as the Director may direct from time to time.

MD5.8 The Dominant Provider shall provide Network Access at the charges, terms

and conditions in the relevant Reference Offer and shall not depart therefrom

either directly or indirectly.

MD5.9 The Dominant Provider shall comply with any direction the Director may

make from time to time under this Condition.

168

Condition MD6 – Requirement to notify charges

MD6.1 Except in so far as the Director may otherwise consent in writing, the

Dominant Provider shall publish charges and act in the manner set out below.

MD6.2 The Dominant Provider shall send to the Director and to every person with

which it has entered into an Access Contract a written notice of any amendment to

the charges on which it provides Network Access or in relation to any charges for

new Network Access (an "Access Charge Change Notice") not less than 28 days

before any such amendment comes into effect.

MD6.3 The Dominant Provider shall ensure that an Access Charge Change Notice

includes:

a. description of the Network Access in question;

b. a reference to the location in the Dominant Provider’s current Reference Offer of

the terms and conditions associated with the provision of that Network Access;

c. the date on which or the period for which any amendments to charges will take

effect (the "effective date");

d. the relevant network tariff gradient.

MD6.4 The Dominant Provider shall not apply any new charge identified in an

Access Charge Change Notice before the effective date.

169

Annex B

Assessment of market power

B.1 Chapter 4 was devoted to a discussion of whether MNOs have SMP in therelevant market. It discussed the criteria that are most relevant to the assessmentof whether there are the conditions for a finding of dominance, given the specificcharacteristics of the relevant market. The Director concluded that each MNO hasSMP in the market for wholesale mobile voice call termination on its network(s).

B.2 The criteria discussed in Chapter 4 were:• market share;• countervailing buyer power;• excessive pricing and profitability; and• ease of market entry and absence of potential competition.

B.3 These are only a subset of the criteria listed in the EU Commission and OftelGuidelines on SMP. Hence, for completeness, this Annex will discuss the othercriteria and explain why they have been considered to be less relevant in thisspecific market.

B.4 This market power assessment analysis focuses only on single dominance.The Director considers that in the market for wholesale mobile voice calltermination, SMP cannot be held by more than one company, since the marketcurrently is a monopoly and no entry appears likely (see Chapter 3). As aconsequence, none of the criteria to assess collective dominance will be reviewedbelow.

Discussion of remaining criteria

Criterion Explanation Analysis/AssessmentOverall size of theundertaking

This criterion refers to theadvantages that may accrueto a firm because of its largesize compared to itscompetitors. Theseadvantages can range fromeconomies of scale inproduction and distribution tobetter finance conditions andmay arise from activitiesundertaken outside therelevant market.

This criterion matters onlyin that the size of an MNOmay give it countervailingbuyer power whennegotiating terminationcharges with another MNO.However, as discussed atlength in Chapter 4, theDirector is of the view thatin this market it unlikely thatin each of the bilateralrelationships between thefour MNOs, any oneoperator has the level ofbuyer power necessary to

170

off-set the other’s monopolyand ensure that calltermination charges are setat cost-reflective level.

Control ofinfrastructure noteasily duplicated

Control of an infrastructurethat is costly or time-consuming to build representsa significant barrier to entry

Access to a mobile networkand to the SIM card detailsof the called party areessential to be able toprovide call termination to aspecific mobile customer.However, even if a firm hadaccess to these facilities,there would still beconsiderable technicalhurdles that would not allowit to terminate the call.

Economies ofscale

Economies of scale existwhen the average cost of afirm falls as its outputincreases.

The Director does not viewthis criterion as relevant inthis market because thepresence of absolute (thisis discussed further inChapter 3) barriers to entryimplies that each MNOoffering voice terminationfaces no existing orpotential competitors and,therefore, cost-advantagesare not relevant.

Economies ofscope

Economies of scope arisewhen a multi-product firm canshare production anddistribution costs across anumber of services.

This criterion is not viewedas relevant because thepresence of absolutebarriers to entry implies thateach MNO offering voicetermination faces noexisting or potentialcompetitors.

Technologicaladvantages orsuperiority

Technological superiority canrepresent an advantage overcompetitors.

This criterion is not viewedas relevant in this marketbecause the presence ofabsolute barriers to entryimplies that each MNOoffering voice terminationfaces no existing orpotential competitors.Hence, no comparisonbetween technologies isrelevant.

171

Product/servicesdiversification

Bundling may supportdominance by foreclosingaccess to the market for partof the bundle. By bundling afirm can also leverage itsmarket power from a marketin which it is dominant to amarket where there ispotential for competition.

This criterion is not viewedas relevant because eachMNO sells termination tooriginating operators whorequest it on a stand-alonebasis and it is not bundlewith other services.

Verticalintegration

Vertical integration canpromote dominance by givinga firm the ability to leveragemarket power into theupstream or downstreammarket.

The MNOs are verticallyintegrated and the size oftheir monopoly in theprovision of mobile voicetermination depends on thenumber of retail customersthey have. However, noneof the MNOs has SMP (SeeOftel’s Review ofcompetition: mobile accessand call origination - April2003) in the retail marketthat it can leverage in themarket for mobile voicetermination. Hence, theDirector does not view thiscriterion as relevant.

A highlydevelopeddistribution andsales network

A well-developed distributionsystem and a strong salenetwork are costly, sometimeseven impossible, to reproduceand, as such, may representan entry barrier.

The Director does not viewthis criterion as relevantbecause the service hereinsold is acquired by only afew major purchasers (otherMNOs and fixed PECNs)and does not require aspecial distribution or salesystem.

Easy or privilegedaccess to capitalmarkets andfinancialresources

Some firms may be capableof commanding better creditconditions, thus gaining a costadvantage over theircompetitors.

This criterion is not viewedas relevant in this market,because the presence ofabsolute barriers to entryimplies that each MNOoffering voice terminationfaces no existing orpotential competitors.Therefore, the cost ofcapital an MNO facescannot give it any specialadvantage in this market.

172

Barriers toswitching

If a large proportion of a firm’scustomers switched toanother provider when it triedto raise prices abovecompetitive levels, the priceincrease would beunprofitable andunsustainable. Hence,switching can represent aconstraint on a firm’s ability tobehave independently of itsexisting customers.

Wholesale purchasers oftermination cannot switch toa different provider from theMNO to which the calledparties are subscribed,because each MNO is amonopolist in the provisionof call termination on itsnetwork(s), or in the case of3’s 2G services voice calltermination provided to itssubscribers.Switching by retailcustomers could act as aconstraint on the MNOs’pricing behaviour. If areasonable number of retailsubscribers were to changemobile network as aconsequence of a rise inthe price of incoming callsabove competitive levels(following an increase incall termination charges),MNOs would beconstrained in their pricingbehaviour.However, the Director couldnot find any evidence that amaterial number of mobilesubscribers change theirnetworks on the basis ofincoming call charges.Consumers appear toswitch mobile networks onthe basis of the cost of therental (if post-pay) and ofoutgoing calls (see Chapter3 for more details).Therefore, switching byretail consumers does notrepresent an effectivethreat that could constrainMNOs when settingtermination charges.

Customers’ abilityto access and use

Limited access to informationon terms and conditions

Wholesale customersappear to have all the

173

information (especially prices), or accessto information that is difficultto use, may reduce thecapacity of consumers to actupon differences betweenproviders. As a result firmsacquire independence ofaction from consumers andcompetition. This criteriondoes not refer to the ability ofconsumers to switch betweenproviders, but to the capacityof first time buyers to make aninformed choice.

relevant information tomake an informed choice.However, they have nochoice between providers,since each MNO is amonopolist in the provisionof voice termination to itssubscribers. Since thetermination charges feedinto the retail price forcalling mobile phones, andthe behaviour of retailconsumers may have animpact on the MNOs’ abilityto set high terminationcharges the informationavailable to retail customersmust also be considered.Better knowledge on theirpart about the price ofcalling each specificnetwork may indirectly forceMNOs to compete on thelevel of terminationcharges. However, asexplained in detail inChapter 3, a number oflinks (not just betterawareness of the cost ofcalling mobiles) would haveto be satisfied so that thebehaviour of callers tomobiles could constraintermination charges. Thus,the Director considers thatthe extent to whichconsumers can access toinformation on the cost ofincoming calls and howeasy this is to understandand used, does not affectthe current assessment ofmarket power. With regardto the called parties there isevidence that, as aconsequence of the pricingarrangements (CPP), they

174

are not concerned aboutthe cost of incoming callswhen choosing theirnetwork. This could changeif callers to mobile changedtheir behaviour andresponded to the differentprices for calling differentnetworks by makingdifferent volumes of calls.

Barriers toexpansion

This criterion refers to theability of competitors toexpand and increase theirmarket share.

This criterion is not viewedas relevant, because thepresence of absolutebarriers to entry implies thatcompetition in the market isnot likely to extend beyondthe existing players andthus the existence ofbarriers to expansionbecomes irrelevant.

Activecompetition onnon-price factors

Market power can be obtainedby successfully differentiatingproduct, either vertically (onthe basis of quality) orhorizontally (on the basis ofdiversity).

This criterion is not viewedas relevant in this marketbecause mobile voice calltermination does not seemto offer much scope forvertical or horizontalproduct differentiation. Inaddition, the presence ofabsolute barriers to entryimplies that competition inthe market is not likely toextend beyond the existingplayers and, thus,diversification, even ifpossible, would be notprovide any additionaladvantage.

International benchmarking

B.5 The 8th Implementation Report published by the European Commission at theend of 2002 provides a comparison of the peak mobile termination charges leviedby the major European MNOs. The chart, copied below in Figure B.1, shows thatUK charges for termination from fixed-to-mobile are amongst the highest in theEU. The average peak per minute rate in the EU is 18.49 Euro cents. UK peakcharges in comparison are:

175

Vodafone 20.96 (Euro cents)T-Mobile 24.90 (Euro cents)Orange 22.09 (Euro cents)O2 19.83 (Euro cents)

B.6 The lowest charges are in Sweden (9.87 Euro cents) and Austria (11.3 Eurocents). However, this comparison does not give a complete picture because UKMNOs have comparatively low weekend charges whereas in other EU countriesthere are no weekend tariffs.

Figure B.1: Fixed-to-Mobile termination charges at peak rate in August 2002(Euro cents per minute)

Source: 8th EU Implementation Report

Excessive pricing in EU

B.7 At least nine other NRAs have taken positive steps to control mobiletermination charges in 2002, although not all have been able to enforce theirdecisions by the end of 2002. Germany has been the only country not tointervene. Finland has a different form of CPP whereby the terminating MNOcharges the calling party directly and therefore does not have wholesale chargesfor mobile termination.

0

5

10

15

20

25

30

A-M

obilk

om

B-B

elgi

com

DK

- TD

C m

ob

E- A

irte

l

E -

Te l

efon

ica

EL

-Cos

mot

e

F- O

rang

e

F-SF

R

Fin

-Rad

io

Fin

-Son

era

I - O

mni

tel

I-T

im

Irl -

Vod

afon

e

I RL

-O2

(dig

if)

L- P

T&

T

NL

-KPN

P-T

MN

S -T

e lia

UK

-O

rang

e

UK

- O2

UK

-T

mob

ile

UK

-V

odaf

one

176

Annex C

Alternative solutions

This Annex discusses solutions, other than a RPI-X cap, that could put somepressure on the level of the mobile termination charges. The Director is of theview that none of the remedies herein discussed will, in the immediate future,create the necessary conditions to bring termination charges down to thecompetitive level. The primary reason for the Director taking this view is that, inorder for these remedies to be effective, they need to be accompanied by changesin the behaviour of the MNOs and/or the called and calling parties which areunlikely to happen in the short term. However, while these changes are unlikely tooccur overnight, it is the Director’s view that it is possible that they will occur overtime as consumers become more aware of relative charges and of methods forkeeping down the cost of making and receiving calls on mobiles.

All bar one (solution 8) of the remedies discussed below have also been examinedby the Competition Commission during its investigation into the prices for calls tomobiles.

The underlying problem

As discussed throughout this document (see Chapter 3), a key reason why theMNOs have a monopoly in the provision of termination services to theirsubscribers is the CPP arrangement adopted in the UK telephony market. Underthe CPP arrangement the calling party (and not the called party) pays the totalprice of retail call. This means that the call termination charge is included in theoriginating network provider’s (either fixed or mobile) cost base and is reflected inthe retail price it sets for calls originating on its network. CPP means there is nolink between the person paying for the calls and hence, for the termination charge(i.e. the calling party) and the person who makes the choice of the terminatingnetwork and could thereby influence the level of the termination charge (i.e. thecalled party). The overall effect of this arrangement is that: (a) in the retail marketthe MNOs have an incentive to keep the price of those services required and paidfor by the subscriber at a level to attract and retain customers, but they have lessincentive to keep the price of calls to mobiles low; and (b) in the wholesalemarket, the MNOs have little incentive to keep call termination charges lowbecause originating operators have to terminate the calls of their customers andhave to pay the termination charge.

Hence, the ideal solution would be to allow the calling party (i.e. who pays for thecall and thus for the termination charge) to choose the network on which toterminate the call. If this were possible, it would provide the MNOs with theincentive to compete on inbound call charges. However, this solution is currentlytechnically not feasible and it is unlikely that it will ever be.

177

Solution 1 – Receiving party pays

One technically feasible solution to the change the incentives that CPP generateswould be to switch to a system of RPP in which the recipient pays for the calls.This could be achieved in a number of ways:

(e) the fixed and/or the mobile operator could allow the calling party to requesta reverse charge by dialling a prefix;

(f) the fixed and/or the mobile operator could allow the calling party to call alocal rate number and initiate an automatic call-back at the behest of themobile subscriber;

(g) the mobile subscriber could be given the opportunity to agree to pay for acall at the start or at some point during the calls; and

(h) the mobile subscriber could be given the opportunity to establish a patternof incoming calls for which they are prepared to pay.

As mentioned above, technically this solution could be implemented in any of theforms just discussed, but the (fixed and mobile) operators are likely have to incursome costs to do so. However, since the MNOs have very little incentive to set upa system that could reduce the revenues they earn from termination services,some form of regulatory intervention would probably be necessary to impose theirintroduction.

The main problem with relying on these remedies to constrain termination chargesis that it is difficult to predict how called parties will react and whether they wouldbe willing to absorb the cost of the call (In the US where the RPP is thepredominant arrangement in the mobile market, the level of penetration hasremained lower and it is likely that this in part is a consequence of the RPParrangement). If the called parties resisted, limited use would be made of thesefacilities and thus no competitive pressure would be placed on mobile terminationcharges. In addition, it seems economically more efficient that the cost of aconsumption decision (i.e. the decision to make a call) should be borne by theperson who takes that decision (i.e. the calling party).

Conclusion on solution 1

The Director is of the view that, at present, this solution could not be relied upon toconstrain termination charges. It is unclear how the parties would behave and,thus, whether any significant pressure would be placed on mobile terminationcharges. In addition, the cost of implementation could be high and the possibleadverse effects of RPP on economic efficiency should also be taken into account.For these reasons the Director believes it would be inappropriate to require theoperators to introduce these arrangements. However, he would be interested inthe views of respondents to the consultation about whether further consideration

178

should be given to the development and possible imposition of forms of RPP in theUK.

Solution 2 - Call-back

Call-back occurs when the direction of a call is ‘reversed’ and the calling party iscalled back by the called party. During the CC investigation one of the MNOsproposed a remedy based on a call-back type of service. This remedy involvesthe MNOs introducing a service that allows the calling party to have a choice ofhow to make a call to a mobile customer, either:

(c) he calls directly (and pays a price which includes the termination charge setby the MNO), as it happens now, or

(d) he sends a text message or leaves a voicemail in the message box of themobile customer.

In either case, the message would inform the mobile customer that the caller hasmade a call. The mobile customer would then have the option of calling back, andwould bear the costs of making the mobile originated call. This service could beimplemented using existing technology and the relevant costs would not be high.

However, there are a number of potential problems with this remedy:

1) It is based on the assumption that an ‘indirect’ call (i.e. the messagingservice or the voicemail) is a good substitute for a ‘direct’ call involving thecall termination service. Hence callers would choose between the directand indirect services only on the basis of relative prices and service qualityand use the indirect service when calling networks with high terminationcharges.

However, for the indirect service to be a good substitute for the caller, themobile customer would need to be willing to call the caller back and to doso speedily. Consumer research and the relatively short duration of calls tomobiles on average suggest that immediacy of contact is an importantconsideration in the caller’s decision to call a mobile number. It is unclearthat mobile customers would behave in the required way. To the extent thatthey did not, the indirect call would not be a good substitute and theproposed remedy would be ineffective. In addition for the remedy to work,callers would need to know all relevant prices – i.e. direct, indirect and call-back prices for calls to each network (since the former two affect the caller’schoice and the latter the mobile customer’s). The evidence points toconsumers being relatively poorly informed about even a single set ofprices for calls to mobiles (see Chapter 3), so the proposed remedy wouldbe likely to be ineffective without consumers becoming much betterinformed.

2) The remedy is based on the additional assumption that the mobilecustomer’s service is made worse by high termination charges, as this

179

would lead callers to use the indirect service instead of directly calling themobile phone. Hence, the mobile customer would seek suppliers providingbetter service – by implication, a mobile network with low terminationcharges, so that callers would use the direct calling option.

However, it is not necessarily true that this remedy makes worse the mobilecustomer’s service, e.g. if the subscriber valued the ability to reply to callsas and when she wished. In this case, the subscriber would have noincentive to seek an alternative network and the proposed remedy would beineffective. In addition, even if the remedy did affect the mobile service, amobile customer would need to receive a better service on another networkto be induced to switch networks (i.e. she would need to receive more directand fewer indirect calls as a result of moving to a network with lowertermination charges). Consumer research and empirical evidence suggeststhat callers to mobiles base their decisions on the general level of pricesand not on the prices for calling particular networks. Mobile customerscould not, on the basis of the current level of price awareness of callers,rely on receiving more direct calls, even if they switched to a network withlower termination charges. The incentive to switch would, therefore,currently be weak. MNOs would recognise this and so would face weakincentives to offer lower termination charges. The proposed remedy wouldbe ineffective at present and simply reduce the overall quality of the servicereceived by mobile subscribers.

3) In addition, the incentives faced by the MNOs would render the serviceineffective, unless very detailed regulation on prices and quality of serviceswere introduced.

The argument is that mobile customers would seek the operator providingthe best inbound service if their service were affected by the introduction ofthis solution. If callers could choose between the direct and indirectservices, this would lead to mobile customers choosing an operator with lowtermination charges because they would receive a greater proportion ofdirect calls on that network. However, the MNO, rather than reducingtermination charges, would have the incentive to reduce the extent ofsubstitutability between the indirect and direct service by providing a lowquality indirect service. The mobile operator determines the ‘generalisedcost’ (i.e. the overall cost to consumers of using the service, which includesboth price and service quality effects – such as ease of use, speed ofcontact etc) to the caller of making the indirect call. It could thus raise thiscost to the level where calling customers did not use the service and wereeffectively forced to pay the excessive charge for the direct service instead.Hence, the proposed remedy would be likely to be ineffective. To ensurethat MNOs did not act in this way would require detailed regulation of thequality of all aspects of the call-back service. This would be a difficult taskand would lead to more detailed and intrusive regulation than a cap ontermination charges. This does not, of course, preclude the possibility thatthe MNOs could develop such a service voluntarily and provide a sufficient

180

quality of service to make it attractive to callers. If they did so, and if callersbecame better informed about the relative prices for calling differentnetworks, and responded to this information by making fewer direct calls tonetworks with high termination charges, the remedy could generate aneffective constraint on termination charges.

Conclusions on solution 2

The Director believes that this remedy is not currently likely to be effectivebecause it is unclear that the indirect call would in practice be a good substitute forthe direct call. The reasons for it being a poor substitute include the importance ofimmediacy of contact and the need for consumers to respond to an implausiblylarge amount of price information. In addition it would be likely to be ineffectivebecause it does not have the appropriate incentive properties to bring competitivepressure to bear on termination charges. MNOs would have sufficient control overkey elements of the indirect service and the incentive to ensure that it would notserve as a competitive constraint on the direct service offered by them.

For these reasons, the Director believes that this remedy would not currently be aviable option to solve the problem of the lack of competitive pressures on mobilecall termination charges. However, if services were developed which were ofsufficient quality that they became used by callers, and if callers become betterinformed about and responded to the relative prices for calls to different mobilenetworks, it is conceivable that this could evolve into a workable remedy over time.

Solution 3 – Call divert

Mobile subscribers could use a single terminal that is both a cordless phone and aGSM phone, thus making and receiving calls using their "fixed line" (via thecordless element of the phone) when within the range of the cordless base stationand bypassing their mobile network.

The introduction of a service based on dual handsets would require the co-operation of the MNO to divert incoming calls based on the called party’s locationand to modify its billing systems to be able to charge each call on the basis of howit was delivered. The MNOs have no incentive to contribute in setting up a systemthat reduces their profits from the provision of termination service, so some form ofregulatory intervention would be necessary to impose the introduction of such asystem.

The main obstacle to this being an effective remedy is that the called party wouldhave to acquire the necessary equipment. The cost of such dual phones iscurrently high and it would have to be offset by the benefits accruing to the calledparty. Most mobile customers, as discussed at length in Chapter 3, consider onlythe charges that face them when making their choice of network so the savingsfrom lower termination charges that would benefit the calling party would beunlikely to affect their choice of network. Hence, at present, such a system isunlikely to become widespread, as mobile customers would not attach much value

181

to the facility, while they would be deterred by its costs. For it to work it isnecessary for callers to become more aware of the relative prices of callingdifferent mobile networks and to respond by making fewer calls to networks withhigh inbound call prices. If the called parties then considered this when makingtheir choice of network, some pressure on termination would be brought about.While this kind of behavioural change is unlikely to happen overnight, it may takeplace over time.

Conclusions on solution 3

In conclusion, the Director considers that this remedy is currently unlikely tointroduce any significant competitive pressure on the level of the terminationcharges. It relies on mobile subscribers taking up a service which would causethem to incur costs but would not, on the basis of the current behaviour of callers,bring them sufficient benefits to justify those costs. However, if callers becomebetter informed about relative prices of calls to different networks and respondedto these prices, it is possible that mobile subscribers would find it worthwhile toincur the extra cost and this could then develop into a workable remedy.

Solution 4 - Mobile virtual network operators with multiple roamingagreements

A mobile virtual network operator (MVNO) is a firm that provides mobile telephonyservices to its customers, but which does not have an allocation of spectrum anduses part of a MNO’s network (the ‘host MNO’). MVNOs come in a number offorms (see Oftel's consultation document on MVNO's, available athttp://www.oftel.gov.uk/publications/1999/competition/mvno0699.htm). A MVNOcould be a simple service provider or it may be able to authenticate its subscriberson one or more mobile networks and route inbound calls to them. The latter typeis referred to as a “full MVNO”.

A full MVNO provides termination services to its subscribers, hence it could havetermination agreements with more than one MNO and select the mobile networkon which to terminate each call on the basis of its charges. This would put somepressure on MNOs to offer them lower termination charges. However, the processof changing the network onto which a mobile is connected currently takes asignificant amount of time (possibly around 15 - 20 seconds). The MVNO has tosignal to its subscriber’s handset (or SIM card) to change network, and thenreceive an acknowledgement back to say this was done successfully, before it canconnect the call. In addition, if the MVNO tries to switch the mobile onto a networkthat has no coverage in the area where the called party is or which is busy due tonetwork traffic. This would introduce even more delay, as the mobile would needto go back to the original network, or try another network. Hence, at presentmultiple roaming appears unattractive. In addition, the MVNO would need to signmultiple roaming arrangements and no MNOs would have the incentive to do this.A regulatory intervention would be necessary to impose such a requirement onMNOs.

182

The main obstacle to this remedy is, however, that, even if the MVNOs were ableto negotiate lower termination charges, they would be unlikely to have theincentive to pass the savings on to callers to mobiles. They would have the sameweak incentive to reduce prices for calls to mobiles as the MNOs. This isbecause, as discussed in Chapter 3, most mobile subscribers do not select theirmobile service provider (MNO or MVNO) on the basis of the prices paid by othersto call their mobiles. Therefore, it seems unlikely that the MVNOs, even if theymanaged to negotiate lower mobile termination charges, would reflect them intolower retail prices to their subscribers rather than in lower charges for terminatinginbound calls.

Conclusions on solution 4

In conclusion, the Director considers it unlikely that an MVNO would introduce anysignificant competitive pressure on the level of the termination charges, at leastwhile the behaviour of calling and called parties in response to price changesremains as it currently is. For MVNOs to have an effect, mobile customers need tochoose their network on the basis of inbound call charges, which they currently donot. If they did so, there would be no need for any remedy, since such consumerbehaviour would lead to competition on termination charges among the existingMNOs.

Solution 5- Multiple SIM cards

As discussed in Chapter 3, because of the CPP arrangement most subscribers donot choose their network on the basis of the price of inbound calls, hence theymay select a network with high termination charges. However, if they couldreceive their incoming calls on mobile networks other than the one to which theysubscribe for making outbound calls, they could choose to receive calls on anetwork with lower termination charges. This could place some pressure on thelevel of termination charges.

For this to happen, the called party must be able to switch his handset betweendifferent networks. It is possible to envisage a mechanism that instructs the calledparty’s mobile phone to switch network automatically when a call is arriving.However, no such mechanism currently exists because of significant technologicaldifficulties. In addition, in order to implement this solution it would almost certainlyrequire a regulatory intervention as no MNO would have the incentive to agree to itgiven its likely negative effect on its revenues from termination services.

Moreover, even if the technical hurdle were overcome, to be effective this solutionwould rely on the called party being responsive to the price of inbound calls sothey would be prepared to incur some cost to reduce the cost to the person callingtheir mobile (as the called party needs to acquire a multiple SIM card holder andsubscribe to more the one network). As discussed extensively in Chapter 3, thisincentive is currently absent.

Conclusions on solution 5

183

The Director considers it unlikely that a remedy based on multiple SIM cardswould be effective given the current behaviour of calling and called parties andbecause current technology does not allow its introduction. However, if callingparties become better informed about the prices of calls to different networks, andresponded to these prices, it is possible that mobile subscribers would adoptmultiple SIM handsets and be induced to switch to low cost networks for inboundcalls. This could then become a workable remedy.

Solution 6 - GSM Gateways

A GSM gateway is a way of converting off-net mobile calls onto on-net calls. Itworks by setting up a mobile call on the network where the call is to be terminated,and then "connecting" this call to the original call. By using a GSM gateway, thenetwork originating the call pays the on-net retail rate to the MNO, rather than theusual wholesale call termination rate.

To illustrate how this solution operates one could consider the case of a fixed userwho calls a mobile subscriber. If the relevant fixed originating operator has GSMgateways, it would take the call to one of them, which would set up the call usingthe relevant MNO’s SIM card. The originating fixed operator can thus connect thefixed call to the MNO call and pay the MNO a termination rate based not on thewholesale call termination rate, but on the retail on-net rate.

GSM gateways exist and are used by operators to provide corporate services.However, it is illegal under current spectrum regulation to use such devices forcommercial services. The Radiocommunications Agency is consulting on possiblechanges to this regulation (seewww.radio.gov.uk/topics/pmc/consult/publicwireless/pubwire.htm).

However, even if GSM gateways were authorised, their introduction would notaddress the problem of a lack of competitive pressure on termination charges.GSM gateways simply exploit an arbitrage opportunity generated by the currentlow price for on-net calls. Their appeal depends only on retail on-net prices beinglower than wholesale call termination charges (and volumes being sufficient tojustify the cost of buying and maintaining the gateways). Hence, MNOs couldeasily undermine their viability by changing the structure of their prices. In addition,GSM gateways are inefficient because they use extra network facilities to convertthe call to an on-net call and, thus, employ more resources than a standard call toa mobile. Their use can also reduce call quality and prevents the correct CallingLine Identity (CLI) being passed on, which impacts on the called parties’ ability toscreen calls and to prevent nuisance calls.

Conclusions on solution 6

The Director considers that GSM gateways on their own, even if their use were tobe legalised, would not introduce significant competitive pressure on the level ofmobile termination charges. In his view, they would be more likely to dissuade

184

MNOs from offering low on-net call prices, as these would have the effect ofproviding a by pass opportunity for their termination monopolies. In addition, GSMgateways are an inefficient means for delivering a call to a mobile because theyemploy more resources than a standard call to a mobile and reduce the quality ofsuch calls. However, if the use of GSM gateways is legalised in future, theDirector will monitor their effect on call termination charges to establish whetherthey might have a role to play in addressing the problem of lack of pressure onmobile termination charges.

Solution 7 - Tying termination charges to the prices for competitivelysupplied services

A solution of this type has been implemented by the Australian Competition andConsumer Commission (ACCC) which had been investigating termination chargesfor calls to mobiles. In its draft report entitled Pricing Methodology for the GSMTermination Service, [December 2000], the ACCC recognised that there isinsufficient competitive pressure on call termination charges to constrain them. Inparticular, it placed great importance on the calling party pays principle and therole of consumer limited knowledge of the prices of calls to mobiles.

After considering different approaches to price control, the ACCC implemented anapproach which combines benchmarking with a tie of call termination charges toother charges which it believed were competitively determined. Briefly, theirfavoured approach involves setting all mobile termination charges equal to thelowest currently existing termination charges. Thereafter, changes in each MNO’stermination charges would be benchmarked against the percentage change of itsweighted average retail prices for the overall mobile packages (access andoutgoing calls).

The advantage of this type of option would be that the benefits from competition inrespect of other parts of the ‘mobile package’ would be passed on to consumers ofmobile call termination.

Conclusion on solution 7

In the Director’s view, this option is not attractive for a number of reasons. First,this approach could distort the competitive markets to which the non-competitiveservice is tied. Secondly, there would be a resultant expansion of regulation intoareas not previously covered. For example, by tying call termination charges tocall origination charges, Oftel would effectively be regulating both. Thirdly, therecould be a departure from the principle of cost reflective charging, since the levelof termination charges would be set not according to the costs of that service, butaccording to the price of some other service, possibly with a different coststructure. Fourthly, it would not address the distorted price structure, which wouldrequire larger decreases in termination charges than in retail prices. In addition,associated with the Australian proposed solution in particular, there are potentialpractical problems in continuing an approach to measurement of the average

185

percentage change of a large set of different and complex tariffs. Any measureused could be inaccurate and/or open to strategic manipulation by the MNO

Solution 8 - Delivering a call further into the terminating network

1 If a call to a mobile phone were not passed to a specific GSM gateway, but to apoint lower down in the mobile network, some components of the networkcould by-passed thus reducing the cost of terminating the call. A call could, forexample be passed directly to the base station controller ("BSC") nearest to thecalled party (This remedy is also discussed in Appendix 3.1 page 116 of theCC report).

2 A call to a mobile subscriber would normally be passed on the network of theMNO to which the called party subscribes at a particular switch (nominated asa gateway switch). Once on the MNO’s network the call would then be passedto the serving mobile switch and from there to the serving BSC. If originatingoperator could directly route the call to that serving BSC the use of two mobileswitches would be avoided. Figure B1 below shows a call that follows the“normal path” (route 1) and a call that is passed directly to the nearest BSC(route 2).

3 Figure B1: solution 8

GMSC

BSC

MSC

Fixedoperator's

1

2

186

4 In order to be able to pass the call to the correct BSC (i.e. the serving BSC) theoriginating operator needs to identify the location of the serving BSC. Thisrequires the exchange of signalling information between the MNO’s homelocation register and visitor location register (HLR/VLR - The home locationregister and the visitor location register are databases used by GSM mobilenetworks to track the location of mobile subscribers, as well as to verify theservices they are entitled to use) and the originating operator’s equivalentdatabase and switch. To Oftel’s knowledge, the exchange of this information isnot standard, hence some form of regulatory intervention would be necessaryto make this possible.

5 In addition to the information on the location of the appropriate BSC, theoriginating operator would require a network of leased lines that linked itsnetwork to all BSCs (around 100 for each mobile network). The cost ofdeploying this solution is likely to be high as some BSCs are in remotelocations. Hence, the amount of traffic to be delivered this way and the costthus saved (in terms of mobile network elements thus bypassed) would have tobe relatively large to justify incurring the cost of setting up this network ofleased-lines. However, the amount of costs that would be saved would besmall. The MNOs would still need to maintain all the network elements from,and including, the BSC down to the individual radio masts and aerials (as wellas other intelligent network functions, such as those provided by theHLR/VLR). The costs of these elements account for the majority of terminationcosts. It is also not clear to Oftel to what extent GSM standards allow BSCs tobe directly connected to public switched telephone networks.

From the above discussion, it is clear that no operator would want to resort to sucha remedy to reduce its mobile termination costs because the costs to implementthe solution are very high. However, the main limitation of this solution is that itdoes nothing to address the problem at hand. The problem is not the costs ofmobile termination, but the charges levied for the service, which are considerablyabove cost. The proposed remedy would not increase the amount of competitivepressure on termination charges and hence would not address the problem.

Conclusions on solution 8

The Director is of the view that this potential remedy is not worthy of furtherconsideration, since it would not increase pressure on termination charges.

Summary of the Director’s views on these alternative solutions

There are essentially four types of remedy to the problem of insufficientcompetitive pressure on mobile termination charges. They are:

(1) a departure from the CPP principle, which is the underlying cause ofthe problem;

187

(2) the introduction of substitute services for the calling party so he canbypass the termination service of the mobile customer’s MNO, which mightinvolve some action by the mobile customer to enable the substitute serviceto be used

(3) the introduction of arrangements by which mobile customers takeactions to reduce the cost of calls to them, and

(4) “borrowing” competitive pressure by linking termination charges tothe charges for more competitive services.

The Director’s views of these groups of options are as follows:

(1) In the Director’s view a move to RPP (solution 1) would be disruptiveand uncertain in its overall economic effect. While such a move should notbe ruled out forever, the Director does not believe it would be appropriate tointroduce RPP at this time.

(2) Currently there are no effective substitute services available forcallers to mobiles (see Chapter 3 on market definition). However, withsuitable action by mobile customers, it is possible for callers to be offeredother options for calling mobile customers (solution 2). This means that for(2) to be effective, the key is what happens under category (3).

(3) There are a number of ways in which mobile customers could makeit easier for callers to call them at a lower rate (solutions 3, 4 and 5 all comeinto this category). For any of these remedies to be successful, a change inconsumer behaviour by both callers to mobiles and mobile subscribers isrequired. In particular, callers to mobiles need to become better informedabout the relative prices for calling different mobile networks, and torespond to this information by making more calls to mobile subscribers onnetworks with low termination charges and fewer calls to networks with hightermination charges. This would induce mobile customers to become moreresponsive to the price of inbound calls so they would be prepared to incursome cost to enable callers to call them more cheaply. It is interesting tonote that, if such a change of behaviour took place, the MNOs themselveswould have an incentive to compete on inbound call charges. Thus, inorder for remedies in category (3) to be effective it is necessary for there tobe a change in consumer behaviour along the line of the type of changethat would of itself remove the problem and the need for a remedy.

Accordingly, whether any category (3) remedy would be technically ortheoretically feasible is not the key question. More important is whether itwould play any part in changing consumers’ information set and behaviour,so that callers were more aware of the prices for calls to specific networksand mobile customers were more responsive to the prices of inbound callsin choosing their mobile network (because of the differing volume ofinbound calls they would receive on different networks). In the Director’s

188

view, none of the proposed remedies would by itself quickly changeconsumers’ behaviour in this way, but each might have a part to play infocussing more attention on inbound call charges and thus contributing to agradual change in behaviour.

A point of particular relevance is that mobile customers currently have theopportunity to exercise choice in a way which might constrain terminationcharges only when they make a choice of network. Such choice, while notirreversible, is exercised infrequently by most customers. Some of theproposed remedies in category (3) would enable mobile customers to takeaction to put pressure on termination charges without switching theirnetwork operator. To that extent the Director believes these remediescould have a part to play in changing consumer behaviour. Thus, while heis unpersuaded of the adequacy of these proposed remedies in the shortterm, he would encourage the fixed and mobile operators to develop theseideas with a view to seeing whether they might bring about changes inconsumers’ behaviour over time.

(4) “Borrowing” competitive pressure. The potential remedies described here inremedies 6 and 7 come into this category. The problem with these remedies isthat they also dilute competitive pressure in the services to which call terminationis linked. Since there is, in a sense, a finite amount of competitive pressure, thelinkage of a competitive service to a monopoly service increases the competitivepressure on the monopoly service at the expense of reducing the competitivepressure on the previously competitive services.

189

Annex D

Cost of Capital

There are a variety of methods for estimating a firm’s cost of capital. It is usuallycalculated as a weighted average of the cost of debt and equity finance.

The cost of capital can be expressed in real terms (after adjusting for inflation) ornominal terms. It can also be expressed in post or pre-tax terms. A pre-tax cost ofcapital should be compared with returns calculated on a pre-tax basis and a post-tax cost of capital with post-tax returns. Oftel uses a pre-tax real cost of capital asa basis for setting charge controls. The following sections outline Oftel’s approachand the values of the key variables in the calculation of mobile operators’ cost ofcapital.

Estimating the Cost of Capital: the Capital Asset Pricing Model (CAPM)

A number of different asset pricing models exist for determining the cost of capital.In addition to the CAPM which measures market risk, with a beta measuredrelative to a market portfolio, there are, for example, multifactor models whichmeasure market risk using multiple betas estimated relative to different factors.The CAPM has a clear theoretical foundation and is simple to implement incomparison to other asset pricing models. This results in the continued wide use ofthe CAPM by almost all regulators of utilities companies and as the usualprocedure in large investment banks and securities houses.

In forming its views about the cost of equity, Oftel uses a number ofmethodologies, but the main emphasis is on the use of the CAPM. Under theCAPM methodology, the cost of equity is built up from three main factors. Theseare:

• the risk free rate;• the equity risk premium; and• the value of beta for the company in question.

The relationship between these factors can be summarised by the followingformula:

Cost of equity = RFR + ( ERP x beta),

where RFR = the risk free rate, ERP = the equity risk premium.

The risk free rate is simply the expected rate of return on a risk free investment.The equity risk premium is the expected return on equities over and above the riskfree rate (that is, it is the expected reward for holding equities compared with thereward for holding risk free assets). The value of beta reflects the variability of

190

returns of the equity of the company in question compared with the variability ofreturns on the equity market represented by an index.

Similarly, the cost of debt can be expressed as:

Cost of debt = RFR + Debt premium,

where the debt premium is the company specific risk premium for corporate debtabove the risk free rate.

The weighted average cost of capital (WACC) takes account of the cost of equityand the cost of debt by weighting each of these by the proportion of equity anddebt respectively in the MNOs’ financial structures in the following way:

WACC = (Cost of equity x (1 – Gearing)) Cost of debt x Gearing,

where Gearing = Debt / (Debt + Equity).

The following sections discuss each of these major components in turn.

Risk Free Rate

The risk free rate of interest is reflected in the calculation of both the cost of debtand the cost of equity. The nominal risk free rate is usually calculated as the yieldon fixed term government debt of certain maturity. There is a range of maturitieson government debt that could be used as the basis for an estimate of the risk freerate. These maturities range from less than 1 year to over 30 years. There arearguments in favour of both short and long-term gilts as the best estimate of therisk free rate. For the purposes of the calls to mobile market review, a maturity ofthree years may be appropriate, as the review is concerned with charge controls tobe applied over a three-year period. However, it is relevant to consider that mobileoperators are required to make longer term investments, for example regardingnetwork infrastructure, hence a long-term gilt may better represent the risk freealternative to the investment made by mobile operators. Oftel has previously takena position between these extremes, and used a period of between 4 and 5 years.Oftel considers the use of 5-year gilts as reasonable, and this period is consistentwith the gilt term used in previous Oftel estimates of the cost of capital.

Oftel uses current estimates of yields on nominal gilts as a proxy for the risk freerate. The objective is to obtain a forward-looking estimate of the risk free rate. Thenominal risk free rate for 5-year gilts in March 2003 ranged from 3.7% to 4.3%9

with an average of 4.0%. This rate compares with a real rate of return of 1.5% forsimilar term index-linked gilts. This difference between the real and nominal rateimplies an inflation rate of approximately 2.5%. The implied inflation rate iscalculated on a geometric basis: (1+nominal rate)/(1+real rate) –1.

9 Source: Bank of England, http://www.bankofengland.co.uk/statistics/yieldcurve/

191

In paragraphs 7.204 to 7.209 of their report, the CC considered real returns onindex-linked gilts over a period from 1985 to 2002 of about 3.5%, and the long-runreal return on Government securities from 1900 to 2000 of about 1.3%, in additionto the then current real rates on index-linked gilts of about 2.2% in forming theirview on a range for the risk free rate of 5.1%-5.3% in nominal terms. The analysisof longer-term evidence in part addresses concerns that interest rates calculatedfrom government securities currently provide too low a benchmark for a risk freeinvestment due to factors such as strong demand from pension funds. Oftel notesthat it is difficult at any point in time to determine whether the current level ofnominal or real interest is too high or too low. In any case, Oftel considers that, ifcurrent rates are artificially low, mobile operators are able to take advantage of thisand borrow at current rates to finance future investment.

In addition to Oftel’s usual method of determining the risk free rate from currentreturns on gilts, Oftel finds some merit in the approach taken by the CC todetermine the risk free rate taking account of redemption yields over a longerperiod of time as well as the current spot rates, in recognition that the spot rate issubject to a certain degree of fluctuation. In determining an updated estimateequivalent to the CC’s range of 5.1%-5.3%, Oftel believes that an estimate of 5%is reasonable, taking account of the fall in current spot rates and reflecting thecontinuation of the downward trend in the underlying rate. Oftel has used this asan upper estimate and 4%, as indicated by the current nominal return on 5-yeargilts, as its lower estimate to obtain a final range of 4%-5% for the nominal riskfree rate. This compares with a rate of 5.2% used in September 2001, the lowerrate reflecting the decline in the risk free rate over the intervening period.

In September 2001, Oftel stated that the inflation risk premium was notsignificantly different from zero. Given the persistence of low inflation rates andlow interest rates and the assumption that conditions have not changeddramatically, Oftel has not made an adjustment for an inflation risk premium in thecalculation of risk free rates.

Equity Risk Premium

The equity risk premium is the difference between the overall return on equitiesand the nominal risk free rate. Its value in the UK reflects the risk of investing inUK equities generally. There is considerable debate about the appropriate methodof calculating the value of the equity risk premium and the calculation isproblematic because different methods produce different values. In particular,methods based on an analysis of current market expectations tend to give lowervalues than those based on analysis of historical estimates from stock marketdata. But determining current market expectation is a difficult and controversialtask.

In its report, the CC refers to two high-profile published studies from last year,namely The Equity Premium, Fama and French, Journal of Finance, April 2002,and Triumph of the Optimists, Dimson, Marsh and Staunton, Princeton University

192

Press, 2002. These publications provide a range of estimates using both historicalaverages of equity returns and evidence of investors’ expectations.

The other UK regulators have all adopted a range of measures of the ERP, forexample:

• OFGEM, in its September 2001 Review of Transco’s Price Control from 2002,http://www.ofgem.gov.uk/temp/ofgem/cache/cmsattach/315_26sep01_pub1.pdf, suggests an ERP of 3.5%, based primarily on survey forecast evidence;

• OFWAT, in Final Determinations: Future water and sewerage charges 2000-0525 November 1999,http://www.ofwat.gov.uk/aptrix/ofwat/publish.nsf/AttachmentsbyTitle/finaldets99part11.pdf/$File/finaldets99part11.pdf, assumes an equity risk premium of3.0%–4.0%. Again, this estimate was based primarily on survey forecastevidence; and

• The CAA, in Economic Regulation of BAA London Airports (Heathrow, Gatwickand Stansted) 2003 – 2008, February 2003, decided to use the CC’s mostrecent range of 2.5% to 4.5%.

Estimating the equity risk premium based on historical data typically leads tohigher values. For example, historical estimates from the London Business Schoolfor figures as at the end of 2000 showed the arithmetic estimate of the real equityrisk premium measured over 101 years in the UK to be 5.6% for bonds and 6.5%for bills10. However, a significant problem with relying on historical estimates is thatthey can vary markedly depending on the period used, as shown by the followingtable.

Table 1: UK mean equity risk premiums over various periods

Period Relative to gilts Relative to treasury bills1900 to 2000 4.4 4.81900 to 1949 2.1 2.41950 to 2000 6.8 7.11960 to 2000 4.6 5.31970 to 2000 3.5 5.61980 to 2000 3.6 7.31990 to 2000 0.4 4.8

Source: Triumph of Optimists, Dimson, Marsh and Staunton, Princeton University Press, 2002

The Monopolies and Merger Commission (now the CC), in its 1998 report on thecost to call a mobile phone estimated the equity risk premium to be in the range of3.5%-5.0%, with a mid-point of 4.25%. In subsequent publications, this estimatehas been revised downwards, due in part to downward trends in historical data. Inits report on calls to mobile in December 2002, the CC estimated a nominal range 10 Triumph of Optimists, Dimson, Marsh and Staunton, Princeton University Press, 2002, Table 32-1, p301.

193

of 2.6%-4.6%. However, in Paragraph 7.265, they noted that the extent ofuncertainty concerning the downward trend in recent years made a degree ofcaution appropriate when implementing this decline, in part to help preventvolatility in the short term. They felt that this factor was most appropriately takenaccount of not by modifying their judgment of the range for the equity risk premiumbut by increasing the overall level of the WACC by 0.25% in real terms.

During the CC’s investigation, all the main parties, other than T-Mobile estimatedan equity risk premium of 5%. T-Mobile, through its consultants Charles RiverAssociates, used an estimate of 5.6% to 6.1%11.

In deciding the appropriate value for the equity risk premium, Oftel has taken intoaccount a range of evidence, both historical and forward-looking. Oftel’sjudgement reflects its recognition of the need to balance both short and long-terminterests of consumers. A low rate of return on capital can bring benefits toconsumers in the short term in the form of lower prices. However, it could damageconsumers’ longer-term interests. The telecommunications industry depends onhigh levels of discretionary investment to support innovation and rapid marketgrowth. The funds for such investment are often internationally mobile. Too low afigure for the cost of capital could deter such investment, thus disadvantagingconsumers in the longer term.

Oftel’s previous view, as stated in its effective competition review of mobile ofSeptember 2001 and supported by expert advice from Professor Julian Franks atthe London Business School (LBS), was that 5% is an appropriate value for theERP. He has considered, in the light of the new academic evidence available, andof the considerations outlined in the previous paragraph, together with theincreased uncertainty affecting world markets, how this view should be revised.Given the recent levels of stock market uncertainty, and the investment imperativeoutlined above, Oftel considers a value of 5% for the equity risk premium to beappropriate for this consultation. In the coming months, Oftel will review the latestevidence available to determine whether this view is still appropriate.

Equity Beta

The value of a mobile operator’s equity beta measures the movements in returnfrom the mobile operator’s shares relative to the movement in the return from theequity market as a whole. It will rise with a mobile operator’s debt equity ratio(gearing), since a higher level of gearing implies higher volatility in the returns toshareholders.

The approach taken by Oftel in its previous market review was to use an index oftelecommunication firms to provide a measure of the equity beta for UK mobileoperators. This approach is based on the assumption that the average equity betaof telecom firms listed on the London Stock Exchange is the best estimate of theequity beta for the UK mobile operators. The latest London Business School Risk 11 Paragraph 7.211 of the CC’s report

194

Measurement Survey (RMS) estimate of the equity beta for wirelesstelecommunication service companies is 1.05 (RMS March 2003).

In its September 2001 statement, Oftel adopted a broad range of potential betaestimates at a 10% gearing on a debt to debt plus equity basis, from 0.94 (the thenmost recent available RMS estimate for Vodafone) to 1.50, which was significantlyin excess of RMS telecom index measures of beta. This estimate was based onindependent source of beta estimates consistent with past Oftel beta estimates,but acknowledged that there is a risk these estimates understate the beta of UKmobile operators. These equity betas were based on a 10% gearing and thengeared up as appropriate for the WACC. The validity of these estimates wasverified by an independent study carried out on Oftel’s behalf by The Brattle Groupin 200212, which produced a wide range of estimates that were consistent withOftel’s chosen range.

Beta estimation is a difficult exercise. One contentious issue in previous reviewshas been the choice of data frequency (daily / weekly / monthly) and estimationperiod (how many years’ worth of data to use, and which period to choose). Thesignificant impact that these factors can have is illustrated in the table below,which provides beta estimates for Vodafone against UK indices calculated onOftel’s behalf by The Brattle Group.

Table 2: Daily and monthly beta estimates for Vodafone

Data Frequency Period Estimate (midpoint)Daily 1996-00 1.64Daily 2001-02 1.36

Monthly 1996-00 0.99Monthly 1997-02 0.91

Source: The Brattle Group

Beta estimation is further complicated by, inter alia, the following issues:

• isolating relevant activities:− excluding overseas activities;− excluding non voice termination domestic activities

• choice of index (international/domestic) to measure risk relative to; and• use of Bayesian adjustment.

During the CC’s recent investigation, the beta estimates shown in the table belowwere submitted by the MNOs and Oftel for a gearing of 10%.

Table 3: Estimates of equity beta of MNO UK operations

12 Issues in Beta Estimation for UK Mobile Operators, The Brattle Group, July 2002.

195

Low beta High beta AverageOftel 0.94 1.50 1.22The Brattle Group (on behalf of Oftel) 1.16 1.56 1.36Vodafone 1.30 1.70 1.50O2 1.22 1.44 1.33Orange 1.45 1.55 1.50CRA (on behalf of T-Mobile) 2.13 2.51 2.32Source: The CC’s report on Vodafone, O2, Orange and T-Mobile, Table 7.17

The CC concluded that a range of 1.0 to 1.6 takes into account all of theuncertainties brought about by taking daily as opposed to monthly returns,choosing the appropriate time period, any differences between UK and overseasactivities and any differences between the regulated and non-regulated operationsof the MNO.

The Director is minded to use the CC’s range of 1.0 to 1.6 at a 10% gearing, ratherthan his previous range of 0.94 to 1.5. He agrees with the CC’s recognition that abalance should be sought to take account of the range of factors outlined above,and considers that the CC’s slightly higher range is appropriate given the recentmarginal increase in RMS betas for mobile companies.

Debt Premium

The cost of corporate debt is made up of a risk free component and a companyspecific risk premium. Historical evidence suggests that blue chip corporate debt,such as that of mobile operators, commands a small risk premium, althoughestimates of this premium vary considerably.

In its September 2001 mobile market review, Oftel estimated the debt premium tobe 1.0% at a 10% gearing, and 1.75% at a gearing of 30%; the non-linear increasein the debt premium as gearing increases being consistent with the availableempirical evidence.

During the CC’s investigation, interested parties submitted a wide range ofestimates for the debt premium. These estimates ranged from the CompetitiveOperators Group’s (COG) estimated average medium term debt premium of0.81%, based on recent debt premiums for water and electricity companies, toOrange’s average estimate of its own debt premium of 5.17%. The CC used arange of 1%-4% at a 10% gearing in its December 2002 report, based on Oftel’slow estimate and the average of the higher premiums paid by MNOs in 2002.

It has been suggested that the debt premium used should be set equal to thecontractual rate on debt currently offered by the MNOs. However, for some of theMNOs, the promised yield is an inaccurate proxy for the debt premium since itdiffers substantially from the expected debt premium. Where the probability ofdefault is significant, ie where the promised yield rates are substantially higherthan the gilt rate, the expected rate on debt (and therefore the cost of debt to the

196

MNO) is lower than the promised yield. The promised yield is effectively themaximum possible return on the bond – it would only be realized if the bondreached maturity. In other cases, the actual return would be lower (for example,substantially below the gilt rate). The size of the premium of the MNOs’ promisedyield on debt over gilts rates submitted to the CC indicates that the probability ofdefault was significant. In calculating the WACC, it is correct to use the expectedcost of debt, which means that the promised yield rate must be adjusteddownwards in order to approximate the expected cost of debt more closely.

Considering the higher premiums paid in 2002 by MNOs averaging 4%, Oftel ismindful to increase its previous estimate of 1.0% to 1.75% to encompass a higherrange of values. After taking account of default probability, Oftel considers a rangeof 1.0% to 3.5% to be reasonable. For the purposes of calculation, Oftel has usedthis wide range of debt premium both at a 10% and 30% gearing, whilstrecognising that the debt premium is more likely to lie at the lower end at a 10%gearing and to be higher at a 30% gearing.

Oftel’s estimate of the mobile operators’ cost of capital is based on beta of debt ofzero for the first one percent of the debt premium and increasing by 0.2 for everyone percent of debt premium above one percent. The debt beta measures theriskiness of the returns on debt. Oftel’s estimate of the debt beta implies that thefirst one percent of premium on mobile operators’ debt is due to liquidity risk ratherthan default risk. Any increase in debt premium beyond that level is attributed tothe risk of default.Thus, on the basis of the current information available, Oftel considers a debtpremium ranging from 1.00% to 3.50% for a gearing of 10% to 30% to bereasonable.

Optimal gearing

Under the standard Capital Asset Pricing Model a firm can potentially lower itsoverall cost of capital by increasing its gearing. This is because debt is generallycheaper than equity as a result of tax advantages to debt.

In its September 2001 review, Oftel used a range of gearing from 10% to 30% ona debt to debt and equity basis. This compares to levels of 20% to 40% previouslyconsidered for BT. In their December 2002 report, the CC adopted an estimate of10% on the basis of Vodafone’s and O2’s actual financial position.

A report submitted to the CC by the COG13 highlights a number of factors whichmay drive up gearing over the coming years including the ability to take on greaterdebt as profitability and stability improve potentially enabling a lower ocerall cost ofcapital.

13 European Mobile Operators Mobile Valuation, Enders Analysis, September 2002

197

Oftel believes that it is still appropriate to consider the wider range of 10% to 30%for the optimal gearing of UK mobile operators to allow a potential for differentcapital cost structures.

Effective Corporate Tax Rate

The calculations set out above are on a post tax basis. Oftel’s financial modelscalculate pre tax returns, so it is necessary to convert the post tax cost of capitalinto an equivalent pre tax figure. This is achieved by dividing the post tax figure bya factor of (l-tc), where tc is the effective corporate tax rate. Oftel has assumed, asdid the CC in their recent report, that the mobile operator’s effective tax rate is thecurrent standard rate of corporate tax of 30%.

Conclusion

Overall, using a broad range of parameters, Oftel estimates the pre-tax nominalcost of capital for UK mobile operators providing 2G services to be between 11.9%and 17.6%. This compares to Oftel’s previous estimate of between 13% and 17%as stated in September 200114. The broad range reflects the uncertaintysurrounding estimation of the key parameters, in particular, betas for the mobileoperators.

As derived in the table below, in pre-tax real terms, Oftel’s estimate for the cost ofcapital ranges from 9.2% to 14.7% with a mid-point of 12%. This compares withOftel’s previous mid-point estimate of 12.5% in September 2001 and the CC’sfigure of 11.25% in its December 2002 report.

Table 4: Estimtes of pre-tax real WACC

Low Gearing High GearingLow

estimateHigh

estimateLow

estimateHigh

estimateRisk free rate 4.0 5.0 4.0 5.0Equity risk premium 5.0 5.0 5.0 5.0Equity beta 1.0 1.6 1.3 2.0Cost of equity (post tax) 9.00 13.00 10.43 14.96

Debt Premium 1.00 3.50 1.00 3.50Cost of debt (post tax) 3.50 5.95 3.50 5.95

Optimal Gearing 10% 10% 30% 30%Corporate tax rate 30% 30% 30% 30%WACC (post tax nominal) 8.45 12.30 8.35 11.99WACC (pre tax nominal) 12.07 17.56 11.93 17.12WACC (pre tax real) 9.34 14.70 9.20 14.26

14 Effective competition review: mobile – 26 September 2001

198

199

Annex E

Explanation of LRIC and the LRIC+ target charge

Use of LRIC as the cost base

Oftel’s view is that the most appropriate and economically efficient basis forregulatory charge controls is forward-looking LRIC. The LRIC of voice terminationis the additional cost an MNO incurs to provide termination. This can also be seenas the cost that the firm would avoid if it decided not to provide voice termination,taking a long run perspective. It corresponds more closely to the charges thatwould prevail in an effectively competitive market than accounting-basedmeasures of cost. It is a fundamental goal of price regulation to mimic the effectsof a competitive market and this consideration underpins the use of LRIC.

The adoption of LRIC as a regulatory costing technique is used widely, forexample by other NRAs in Europe, and by the FCC in the US. It has also beenidentified as the most appropriate methodology to use for setting interconnectioncharges by the European Commission in its 1998 Recommendation onInterconnection. For further details, see The Use of Long Run Incremental Cost(LRIC) as a Costing Methodology in Regulation, 12 February 2002,http://www.oftel.gov.uk/publications/mobile/ctm_2002/lric120202.pdf.Furthermore, the Competition Commission (CC) has agreed, as stated inParagraph 2.251 of its December 2002 report on mobile voice termination15, withthe use of LRIC as the appropriate costing methodology for setting terminationcharges.

Oftel’s view is that the only relevant costs for the purposes of setting the chargecontrols are those relevant to 2G voice termination. This excludes 3G costs whichthe MNOs can recover through their unregulated 3G charges. The CC took thesame view as Oftel on this issue as stated in Paragraph 2.251 of their report.

LRIC model

The purpose of the LRIC model used by Oftel is to derive the costs of areasonably efficient 2G mobile operator in the UK. In April 2002, Oftel madeavailable the latest version of the model that considered a voice-only network. Themodel and supporting documentation are available athttp://www.analysys.com/ukmobilelric. Further detailed papers are also available:Source of algorithms, data, assumptions and estimates,http://www.oftel.gov.uk/publications/mobile/ctm_2002/analysis300102.pdf; andManual for the Oftel LRIC model,http://www.oftel.gov.uk/publications/mobile/ctm_2002/slides300102.pdf

15 Competition Commission’s report on Vodafone, O2, Orange and T-Mobile, December 2002,http://www.oftel.gov.uk/publications/mobile/ctm_2003/index.htm.

200

In designing the model, Oftel considered five key issues:

the length of the time period over which cost behaviour would be considered;the definition of the increment;the definition of common costs and how these should be recovered;the level of efficiency to be assumed; andthe depreciation method to be used.

The time period

The relevant time period for LRIC is the long run, which can be defined as theperiod over which all assets are replaced. Using short-run costs is impracticalbecause of their volatility, and because they are unlikely to promote efficientdecisions by consumers, and might not allow the operators to recover their costs.For further details see Network Common Costs, 19 February 2002,http://www.oftel.gov.uk/publications/mobile/ctm_2002/network_costs.pdf.

The increment

In modelling the cost of terminating calls, Oftel has defined the increment as thetraffic on all services and then, as a second step, allocated some of the cost of theincrement to the voice call termination service. In the long run, and starting fromcurrent traffic volumes, it is not necessary to define the increment as a singleservice since the choice of the increment should not significantly affect the costcalculation. This is because, even if equipment is shared between services in theshort run, the quantity of equipment required is related to the demand for particularservices. For further details see Network Common Costs (referred to above) andDifferent Views of Oftel and MNOs on Network Common Costs, 27 May 2002,http://www.oftel.gov.uk/publications/mobile/ctm_2002/common_cost0602.pdf.

Common network costs

The only common costs in a mobile network are site acquisition and lease costs ofthe minimum number of base station sites needed to provide coverage across theUK, together with the cost of the network management system. For further detailssee Network Common Costs and Different Views of Oftel and MNOs on NetworkCommon Costs (referred to above). Oftel believes that it is reasonable that thesecosts should be recovered by an equi-proportional mark-up (EPMU). It isappropriate to include a mark-up over LRIC to allow the MNOs the opportunity torecover all of their reasonably incurred costs, including common costs. The use ofEPMU to allow for common cost recovery is reasonable, given the undesirability orimpracticality of other approaches. For example, as discussed in Chapter 7,although in theory Ramsey prices are economically efficient (under certaincircumstances), deriving reliable estimates is completely impractical.

The level of efficiency

201

Oftel has based the LRIC model on a ‘reasonably efficient’ operator. In order toprovide a cross-check that such efficiency is achievable, Oftel has drawn on theinformation in, and results from, the CC’s review of the model that compared itsoutputs with financial and operating information from the MNOs to test thereasonableness of the modelling assumptions. The resulting adjustments arediscussed in detail below.

Economic depreciation

The depreciation approach selected by Oftel for the LRIC model is economicdepreciation16. This matches the cost of equipment to its actual and forecastusage over the long term. As a consequence, there is relatively little depreciationin years where utilisation is low and relatively high depreciation in years of fullequipment utilisation. By contrast, most forms of accounting depreciation arerelatively simple. The usual method is to take the actual price paid for equipment(or its replacement cost) and to divide by the expected equipment life to reach adepreciation charge for the year. The timing of cost recovery under economicdepreciation varies from that under accounting depreciation. Between 2001 and2006 the use of economic depreciation results in a higher pence per minutes costof terminating calls whilst in years prior to 2001, economic depreciation wouldhave resulted in lower costs compared to an equivalent calculation based onaccounting straight-line depreciation. For further details see Accountingdepreciation cost based estimates, 3 May 2002,http://www.oftel.gov.uk/publications/mobile/ctm_2002/account_let0502.pdf.

In its review of the charges for calls to mobiles, the CC agreed with the generalprinciples outlined above and agreed that the April 2002 LRIC model was asuitable starting point for the assessment of the costs of terminating calls onmobile networks (Paragraph 2.287 in their report).

Adjustments to the LRIC model output

The charge controls set out in this market review are based on the costs ofterminating calls on a 2G mobile network as calculated by the April 2002 LRICmodel. However, in the light of the CC’s investigation and their report of December2002, Oftel considers it appropriate to consider a number of issues and potentialadjustments concerning the output of the April 2002 model. These issues are asfollows and are discussed in turn below.

Cost of capital;Comparison with MNO data;Market share; and

16 For further details of the conceptual underpinnings, see Calls to mobile: economic depreciation,September 2001, http://www.oftel.gov.uk/publications/mobile/depr0901.htm. For a discussion of thecost path over time using economic depreciation in the LRIC model for key assets, see AdditionalInformation Concerning Oftel’s LRIC Model, 12 February 2002,http://www.oftel.gov.uk/publications/mobile/ctm_2002/lric_more120202.pdf.

202

Non-network common costs.

Cost of capital

The appropriate cost of capital in the context of this market review is the cost ofcapital for a reasonably efficient 2G mobile operator in the UK, and in particular,the cost of capital regarding 2G termination services.

Oftel has undertaken a fresh analysis of the LRIC model made available in April2002, which calculates the LRIC on the basis of a 12.5% pre-tax real cost ofcapital in 2002/03 and subsequent years. each of the components of the CAPMused to derive an estimate for the cost of capital in the light of more recentinformation. On this basis, Oftel estimates the pre-tax real cost of capital to be inthe range of 9.2% to 14.7% with a mid-point of 12%. Further details of thederivation of this range are provided in Annex D.

The resulting adjustment to the output from the April 2002 LRIC model, which wasbased on a higher cost of capital of 12.5%, is shown in the table below.

Table 1: Adjustment to April 2002 LRIC model output for 12% cost of capital

Pence per minute (real 2000/01) 2001/02 2002/03 2003/04 2004/05 2005/06Combined 900/1800MHz 0.03 0.38 -0.08 -0.07 -0.061800MHz 0.03 0.50 -0.11 -0.09 -0.08

As expected, lowering the cost of capital from 2003/04 results in a lower pence perminute cost of termination in 2003/04 and subsequent years. Less intuitively, itresults in a slight increase in the pence per minute cost in prior years. This resultsfrom the use of economic depreciation as the lower cost of capital in later yearsenables a new entrant to charge lower cost recovery in later years as a result ofthis constraint. However, in order to ensure full cost recovery, it may be necessaryfor the incumbent to recover more cost in earlier years.

Comparison with MNO data

In order to address concerns over the accuracy of the LRIC model, Oftel hasundertaken a comparison between the outputs of the model and actual costaccounting data from the mobile operators. Oftel has derived adjustments to beapplied to the output of the LRIC model following the methodology below, asproposed by the CC in its recent inquiry:

i. separating network from non-network costs in the MNOs’ information;ii. averaging across the resulting information from the four MNOs;iii. averaging the outputs of Oftel’s LRIC model for the combined

900/1800MHz and 1800MHz operators;iv. comparing ii. and iii. at the level of total network costs to derive (separately)

capital and operating cost adjustments; and

203

v. comparing the actual equipment deployments of the combined900/1800MHz and 1800MHz MNOs to derive the 900MHz adjustment.

As the starting point for the analysis, Oftel has taken an average gross book value(GBV) across the four MNOs of £2,680m as at September 2001 (see Table 7.6 ofthe CC’s report) to compare with that calculated by the LRIC model in order todetermine the appropriate capital cost adjustment. Similarly, Oftel has taken theaverage of network operating costs (excluding depreciation) across the fouroperators of £338m for the year to March or December 2001 (see Table 7.7 of theCC’s report) in order to determine the appropriate operating cost adjustment.

The LRIC model considers a voice-only network whilst the information submittedby the MNOs was for their total network including both voice and data services.Thus before comparing with the LRIC model output, a ‘data correction’ factorshould be applied to the cost information provided by the MNOs to obtain a figurefor the costs incurred by the MNOs relating only to voice services. During the CC’sinvestigation, the MNOs submitted estimates of total capacity dedicated to data in2001 that varied from 0% to 14% (Paragraph 7.105 of the CC’s report). Oftelbelieves that 5% is a reasonable estimate of the proportion of capacity installed tohandle existing and forecast data traffic in September 2001 as it would allow oneor more traffic channels to be reserved in some areas, and none in other areaswhere sufficient excess capacity already exists. This estimate is consistent withthat used by the CC in Paragraph 7.108 of their report. Assuming that thepercentage of capacity used for data is appropriate as the percentage ofunderlying costs that can be attributed to data, Oftel considers that applying a 5%reduction to the MNOs total voice and data network cost figures is conservative as5% lies below the average of the figures submitted by the MNOs17, and a lowerdata correction factor results in a higher cost figure to compare to the LRIC modeloutputs and thus ultimately a higher adjustment factor to be applied to thoseoutputs. Oftel believes that the figures resulting from these adjustments are morelikely to overstate efficiently incurred costs than understate them due to the belowaverage data correction factor and the fact that the data underpinning theadjustments reflect the average of any current inefficiency of the MNOs rather thanthe efficient level of costs.

In order to determine the capital cost adjustment, Oftel has compared the averageGBV of the four MNOs at September 2001, reduced by the data correction factorof 5%, with the average output (across combined 900/1800MHz and 1800MHzoperators) from the LRIC model for 2001/02. The LRIC model calculates a nominalGBV for combined 900/1800MHz operators of £1,879m and for 1800MHzoperators of £2,044m in 2001/02. This results in an average of £1,962m.Comparing this model average with the average of £2,680m submitted by theMNOs reduced by 5% implies that an upward adjustment factor should be applied

17 The CC states in paragraph 2.295 of their report that “In the light of a range of estimates ofequipment that was dedicated to data, we considered that taking a figure slightly below the averagewas reasonable because it would be unlikely to lead to an eventual understatement of the ppm ofterminating calls.

204

to the capital costs in the LRIC model of 29.8% in 2001/02. Oftel notes that thisadjustment is very similar to the adjustment used by the CC in paragraph 7.121 oftheir report, based on their comparison of GBV and the number of items ofequipment.

Oftel has also undertaken a comparison of the average MNO network operatingcost with the output of the LRIC model. The LRIC model calculates nominaloperating costs for combined 900/1800MHz operators to be £339m in 2000/01 –the appropriate year to compare for accounts ending in March 2001. Similarly, themodel calculates nominal operating costs for 1800MHz operators to be £370m and£403m in 2000/01 and 2001/02 respectively, which gives a weighted average of£395m as an estimate for the calendar year 2001 to compare with the accounts ofthe 1800MHz operators. The average of the combined 900/1800MHz and1800MHz figures results in an average LRIC model figure of £367m. Comparingthis model average with the average of £338m submitted by the MNOs reduced by5%, Oftel believes that it is appropriate to apply a downward adjustment tooperating costs in the LRIC model by 12.5% in 2001/02. This compares to theadjustment of -11% used by the CC in paragraph 7.122 of their report.

In order to apply these adjustments to equipment costs and operating costsseparately, a split of total costs between capital and operating costs must bederived. Oftel has applied proportions of 57% and 43% for capital and operatingcosts respectively for a combined 900/1800MHz operator18, and 53% and 47% forcapital and operating costs respectively for a 1800MHz operator19 in order to alignthe model to the cost accounting data in 2001/02. Having established the correctbase year of costs, the adjustments for subsequent years have then beendetermined using the overall model forecast cost trend.

Oftel recognises that there is a case for varying the capital / operating costproportions over time in calculating the adjustments for subsequent years, as theLRIC model indicates that these proportions do vary year by year, with operatingcosts growing in importance relative to capital costs. However, on balance Oftelhas chosen to take the more conservative approach of establishing the correctbase year of costs, and using the overall model forecast cost trend for anyadjustments in subsequent years.

Oftel notes and accepts the CC’s conclusion that, at current traffic levels of theMNOs, combined 900/1800MHz and 1800Mhz operators employ a similar amountof network equipment and so have similar costs (see Paragraphs 2.301-2.307 ofthe CC’s report). Since the LRIC model derives a higher cost figure for 1800MHzoperators, Oftel believes that a further adjustment to the final pence per minutecost is appropriate to realign the costs of the two types of operator. There are, 18 The operating cost accounting data provided by the combined 900/1800MHz operators alignswith the 2000/01 financial year. Oftel has thus used capital / operating cost proportions as in theLRIC model for a combined 900/1800MHz operators in 2000/01.19 The cost accounting data provided by the 1800MHz operators is most closely aligned with the2001/02 financial years. Oftel has thus used capital / operating cost proportions as in the LRICmodel for an 1800MHz operator in 2001/02.

205

however, two sources of this difference and it is only appropriate to adjust for thefirst one. The first source of the difference reflects the fact that, on an accountingdepreciation basis, the LRIC model estimates that the cost of 1800MHz operatorsis higher by about 0.2p per minute for terminating calls when compared to thecombined 900/1800MHz operators in the years 2003/04-2005/06. The secondsource of the difference, which is quantitatively more important, derives from theuse of economic depreciation to obtain the path of costs over time. As describedabove, under economic depreciation cost recovery is deferred from earlier years,in which utilisation was lower, to later years, in which higher levels of utilisation areexperienced. However, this effect is more pronounced for 1800MHz networks,because the characteristics of their spectrum imply that such operators have asmaller maximum cell radius and so experience lower site utilisation in the earlieryears of their operation past than the combined 900/1800MHz operators. Theconclusion that, currently, combined 900/1800MHz and 1800MHz operatorsemploy a similar amount of network equipment and so have similar costs does notimply any need to adjust for the second source of difference, but only to removethe difference in the first case.

Taking account of the first point raised in the previous paragraph implies that thegap between the estimated costs for the two types of operators should be reducedby 0.2ppm. However, the operating and capital cost adjustments discussed abovealready bring the average total network cost in the LRIC model in line with theMNOs’ accounting data and so any further adjustment should ensure that theaverage across both types of operator remains unchanged. Hence, theappropriate adjustment is to add 0.1ppm to the costs of the combined900/1800MHz operators, and subtract 0.1ppm from the costs of the 1800MHzoperators so that the average adjustment across all MNOs for this effect is zero.

The derivation of the net adjustment to be made for combined 900/1800MHzoperators and 1800MHz operators in the years up to 2005/06 is shown in the tablebelow.

Table 2 – Net adjustments following comparison with MNO data

Combined 900/1800MHz 2001/02 2002/03 2003/04 2004/05 2005/06April 2002 model LRIC (12% WACC) 4.86 4.57 3.90 3.70 3.53Capital cost adjustment (ppm) 0.82 0.77 0.66 0.63 0.60Operating cost adjustment (ppm) -0.26 -0.25 -0.21 -0.20 -0.19900MHz adjustment 0.10 0.10 0.10 0.10 0.10Net adjustment 0.66 0.63 0.55 0.53 0.51

1800MHz 2001/02 2002/03 2003/04 2004/05 2005/06April 2002 model LRIC (12% WACC) 6.06 5.67 4.77 4.50 4.27Capital cost adjustment (ppm) 0.95 0.89 0.75 0.71 0.67Operating cost adjustment (ppm) -0.36 -0.33 -0.28 -0.27 -0.25900MHz adjustment -0.10 -0.10 -0.10 -0.10 -0.10Net adjustment 0.49 0.46 0.37 0.34 0.32

206

Market share

The output of the LRIC model is based on the costs of an average operator with a25% market share in 2001, declining to 20% by 2010 following the entrance of thefifth operator, “3”.

There is a question whether the fair charge should be adjusted to take intoaccount the position of an MNO with a market share of total traffic lower than theaverage. This question is open to more than one reasonable judgement. Points infavour of implementing an adjustment are as set out by the CC in Paragraphs2.275-2.280 of their report: that is, in the short term the fair charge should take intoaccount the extra cost of an MNO with a market share of traffic lower than theaverage to ensure that it is not unduly disadvantaged by its smaller scale. Incontrast, it can be argued that a market share adjustment is inappropriate, asdetermining the fair charge based on an average operator with average marketshare is the most competitively neutral position to adopt rather than modellingspecific operators.

In the context of this market review the key cost figure is the fair charge in2005/06. Therefore, Oftel concludes that the point is academic. Even in the casethat a market share adjustment were implemented, as proposed by the CC, thenthis adjustment would diminish to zero by 2005/06 by design. The CC state inParagraph 2.277 of their report: “However, over a period of two to three years wethink that an MNO with a lower than average market share has the opportunity tocapture at least an average share of the market. Therefore, by 2005, we wouldexpect there to be no need for any extra cost due to low market share,…”. Thuseither approach results in no adjustment to the fair charge in 2005/06.

Non-network common costs

In addition to the LRIC for network costs of incoming calls, two further additionsare appropriate in determining the total cost of the incoming calls service: a mark-up for any appropriate share of non-network costs and a mark-up to take accountof network externalities. The first issue is addressed here, the second in Annex F.

In general, costs caused by the caller are already included in the LRIC costs.Furthermore, any costs associated with benefits to the caller are not relevant herebut are captured by the mark-up for network externalities. Potentially relevant non-network costs can be grouped into two categories:

(a) customer acquisition, retention and service costs – including advertising andmarketing, handset costs, discounts, customer care, and billing; and

(b) administrative costs – general overheads.

Oftel does not believe that there any exceptional items that can be casually relatedto termination services. In particular, amortisation of 3G licence costs are causallyrelated to 3G services only, and not to 2G voice termination.

207

Regarding the first category, customer acquisition costs are not causally related tothe terminating calls service. The operators choose to acquire retail customers,and the fact that calls are then made to those customers does not result in anyadditional customer acquisition costs. The number of retail subscribers drivescustomer acquisition costs and not the volume of wholesale terminating minutessupplieed. Similarly, other non-network costs in this category such as marketingand advertising costs and customer service are caused by services other thanvoice call termination. Furthermore, whilst a very small element of customerservice costs may be described as common across traffic (in particular, customercall centre costs incurred in helping to resolve technical problems that affectincoming calls), customer service costs vary with either outgoing traffic volume ornumber of subscribers and so are not common costs relevant to termination, whichis a wholesale service. It is therefore inappropriate to recover any of the costsunder (a) in either the LRIC of termination or the mark-up for common costs. Thebenefit that callers to mobiles derive from such costs, in terms of acquisition andretention of additional subscribers, is captured in the externality surcharge. Toinclude such costs also in the LRIC or common costs would amount to doublecounting.

Administration costs include the costs that should be recovered across all areas ofthe business, including both network and retail services. These consist ofoverheads for non-network depreciation (IT, furniture and office equipment),property costs, human resources, finance and legal costs, and IT overheads.Based on information on administration costs for 2001 submitted by the MNOs tothe CC during its investigation (see Paragraph 2.329 of their report), Oftel hastaken £130m a year as an estimate of the reasonably efficient administrationoverheads incurred by a mobile operator.

Given that these costs are common to the supply of termination and other servicesthey should be allocated across all the services concerned. To do so, the totalcosts of the average MNO are derived. Using average information for the fourMNOs, Oftel has taken average network costs in 2001 to equal £607m, and retailcosts - customer acquisition, retention and service costs (before handsetrevenues) - to average £1,276m. Adopting an equi-proportionate mark-up, 32%(£607m / (£607m + £1,276m)) of the £130m administrative overheads can beattributed to network activities. Dividing the resulting administrative costattributable to network by the average total minutes for the year (comprisingincoming, outgoing and on-net voice minutes) of 15.5 billion20 results in a non-network common cost mark-up to the LRIC of voice call termination of 0.27ppm.This figure is rounded-up to 0.3ppm.

During the CC’s consultation, the MNOs generally said that their administrationcosts level could be expected to remain relatively constant with changes in traffic,unless there was a significant increase in traffic or subscriber numbers. Given thatthe forecast traffic in the LRIC model grows only slowly until 2005/06, Oftelconsiders it reasonable that the estimated mark-up of 0.3ppm for non-network 20 Based on information provided to the CC by the MNOs – see Table 7.14 of the CC’s report.

208

common costs should remain constant over this period in real terms. This isconsistent with the CC’s conclusion.

Conclusion

In conclusion, the table below sets out the original LRIC+ figures from the April2002 model and the adjustments discussed above to derive the LRIC+ figures thatinclude the EPMU for network and non-network common costs. The fair charge isthen determined by the sum of the adjusted LRIC+ figures below and thesurcharge for the network externality (discussed in Annex F).

Table 3: Adjusted LRIC+ figures in pence per minute (real 2000/01)

Combined 900/1800MHz 2001/02 2002/03 2003/04 2004/05 2005/06Original April 02 model (12.5% WACC) 4.83 4.20 3.98 3.77 3.58Cost of capital adjust (12.0% WACC) 0.03 0.38 -0.08 -0.07 -0.06Capital / Operating cost adjustment 0.66 0.63 0.55 0.53 0.51Non-network common cost markup 0.30 0.30 0.30 0.30 0.30LRIC+ (excluding network externality) 5.82 5.50 4.75 4.53 4.33

1800MHz 2001/02 2002/03 2003/04 2004/05 2005/06

Original April 02 model (12.5% WACC) 6.03 5.17 4.88 4.60 4.35Cost of capital adjust (12.0% WACC) 0.03 0.50 -0.11 -0.09 -0.08Capital / Operating cost adjustment 0.49 0.46 0.37 0.34 0.32Non-network common cost markup 0.30 0.30 0.30 0.30 0.30LRIC+ (excluding network externality) 6.85 6.42 5.43 5.15 4.89

209

Annex F

Evaluation of the surcharge for the network externality

Introduction

An externality is a cost or benefit that is imposed on one or more parties by aconsumption (or production) decision taken by another party. Since the cost (orbenefit) does not affect the party who makes the decision, the latter in generaldoes not take this cost (or benefit) into account. Hence the presence ofexternalities can lead to sub-optimal decisions. In principle, to achieve economicefficiency, prices should be adjusted to ‘internalise’ the externality, so as to providethe decision-maker with an incentive to consider the effects of her decision onothers. To the extent that the externality is already internalised by the marketparticipants, either by consumers or producers, such corrective action is notneeded.

If a market is characterised by a positive consumption externality (i.e. one thatgenerates a benefit to others) and the externality is not (completely) internalised,the level of consumption will be below the efficient level. Each consumer willconsume up to the point where her marginal private benefit is equal to the price.However, the efficient level of consumption is the one at which the marginal costequals the marginal social benefit (which, due to the externality, is higher than themarginal private benefit).

The decision by a consumer to be part of a telecoms mobile network generates anumber of externalities, the most relevant of which is referred to here as the‘network externality’. The network externality is the benefit obtained by existingtelephone users when a person decides to become a new mobile subscriber,because they will be able to call or be called by her (callers value these calls) or,at least, will have the option to do so (the so-called ‘option value’). However, theperson deciding whether to join a mobile network will only take account of thebenefit that she obtains and not the benefit that her decision generates for others(the ‘external benefit’). Hence a person may choose not to join the networkbecause her private benefits do not cover the price of becoming a subscriber,even though economic welfare would be enhanced if she did, because of thebenefits obtained by others. As a consequence the number of subscribers isbelow the efficient level.

In some cases consumers internalise externalities themselves. For example,some people may contribute to the costs of acquiring the mobile handset and paythe relevant subscription for others with whom they have substantial community ofinterest (e.g. parents for their children or firms for their employees). However, notall of the network externality is thus internalised and, therefore, there is anargument for introducing some corrective pricing to reflect this un-internalisedexternality.

210

For some consumers there is an argument, therefore, that prices should be belowcost, i.e. subsidised, to entice them to become (or remain) mobile subscribers.The set of consumers for whom this consideration is relevant are the ‘marginalsubscribers’, who are defined as those subscribers who would not join (or remainon) the network in the absence of a subsidy, but whose addition to, and retentionon, the network increases economic efficiency because of the external benefitsgenerated. It should be noted that subsidies provided to infra-marginalsubscribers do not affect welfare, because these would subscribe even in theabsence of a subsidy (this is indeed the definition of infra-marginal subscribers).

If a subsidy is to be provided to marginal subscribers by the MNOs, then they needto have the opportunity to recover such costs. Hence, Oftel considers that it wouldbe appropriate for the wholesale termination charge to include a contributiontowards the recovery of these subsidies. This will be referred to as the ‘externalitysurcharge’.

Deriving the optimal externality surcharge in theory

In principle a variety of factors and considerations is relevant when seeking toderive a figure for the externality surcharge. Discussed below is a non-exhaustivelist of the factors in the relevant economic theory.

Size of subsidy

Number of marginal subscribers

The number of marginal subscribers depends on estimates of the marginal costand the marginal social benefit. The latter comprises the sum of the marginalprivate benefit and the external benefit.

The marginal private benefit is captured in the demand for mobile subscription.Hence, the rate at which the marginal private benefit declines as the volume ofsubscribers increases depends on the elasticity of demand.

The size of the external benefit relative to the private benefit is referred to as theGross Externality Factor or the Rohlfs-Griffin factor (the ‘R-G factor’). A value of 1for the R-G factor would imply that there were no external benefits, or that theywere entirely internalised. This is therefore the lower bound. A value of 2 wouldimply that the external benefit to existing subscribers as large as the privatebenefit obtained by the marginal subscriber. On the basis that it is reasonable tosuppose that the external benefit to existing subscribers is unlikely on average toexceed the private benefit to the marginal subscribers, a value of 2 for the R-Gfactor provides a reasonable upper bound. However, the value of 2 does not allowfor any internalisation of the externality, and the value of 1 does not allow for any(uninternalised) externality at all. As explained above, there is evidence that someinternalisation of this externality takes place. Hence, given these considerations,the Director believes that a reasonable range for the Rohlfs-Griffin factor is 1.3 to1.7.

211

Targeting

A key issue that affects the size of the required subsidy is the extent to which it isfeasible for MNOs to target the provision of subsidies on marginal subscribers.One possibility is the scenario where absolutely no targeting can be achieved, i.e.the MNOs charge the same price to all subscribers, both marginal and infra-marginal ones. However, this scenario is clearly unrealistic, since the MNOs offera variety of tariff packages and engage in substantial price discrimination. At oneextreme there is the scenario with perfect targeting, i.e. the MNOs charge abespoke price to all marginal subscribers and provides no subsidy to the infra-marginal ones, which would result in the lowest required subsidy. There are otherpossibilities. For example, it could be assumed that the MNOs were able toseparate out marginal from infra-marginal subscribers, but among the group ofmarginal subscribers were unable to offer different prices to different consumers.Or it could be assumed that the MNOs were unable completely to separate thesegroups and that some infra-marginal subscribers were able to ‘free-ride’ on thesubsidy intended for the marginal subscribers.

Funding of subsidy and welfare trade-offs

Welfare trade-off

The economically efficient externality surcharge reflects a trade-off between twofactors. First, a larger surcharge on termination leads to an incremental welfaregain to the extent that it induces additional subsidies to be provided to marginalsubscribers. As the termination charge rises, each further increase would result ina smaller incremental welfare gain, because additional subscribers yield aprogressively smaller benefit, as they can be expected to call and be called lessthan customers with a higher willingness to pay.

Second, a higher termination charge leads to higher retail prices for calls tomobiles and so imposes an incremental welfare loss on callers to mobile. As thetermination charge rises, each further increase would result in a larger incrementalwelfare loss, because calls to mobile would become progressively more expensiveand the more that they are priced above cost, the greater the welfare lossessuffered by callers to mobile.

The economically efficient externality surcharge is the level at which theincremental welfare gain and incremental welfare loss are equal. At lowertermination charges there remain net welfare gains that can be exploited byincreasing the charge and the subsidy. At higher termination charges the welfarelosses imposed on callers to mobile are larger than the benefits from additionalsubscribers, so that welfare can be increased by reducing the termination chargesand the subsidies. An important implication is that it is economically efficient for

212

some marginal subscribers not to be induced onto the network.21 This is becausethe welfare loss from doing so exceeds the welfare gain.

Funding the subsidy across all services

The economically efficient funding of the subsidy to (some) marginal subscriberswould involve surcharges on other mobile services as well as termination,including mobile originated calls and subscription. This is because of the relevantwelfare trade-offs.

It would not be the most economically efficient outcome if the entire subsidy werefunded through termination charges. As discussed above, it would imposesignificant and increasing welfare losses on callers to mobile. A moreeconomically efficient outcome could be achieved by funding the subsidy throughall prices. Although this would lead to welfare losses on those other services, thesize of such losses would be smaller than the incremental welfare losses thatwould arise if the subsidy were funded solely from a surcharge on termination.

The surcharge that should apply to each service would depend on the relativesizes of their elasticities of demand. The service with the demand that was mostinelastic, i.e. least sensitive to price, should have the largest surcharge. This isbecause a given price increase for a service with a more inelastic demand wouldresult in a smaller reduction in volume, and consequent loss of welfare, than for aservice with a more elastic (price-sensitive) demand.

In the comparison of the relative sizes of the elasticities, both own-price and cross-price elasticities are relevant. The own-price elasticity of service A is thepercentage change in demand caused by a small change in the price of A. Thecross-price elasticity of service B with respect to A is the percentage change indemand of B caused by a small change in the price of A.

Additional considerations since retail prices are unregulated

Imperfect competition

If the retail mobile market is imperfectly competitive, then some of the surchargeon termination will be taken by the MNOs in the form of supra-normal profit, ie notall of the surcharge will flow through into lower retail prices and subsidies tomarginal subscribers. This is another source of inefficiency in convertingsurcharge on termination into subsidy to marginal subscribers. The result that thesurcharge should be lower is unambiguous in theory when all welfare weight isplaced on consumer surplus. For further details, see Oftel’s Response to theCompetition Commission’s Letter on Externalities of 28 March, 15 April 2002

21 It is assumed that the ‘first best’ outcome is not feasible, ie all outgoing and incoming call pricesat marginal cost and the subscription price below cost (because of the externality), since it wouldresult in a loss for the MNO.

213

(pages 11-15),http://www.oftel.gov.uk/publications/mobile/ctm_2002/externalities150402.pdf.

Deriving the externality surcharge in practice

As can be seen from the discussion above, the assessment of the economicallyefficient externality surcharge in theory is a complex and multi-faceted issue. Awide variety of factors should be considered when setting this surcharge.Specifying a model that captures all of the relevant conceptual points would be avery difficult task, given the complexity of some of the factors. In addition, thereare many parameters whose variables cannot reliably be estimated in practice,such as the elasticities of demand.

Given these difficulties, the Director considers that it should not rely on a singleestimate of the externality surcharge. As discussed below, there are severaldifferent quantification exercises that have been carried out that the Directorconsiders useful. Each of them, considered individually, is incomplete because itonly captures a sub-set of the factors and/or because some of the assumptionsused (e.g. elasticities) cannot in practice be robustly derived from empirical data,but they are helpful in informing the reasonable value for the externality surcharge.

The CC approach

The CC calculated an optimal surcharge of 0.45 pence per minute22. It arrived atthis figure by assessing the subsidy to the handset necessary to attract/retain onmobile networks the efficient number of subscribers, allowing for targeting ofsubsidies to marginal subscribers.

More specifically the CC’s methodology consists in evaluating three elements:-

i) the subsidy to non-current subscribers;ii) the subsidy to existing subscribers; andiii) the extra subsidy to existing subscribers for whom the subsidy to non-

current subscribers is more attractive.

(i) The subsidy to non-current subscribers is calculated as the product of theoptimal subsidy per subscriber (assumed to be uniform for all) and the number ofrelevant non-subscribers. This number includes both non-current subscribers whoare marginal, and allows for some ‘free-riding’ on the subsidy from non-currentsubscribers who are infra-marginal.

(ii) The subsidy to existing subscribers is calculated as the product of theaverage subsidy per subscriber (calculated assuming a different, bespoke subsidyto each subscriber), and the number of existing subscribers deemed to bemarginal.

22 See Chapter 2, Chapter 8 and Annex 4 of the CC Report.

214

(iii) The CC relaxed the assumption that perfect targeting is possible, toconsider that existing subscribers, whose bespoke subsidy would be lower thanthe subsidy offered to non-current subscribers, would take advantage of this offer.This was because the CC considered that it would be very difficult to preventexisting subscribers from taking advantage of the subsidy offered to non-currentsubscribers.

Further details of the CC’s calculation are set out in Chapters 2 and 8 and Annex 4of the CC report.

Rohlfs’ model

Introduction

During the CC’s inquiry, Oftel commissioned Dr Rohlfs to develop a model forestimating optimal prices for mobile services, taking account of the networkexternality. The model is described and the key assumptions are set out in AModel of Prices and Costs of Mobile Network Operators, Dr Rohlfs, 22 May 2002,http://www.oftel.gov.uk/publications/mobile/ctm_2002/main_report.pdf.

In particular, Dr Rohlfs ensured that his demand assumptions were consistent withcredible underlying economic relationships. Specifically, as set out above, it isreasonable to take the view that the Rohlfs-Griffin factor – the ratio of the externalbenefits (accruing to existing subscribers) to the private benefits (accruing to themarginal subscriber) - is in the range 1.3 to 1.7. To a large extent the externality isreflected in the cross-elasticities of demand. For example, the demand for calls tomobile increases when the mobile subscription price falls, because this leads to anincrease in the number of mobile subscribers, who are then called by the pre-existing customers.

Removing the effect of the externalities embodied in cross-elasticity assumptionsfrom the Rohlfs-Griffin factor yields the option externality factor (or net externalityfactor’). This is the ratio of the external benefit arising just from the option to callthe new subscriber to the private benefit accruing to the marginal subscriber. Theoption externality reflects the value placed by consumers on the ability to call anew subscriber on his mobile phone, as distinct from actually calling him (which ispart of the cross-elasticities and the Rohlfs-Griffin factor). It is reasonable tobelieve that the option externality is relatively small and so that the net externalityfactor lies between about 1 and 1.1. Such considerations are reflected in theassumptions and sensitivities implemented by Dr Rohlfs23.

Dr Rohlfs developed an initial Ramsey pricing model – this forms the basis for theupper bound figure, discussed below. Dr Rohlfs also extended the model tocapture the effect on the externality surcharge of allowing for the targeting ofsubsidies - this forms the basis for the lower bound figure, discussed below. A

23 See response under point 5 in Ramsey pricing – Oftel’s response to letter of 4 July, 12 July 2002 http://www.oftel.gov.uk/publications/mobile/ctm_2002/ramsey0702.pdf]

215

further variant was the principal-agent model which explicitly models the retailprice-setting behaviour of MNOs, albeit in simplified form.

Upper bound

Dr Rohlfs’ initial model derives an estimate of the externality surcharge assumingno targeting, i.e. all subscribers pay the same prices. His base case set ofassumptions yields a figure for the mark-up over LRIC of 0.77 ppm. This shouldbe treated as an upper bound, because it makes no allowance for targeting, or forinefficiencies in converting the surcharge into subsidies to marginal subscribers.

However, Dr Rohlfs’ figure is a mark-up over LRIC, whereas the figure derived forthe externality surcharge will be added to LRIC plus EPMU for (network and non-network) common costs. The mark-up derived in Dr Rohlfs’ model reflects twoeffects: a mark-up for common costs and a surcharge for the network externality,whereas the Director is seeking to quantify only the latter effect. This means that itis appropriate to make a downward adjustment to the mark-up derived in DrRohlfs’ model to take out the part of the mark-up relating to common costrecovery. A reasonable method to make this adjustment is to subtract the EPMUfor network costs24 derived when calculating the fair charge. The EPMU fornetwork costs in the fair charge is about 0.22 ppm for 900 MHz and about 0.33ppm for 1800 MHz. Subtracting an average of 0.28 ppm yields an adjusted upperbound based on Rohlfs’ analysis of 0.49 ppm.

Lower bound

Dr Rohlfs extended his model to derive estimates of the externality surchargewhen a high degree of targeting is allowed for. The degree of targeting modelledwas the ability to target the set of marginal subscribers, but offering them all thesame price (rather than bespoke prices as under perfect targeting). This wasmodelled using ‘2- and 3-part pricing plans’25 –

From this analysis of targeted subsidies, Dr Rohlfs estimated that the externalitysurcharge would be 0.06 ppm for the 2-part pricing plan and 0.05 ppm for the 3-part plan. The ability of MNOs to target subsidies on marginal subscribers may fallshort of the amount assumed in the 2- and 3-part pricing plans. These estimatesof the surcharge with targeted subsidies should therefore be regarded as providinga lower bound.

Principal-agent model

The Director also notes that Dr Rohlfs has attempted to quantify the implicationsfor the externality surcharge of private incentives to provide subsidies and

24 The costs in Rohlfs model include all network common costs, but do not include non-networkcommon costs.25

For further details see Estimates of Targeted Subsidies, Dr Rohlfs, 19 June 2002http://www.oftel.gov.uk/publications/mobile/ctm_2002/rohlfs0602.pdf

216

imperfect competition in the retail mobile market. Dr Rohlfs has referred to thisanalysis as his principal-agent (‘PA’) model in which the regulator is assumed tomaximise economic efficiency, subject to the constraint that MNOs pursue theirown agendas in setting retail prices under imperfect competition. In his paper, DrRohlfs reports the results of his analysis as follows:

“The PA model yields lower prices than the Ramsey model for both fixed-to-mobile (FTM) and off-net prices. The intuition behind this result is as follows:the PA model takes a more pessimistic view than does the Ramsey modelabout how well MNOs will use the revenues that they accrue through mark-upson mobile-termination rates. In particular:

It is assumed that MNOs will not necessarily pass all the increased revenues on inthe form of lower prices; and it is assumed that they will not give full weight to theexternal benefits of subscription to fixed subscribers.

Thus, they are likely to use an inefficiently large proportion of the incrementalrevenues to lower usage prices, as opposed to the subscription price. Thispractice would not cause any problems if all subscribers had the same usage. Inthat case, buy-through would ensure reductions in usage prices and subscriptionprices have equivalent impacts on demand for subscription. In reality, mobileusage varies enormously among subscribers. MNOs may lower usage prices tostimulate additional usage by infra-marginal subscribers. Such reductions wouldhave relatively little effect on marginal subscribers (who typically have low usage)and would therefore have relatively little effect on demand for subscription. As aresult, fixed subscribers would derive relatively little benefit from FTM mark-upsabove incremental cost.

Since MNOs are expected to use the revenues less effectively in terms ofmaximizing economic efficiency, it is optimal to have lower FTM mark-ups andgenerate less of those revenues. The

I derive optimal mark-ups over incremental cost for FTM usage of 0.14 to 0.42ppm using the PA model, which takes account of the likely non-optimalresponse of MNOs to the prices that Oftel determines.”26

As for the upper bound estimate, a downward adjustment should be applied to thisfigure to subtract the mark-up for common cost recovery. This yields an estimatedexternality surcharge of 0.14 ppm or less.

Surcharge in previous caps

The previous caps in place on Vodafone and O2, which were implementedfollowing the CC’s 1998 inquiry, made an allowance for the externality surchargeof 0.5 ppm. For further details of the derivation of this figure, see the Reports onreferences under section 13 of the Telecommunications Act 1984 on the charges

26 A Model of Prices and Costs of Mobile Network Operators, 22 May June 2002

217

made by Vodafone and Cellnet for terminating calls from fixed-line networks by theMonopoly and Merger Commission (December 1998).

Other estimates provided during the CC’s inquiry

Consultants working on behalf of the MNOs provided various pricing models andestimated externality surcharges during the course of the CC’s inquiry, such asDotEcon on behalf of O2, Frontier Economics on behalf of Vodafone and CharlesRiver Associates on behalf of T-Mobile. For further details of these models, seeChapters 8 and 9 of the CC report.

The Director does not consider that any of these models provided useful resultsthat should be taken into account in this market review. This is because suchmodels were subject to methodological flaws and/or implausible elasticityestimates. As discussed above, it is reasonable to believe that the R-G factor liesbetween 1.3 and 1.7 and certainly between 1 and 2. The elasticity assumptionsused in these models, including the cross-price elasticities, imply significantlylarger values for the R-G factor. Therefore, the Director considers that theelasticity assumptions used in these models are implausible and cannot be reliedupon, because they are inconsistent with credible underlying economicrelationships. Further difficulties with these models are set out in Dr Rohlfs’comments reported in Appendix 9.1 of the CC report.

Summary of estimates and conclusion

The quantifications described above are summarised in Table X.1 below.

Table X.1: Quantifications of the optimal externality surchargeRohlfs’ model– upper bound(untargetedsubsidies)

Surcharge inearlier caps(following 98MMC report)

CC report2002

Rohlfs’ PAmodel

Rohlfs’modelwith targetedsubsidies

0.49 ppm 0.5 ppm 0.45 ppm 0.14 ppm 0.05 ppm

As discussed above, the Director believes that none of these quantifications canbe solely relied upon, because they are all incomplete or rely on parameterswhose values are quite uncertain and difficult to quantify (such as elasticities ofdemand). However, it considers that each provides a useful insight. Based on theavailable evidence, the Director considers that a reasonable externality surchargeis 0.4 ppm.

218

Annex G

Fair target charge and the level of charge control

Structure of the charge control

As described in Chapter 7 Oftel’s approach to controlling charges is to set a ceilingon the weighted average charge that an operator can levy during each year of thecontrol. This involves identifying the fair target charge in 2005/06, the final year ofthe proposed control, and then calculating the appropriate percentage by whichthe current charge should fall each year from the average existing charges in2002/03, allowing for inflation. The formula for this calculation is generallydescribed as RPI (Retail Price Index) – X.

Oftel considers that the charge control regime should last until March 2006 when itwill be appropriate to undertake a subsequent review of the market. Hence thereare three periods over which the proposed control will be in force:

a) Period ending 31 March 2004;b) 1 April 2004 to 31 March 2005 (2004/05); andc) 1 April 2005 to 31 March 2006 (2005/06).

Oftel proposes that the target average charge (TAC) at the end of the first periodshould be set as an absolute target in pence per minute and thus specified interms of the RPI-X formula only for the second and third periods of the control.The underlying intention for this adjustment to the structure of the charge controlsis to allow the charges of the four operators to be aligned in two ways: firstly, it isdesirable for Orange and T-Mobile to have the same TAC, reflecting the identicalefficient costs that they incur as 1800MHz operators27. Secondly, although the faircharge differs in each year for combined 900/1800MHz (Vodafone and O2) and1800MHz operators (Orange and T-Mobile), it is desirable to align the targetcharges of the two types of operators in this first period so that the TAC for each isthe same amount above the fair charge. Alternatively stated, Oftel considers itappropriate to align the target charges so that the gap between the TAC for thetwo types of operators equals the gap between the fair charge for each type ofoperator at the end of the first period. Thus one type of operator would not have anadvantage over the other which potentially might result in a distortion in retailcompetition.

Fair target charge

Before determining the level of the charge control, it is first necessary to determinethe fair target charge at the end of the control period. The fair target charge isbased on the LRIC of termination plus two mark ups, the first an equal

27 Currently Orange and T-Mobile have different average charges.

219

proportionate mark up for the recovery of network and non-network commoncosts, and the second an allowance for the value of the network externality.

The derivation of the appropriate LRIC including the mark up for recovery ofcommon costs is discussed in Annex E, and the value of the network externalitysurcharge is addressed in Annex F. The sum of these determines the fair chargein each year as shown in pence per minute (expressed in real 2000/01 terms) inthe table below.

Table 1: Fair charge in each year in pence per minute (real 2000/01 prices)

900/1800MHz (ppm, real2000/01)

2001/02 2002/03 2003/04 2004/05 2005/06

LRIC+ mark up for common costs 5.82 5.50 4.75 4.53 4.33Network externality 0.40 0.40 0.40 0.40 0.40Fair charge 6.22 5.90 5.15 4.93 4.73

1800MHz (ppm, real 2000/01) 2001/02 2002/03 2003/04 2004/05 2005/06LRIC+ mark up for common costs 6.85 6.42 5.43 5.15 4.89Network externality 0.40 0.40 0.40 0.40 0.40Fair charge 7.25 6.82 5.83 5.55 5.29

For the purpose of the charge control, the critical charges are the fair charges in2005/06 which are the target charges for the end of the control period. The fairtarget charges of 4.73ppm and 5.29ppm (in real 2000/01 terms) for combined900/1800MHz and 1800MHz operators respectively are very similar to the chargesfor 2005/06 determined by the Competition Commission (CC) and set out in Table2.11 of their December 2002 report.

Regarding the difference in fair charges for combined 900/1800MHz and 1800MHzoperators, as discussed in Annex E, Oftel notes and accepts the CC’s conclusionstated in Paragraph 2.301 of their report that, at current traffic levels of the MNOs,both types of operators employ a similar amount of network equipment and sohave similar costs. However, the difference in the fair charge reflects thedifference in LRIC derived from the use of economic depreciation to obtain thepath of costs over time. Under economic depreciation cost recovery is deferredfrom earlier years, in which utilisation was lower, to later years, in which higherlevels of utilisation are experienced. This effect is more pronounced for 1800MHznetworks, because the characteristics of their spectrum imply that such operatorshave a smaller maximum cell radius and so experience lower site utilisation in theearlier years of their operation than the combined 900/1800 MHz operators, thusresulting in a higher fair charge in the future.

Calculation of X

The value of X for the charge control can be calculated from the starting charge (atthe beginning of the first period) and the fair target charge (in 2005/06). Thestarting charge is derived from the nominal prices charged in 2002/03 after

220

implementation of the 15% reduction in real terms as recommended by the CC, totake effect within the period from 1 April to 25 July 2003. Oftel requestedinformation from the MNOs in March 2003 regarding their termination charges andcall volumes, enabling the average charge for 2002/03 to be calculated using2002/03 time of day charges and 2001/02 call volumes, consistent with standardpractice in previous proposed charge controls. This confirmed the average chargefor combined 900/1800MHz operators to be 9.35ppm (as required by the existingprice controls on Vodafone and O2) and showed the average charge to be11.0ppm for 1800MHz operators.

By considering the starting charges and target charges in real terms (2000/01prices), the value of X can be deduced from the real percentage reduction requiredto reach the fair target charge in 2005/06 after three successive applications of theRPI-X control. The following table shows the results this calculation for combined900/1800MHz and 1800MHz operators separately.

Table 2: Derivation of average real percentage reduction over three periods

combined900/1800MHz operators

1800MHz operators

Starting charge (2002/03)Nominal ppm 9.35 11.00Real ppm (2000/01 prices) 9.02 10.61After 15% real reduction (by 25 July 2003) 7.67 9.02

Target fair charge (2005/06)Real ppm (2000/01 prices) 4.73 5.29

Annual real reduction over three periods(rounded)

15% 16%

As described above, Oftel proposes to set the TAC for the first period as anabsolute target in pence per minute so that the gap between the target charges forthe two types of operators exactly matches the gap in the fair charge at the end ofthis period (the end of 2003/04).

From Table 2, the termination charge for combined 900/1800MHz operators at thestart of the control period is 7.67ppm (in real 2000/01 prices). Applying a 15%reduction in real terms consistent with the RPI-15 charge control implied by thecalculation outlined in Table 2 results in a target charge of 6.52ppm (in real2000/01 prices) for combined 900/1800MHz operators at the end of the firstperiod. From Table 1, the gap in the fair charge at the end of the first period (endof 2003/04) is 0.69ppm (5.83ppm – 5.15ppm) which when added to 6.52ppm givesa target charge of 7.21ppm (in real 2000/01 prices) for 1800MHz operators.Expressed in nominal terms, the target average charge at the end of the firstperiod is 6.95ppm for combined 900/1800MHz operators and 7.70ppm for1800MHz operators.

221

To complete the calculation, the value of X for the RPI-X control in the remainingtwo periods can be derived from the real percentage reduction necessary toreduce the TAC at the end of the first period to the fair target charge in 2005/06.The following table shows the results of this calculation for combined900/1800MHz and 1800MHz operators separately.

Table 3: First period TAC and derivation of value of X for 2nd and 3rdperiods

combined900/1800MHz operators

1800MHz operators

1st period target charge (real 2000/01 ppm) 6.52 7.211st period target charge (nominal ppm) 6.95 7.70

Final target fair charge (real 2000/01 ppm) 4.73 5.29Value of X for 2nd & 3rd periods (rounded) 15% 14%

In summary, Oftel proposes that the target average charge for the first periodshould be set at 6.95ppm for combined 900/1800MHz operators and 7.70ppm for1800MHz operators, followed by a charge control of RPI-15 for combined900/1800MHz operators and RPI-14 for 1800MHz operators in the remaining twoperiods.

222

Annex H

Charge cap mechanism

Introduction

The charge caps on the 2G voice termination charges of O2, Orange, T-Mobileand Vodafone will act on the weighted average charge. Each operator hasdifferent termination charges for three times of day/days of week: daytime, eveningand weekend and Oftel does intend to regulate the specific level of these charges.

The charge cap constraint is that the average charge set by the mobile operator(the Average Interconnection Charge or AIC) must not exceed the charge withwhich the regulated operator is required to comply (the Target Average Charge orTAC) for that control period.

Average Interconnection Charge

This section contains a description of how Oftel proposes the AIC should becalculated.

Average monthly charge

The first step in the calculations of the AIC consists in assessing the averagemonthly charge. The average charge in each month is calculated by:

• taking the termination charges by time of day (i.e. Daytime, Evening,Weekend); and

• deriving the average using as weights the share of call minutes of each timeof day period in the same month experienced during the prior financial year.

This discussion makes the simplifying assumption that any changes in terminationcharges occur on the first day of a month. If charges were to change during thecourse of the month, volume weights for parts of months would be needed,corresponding to the periods during which different termination charges were inforce.

This is shown in table X.1 with illustrative figures.

Table X.1: Illustration of calculation of average monthly chargeCharges in Octoberof Year 2

Volume shares bytime of day inOctober of Year 1

Weighted averagecharge in October ofYear 2

Daytime 15 55%Evening 10 25%Weekend 5 20%

223

Total 11.75Note: Figures are purely illustrative.

Average charge over the control period

The second step in the calculations of the AIC consists in deriving the averagecharge for the whole control period. The control period in general runs from 1stApril of one year to 31st March of the following year. Hence, the average charge inthe control period, the AIC, is equal to the mean of the twelve average monthlycharges weighted by the share of call minutes in each month experienced in theprior year, as illustrated in Table X.2. However, if the control period is shorter,clearly the calculations will be reduced to cover only the relevant months anddays.

Table X.2: Illustration of calculation of Average Interconnection ChargeMonthly averagecharges in Year 2

Volume shares bymonth in Year 1

April 12.3 7%May 12.3 7%June 12.3 7%July 10.9 7%August 10.9 8%September 11.75 8%October 11.75 8%November 11.75 9%December 11.75 9%January 11.75 10%February 11.75 10%March 11.75 10%

Total 11.74Note: Figures are purely illustrative

The AIC in a control period is, therefore, calculated as:

Charges by time of day and month in this control period,multiplied by volume weights by time of day and month in theprevious year

Variation of AIC with changes in weights

The AIC varies with changes in the weights, even at unchanged time of daycharges. For example, it has been observed in recent years that there have beensignificant increases in mobile termination volumes in evening and weekendperiods relative to daytime. This means that the annual average charge

224

decreases from one year to another, even at unchanged day, evening andweekend charges, because of the rise in the evening and weekend weights. Thisis illustrated in Table X.3, which shows a fall in the weighted average charge of 6%from 11.75 to 11.0, solely because of changes in the weights.

Such a change in the AIC could thus contribute towards satisfying compliance withthe TAC set for that control period. In other words, if the weights were to changeby a sufficient amount (and in the right direction), it is possible that compliancewith the TAC could be achieved without reducing any charges. Indeed, it istheoretically possible that a situation could arise in which the regulated operatorwould be able to increase charges and still comply with the target average charge.Movement of the weights in this direction would provide the regulated operatorwith an unearned gain.

Table A6.3: Illustration of reduction in average charge under current methodCharges inYears 2and 3

Volumeshares inYear 1

AIC in Year 2Chargesin Years 2and 3

Volumeshares inYear 2

AIC in Year 3

Daytime 15 55% 15 45%Evening 10 25% 10 30%Weekend 5 20% 5 25%6 Total 11.75 11

Note: Figures are purely illustrative

It can be noted that this effect is symmetric. That is, if the time of day weights wereto move in the opposite direction, for example with an increase in the proportion oftraffic in the daytime, the AIC would increase, even with no change in anytermination charge by time of day. This would make it more difficult for theregulated operator to comply with the TAC and impose a loss on the regulatedoperator.

The Target Average Charge

This section contains a description of how Oftel proposes the TAC should becalculated when is specified on the basis of an RPI-X formula.

The charge cap constraint when expressed with an RPI-X formula means that theTAC in any year needs to be lower than the TAC in the previous control period byat least RPI-X%. However, if there had been changes in the weights from oneyear to the other, this should be considered when assessing the appropriate TAC.

Oftel proposes that the TAC in each control period should be derived using asweights the volume of call experience in the prior year. The TAC in every controlperiod would thus depend on the weights used to calculate the AIC in that samecontrol period. This would guarantee consistency in the weights used forcalculating the two charges for the same control period.

225

If in the last control period the AIC was equal to the TAC

Assume initially that in the last control period the AIC exactly satisfied the TAC.

X Then the TAC for the current control period would be equal to:Charges by time of day and month last year,

multiplied by volume weights by time of day and month previous year,multiplied by the sum of 100 per cent and the Controlling Percentage

The proposed method is shown with illustrative figures in table X.4

Table X.4: Illustration of new method to calculate target average chargeAveragecharges inYear 2

Volumeshares inYear 2

TAC inYear 3

Charges inYear 3(if unchanged)

Volumeshares inYear 2

AIC inYear 3

Daytime 15 45% 15 45%Evening 10 30% 10 30%Weekend 5 25% 5 25%

Total 11 11TAC Assume RPI-X =

-6.4%10.30

Note: Figures are purely illustrative

If in the last control period the AIC was not exactly equal to the TAC

The expression contained in X above for the TAC is correct only if in the lastcontrol period the AIC was equal to the TAC. That expression needs to bemodified if the assumption on the AIC does not hold.

Consider the case in which in the last control period the AIC that was lower thanthe TAC (which is referred to as an ‘under-shoot’). If the TAC was set using theexpression in X above, the operator would obtain no credit for the ‘under-shoot’last year i.e. the TAC in this control period would be lower than if there had beenno under-shoot.

To avoid this effect, an adjustment should be made to the formula for the proposedmethod. The final expression for the proposed method to calculate the TAC for thecurrent control period is:

Charges by time of day and month previous year,multiplied by volume weights by time of day and month previous year,multiplied by the sum of 100 per cent and the Controlling Percentage,

divided by the sum of 100 per cent and the Adjustment Factor for under/over-shootprevious year

where the Adjustment Factor is the percentage by which AIC in the last controlperiod was different from the TAC and is:

226

• Negative for an under-shoot (i.e. AIC below TAC) - thus the AdjustmentFactor would increase the TAC in this control period, because the originalexpression (set out in X above) is divided by a number less than one. Thisis in order that the target for this control period is calculated as if there hadnot been an under-shoot in the last control period.

• Positive for an over-shoot (i.e. if the operator has failed to comply with thecap by setting the AIC above the TAC) - thus Adjustment factor wouldadjust the TAC downwards by dividing the original expression (set in Xabove) by a number greater than one. This would avoid that the regulatedoperator, who had failed to comply with the charge cap in the last controlperiod r, does not benefit from a higher TAC.

The Adjustment Factor is purely intended to ensure that the TAC in each controlperiod is not distorted by any under- or over-shooting in the previous controlperiod. It guarantees that the target is calculated as if there had been neitherunder- nor over-shoot. There is a separate, additional question of the adjustmentsthat should be made to the target in order to compensate customers for anovershoot in the prior control period or the regulated operator for an undershoot(where carryover is permitted). These adjustments are not discussed here.

227

Annex I

The treatment of ported numbers

Portability of mobile numbers has been available in the UK since January 1999.Number portability is a facility provided by one MNO to another which enablesmobile subscribers to keep their telephone numbers when switching betweenthose operators.

Under the current technical arrangements for the provision of mobile numberportability, calls to a ported number are routed via the donor MNO to theterminating MNO. This means that the donor MNO, in effect, acts as a transitcarrier conveying the call from the originating operator to the terminating MNO.The commercial arrangements provide that the terminating MNO receives thetermination charge of the donor network (less a transit fee). This means that forthese calls the terminating MNO receives an amount which is different from itsown termination charges. Hence, calls to ported numbers are ‘ported in’ from theperspective of the terminating network and ‘ported out’ from the perspective of thedonor network.

The level of porting of mobile numbers has now increased: 1,7 million mobilenumbers (out of 34 million) have been ported between January 1999 andSeptember 200228. The volume is becoming significant enough to warrant properconsideration of how they should be treated in the control on off-net terminationcharges. Furthermore, at the level of an individual operator the importance ofported numbers can be proportionally much larger.

Options

The Director believes that it is undesirable that ported numbers are completelyomitted from the charge control. Given the growing level of ported numbers, to doso would have the effect of diminishing the effect of the control and leaves openthe possibility of manipulation of the controls by the MNOs. Hence, if theterminating network continues charging the donor network’s termination charge,the Director believes that there are two options available:

1. To include in the AIC29 all direct call minutes plus call minutes ported-in; or2. To include in the AIC all direct call minutes plus call minutes ported-out.

Option 1: including ported-in minutes

28 Source: Oftel data.29 The Average Interconnection Charge is the average charge set by the regulated MNO over eachcontrol period. The charge control proposed in this review requires that, at the end of each periodof the control, the AIC set by the regulated MNO over that period does not exceed the TAC (thecharge with which the operator is required to comply). (More details on the AIC and the TAC forthis control are contained in Annex G).

228

Under Option 1 the charge control on each MNO would include call minutes toported-in numbers whilst excluding call minutes to ported-out ones. To explainthis distinction, consider the treatment of calls ported in and out in operator A’scharge control. Calls to numbers that are ported-in are calls to customers whowere formerly with operator B, but who ported their numbers to operator A. Thetermination charges for such calls would be regulated under the control applying tooperator A (since the calls terminate on A’s network). Calls to numbers that areported-out are to customers who ported their numbers from operator A to operatorB. The termination charges for these calls would be excluded from A’s capbecause they do not terminate on A’s network (and are included in B’s chargecontrol).

Option 1 has the advantage that it would result in average termination charges thatbest reflected the terminating operator’s average costs. For a ported-in minute theapplicable termination charge (under current arrangements) is the donor operator'scharge, but the applicable cost is that of the recipient operator, because the call isterminated on the recipient's network.

For illustration, assume that there network A and network B have different costsand so different TACs: 5 for operator A and 6 for operator B. Assume for simplicitythat the weights for each operator reflect the receipt of 2 'direct' (i.e. non-ported)termination minutes and 1 ported-in minute. Assume that B sets its directtermination charge at 6. This will be received by A for its ported-in minute(ignoring the transit charge retained by B). Hence, in order to comply with theTAC of 5, operator A is forced to set a charge for its direct termination no higherthan 4.5.30 A's direct termination is therefore below cost, but this is offset by thereceipt for the termination of ported-in minutes, which is above cost. As aconsequence, its average charge is the same as its average cost (assumed equalto the TAC in this example).

Although under Option 1 charges would be cost reflective on average, this wouldnot be true at the disaggregated level. Since retail prices are set by originatingoperators taking account of the termination charges that they pay, prices for callsto each of ported and non-ported numbers would not be cost reflective. Thiswould distort the pricing signal for calls to all mobile numbers. Using theillustrative example, the retail price of calls to numbers ported in to operator Awould be too high, because they would reflect the termination charge of 6 whenthe cost of termination on the recipient network was only 5. The retail price of callsto A’s non-ported numbers would be too low, because they would reflect the directtermination charge of 4.5 (compared to the cost of termination of 5).31

30 4.5 x 67% + 6 x 33% = 5

31 The retail prices for calls terminating on operator B’s network would not be fully cost reflectiveeither, but the errors would be in the other direction, ie it would be the price of calls to numbersported in to B that would be too low (4.5 compared to cost of 6).

229

However, the main shortcoming of this option is that MNOs would have difficulty incomplying exactly with their control. This is due to two reasons. First, each MNOwould need to forecast accurately the charges that all other MNOs set fortermination during the forthcoming control year, since these would form part of itsown AIC via ported-in minutes. This is a demanding condition, since, even if theaverage charge is known, each MNO has flexibility on how to set its charges bytime of day and over the timing of charge changes. In addition, even theknowledge of other MNOs’ charges for the control year ahead would not besufficient. Each MNO, to satisfy its cap exactly, would have also need to know theweights in other MNOs’ caps32. To illustrate, in the example above A satisfies itscap exactly, but B does not. This is because B’s direct charge is 6 and its ported-incharge (set by A) is 4.5, hence its average charge is 5.5, which is 0.5 less than themaximum permitted (i.e. its TAC). There is only one set of charges that allowsboth MNOs to exactly satisfy their caps: if A charges 4 for its direct termination(less than its own cost of 5) and B charges 7 for its direct termination (more thanits own cost of 6).33

Option 2: including ported-out minutes

Under Option 2 the charge control on each MNO would include call minutes toported-out numbers whilst excluding call minutes to ported-in ones. Option 2implies that each MNO's average charge would not match its average cost and,thus, the MNO may have a windfall gain or loss (depending on whether the donornetworks from which it receives the ported-in minutes have higher or lowertermination charges). However, under this option the termination charges for callsto non-ported numbers would reflect the relevant costs so that retail prices for callsto non-ported numbers would provide the right price signal to callers. Terminationcharges and retail prices for calls to ported numbers would instead be too high ortoo low depending on whether the donor network’s termination charge was aboveor below the terminating network’s termination cost. This problem would not ariseonly if the TACs for all MNOs were the same or similar.

Conclusion

In the Director’s view, a variant of Option 1 would provide the most appropriatesolution to this issue. The Director proposes to include call minutes to ported-innumbers in the control on off-net termination charges, with the proviso that he mayconsent to a variation in the volumes used in the calculation of the AIC so as toexclude ported-in minutes. This consent shall be requested separately for eachperiod of the control and the Director may give his consent in one period and thendenied in the following year, if he is concerned that there might be manipulation ofthe charge control arrangements to set excessive termination charges for calls toported-in numbers.

32 The discussion above illustrates the operators’ argument that they have no control over the levelof the charge for ported-in minutes. This is a result of the way that they have agreed theinterconnect arrangements for terminating ported numbers.33 The illustrative examples are set out in Table 2 attached.

230

Table I.1: Illustrative examplesOperator A Operator BCharges Weights Charges Weights

TAC (assumed equal to average cost) 5 6Example 1Direct termination minutes 4.5 66.7% 6.0 66.7%Ported-in minutes 6.0 33.3% 4.5 33.3%AIC 5 5.5Amount AIC is above (below) TAC - (0.5)Example 2: both caps satisfied exactlyDirect termination minutes 4.0 66.7% 7.0 66.7%Ported-in minutes 7.0 33.3% 4.0 33.3%AIC 5 6Amount AIC is above (below) TAC - -

231

Annex J

List of questions

Chapter 3

Question 1Do respondents agree with the market definition? If not respondents shouldplease provide evidence that might support an alternative view

Question 2Will future developments in the mobile market, including changingconsumer behaviour, increase competition in voice call termination? Ifrespondents think they will evidence should be provided explaining how thiswill occur.

Chapter 4

Question 3Do respondents agree with Oftel’s analysis of the effect that CPP has on thecompetitiveness of the market for mobile voice call termination?

Question 4Do respondents agree with the criteria used in assessing SMP? If not, howwould they suggest developing the analysis?

Question 5Do respondents have any evidence of the existence of buyer powersufficient to constrain the level of mobile voice termination charges? Is thereevidence that such countervailing buyer power will develop in the nearfuture?

Question 6Do respondents agree that all MNOs have SMP in the provision of mobilevoice call termination over their networks? If not, please provide anyevidence to support a different conclusion.

Question 7Do respondents believe that future developments are likely to change thisanalysis in the near future? If so, please provide any evidence of how thiswill happen.

232

Chapter 6

Question 8Do you agree that Option A is the most appropriate and proportionateregulation of 3G mobile voice call termination services terminated on:• Vodafone’s 3G network?• T-Mobile’s 3G network?• O2’s 3G network?• Orange’s 3G network?

Question 9Do you agree that Option C is the most appropriate and proportionateregulation of:• Vodafone’s provision of mobile voice termination services terminated on

its 2G network?• T-Mobile’s provision of mobile voice termination services terminated on

its 2G network?• O2’s provision of mobile voice termination services terminated on its 2G

network?• Orange’s provision of mobile voice termination services terminated on its

2G network?

Question 10Do you have any comments on the justification or detail of the proposed

remedies on the MNOs?

Question 11Do you agree that Option AA is the most appropriate and proportionateresponse to the provision of 3G mobile voice call termination servicesterminated on 3’s 3G network?

Question 12Do you agree that Option BB is the most appropriate and proportionateresponse to 3’s provision of 2G voice call termination?

Question 13Do you have any comments on the justification or detail of the proposedremedies on 3?

Question 14Do you agree that Option BBB is the most appropriate and proportionateresponse to the provision of mobile voice call termination servicesterminated on Inquam’s network?

Question 15Do you have any comments on the justification or detail of the proposedremedy on Inquam?

233

Annex K

Glossary

Bottom-up approach (to modelling of costs) – the calculation of costs byidentifying and summarising the costs of the items of equipment, manpower andother resources required. Contrasts with top-down approach, which involvesremoving from a known total the costs which are not relevant to the activity inquestion.

Called party – a person who receives a call.

Calling party – a person who initiates a call.

Call origination – see originating operator.

Voice call termination – see terminating operator.

CC – Competition Commission (formerly Monopolies and Mergers Commission).

Common costs – Costs that are incurred in the supply of all or a group ofproducts or services provided by the company and that do not arise directly fromthe production of a single good or service.

Cost of capital – a firm’s cost of capital can be defined as the rate of return thatcould be earned in the capital market on securities of equivalent risk. In general,the higher the riskiness of the firm's activities, the higher its cost of capital, sinceinvestors typically require compensation for greater risk. For a firm financed bydebt and equity, the cost of capital will be a weighted average of its cost of capitalfrom both sources.

Current Cost Accounting (CCA) – an accounting convention, where assets arevalued and depreciated according to their current replacement cost whilstmaintaining the operating or financial capital of the business entity.

Equal Proportionate Mark Up (EPMU) – a means of recovering fixed andcommon costs through the addition of a mark-up on top of incremental costs. Thecosts to be recovered are allocated across a range of services so that eachservice is allocated the same mark up as a percentage of its incremental cost.

Fixed Network Operators (FNO) – operators providing fixed as opposed tomobile telephony services.

FAC (fully allocated costs) – a system of cost allocation based on HCA.

GPRS (General Packet Radio Service) – an extension to GSM standard to includepacket data services.

234

GSM – Global System for Mobile Communication. A European system for digitalmobile phones.

GSM 900/1800 MHz – GSM in the 900 and 1800 MHz frequency bands.

HCA – Historic Cost Accounting – a universally recognised accountingconvention. Costs, turnover, assets and liabilities are generally recorded at thevalue when the transaction was incurred and where assets are valued anddepreciated according to their cost at the time of purchase.

HSCSD (High Speed Circuit Switched Data) – an enhancement to currentmobile networks that allows several circuit switched channels to be combined fordata transmission, to a maximum of eight.

Internet – a global network of networks, mainly narrowband, accessed by userswith a computer and a modem via a service provider.

IP – Internet Protocol – packet data protocol used for routing and carriage ofmessages across the Internet.

Long Run Incremental Cost (LRIC) – the cost caused by the provision of adefined increment of output given that costs can, if necessary, be varied and thatsome level of output is already produced.

MMC – Monopolies and Mergers Commission. (now renamed the CompetitionCommission).

MNO – Mobile Network Operator, ie Vodafone, O2, Orange, T-Mobile or Inquam.

Mobile number portability – where a customer taking a service from a mobileoperator (eg Vodafone, Orange) or service provider (eg People’s Phone) canretain their telephone number when they change to a different mobile operator orservice provider.

MVNO – Mobile Virtual Network Operator – an organisation which provides mobiletelephony services to its customers, but does not have allocation of spectrum.

Network Externality – the effect on a third party when a person decides tobecome a new subscriber to a network which is not taken into account when thisdecision is made. In this case the third party values the calls that they make to andreceive from the new subscriber.

Off-net call – call from one mobile network which terminates on a different mobilenetwork.

On-net call – call from one mobile network terminating on the same mobilenetwork.

235

Originating network – the network to which a caller who makes a call is directlyconnected.

Originating operator – operator on whose network the call originates, ie theoperator to whom the customer subscribes.

Outpayments – these are the payments made by one network operator to anotherfor the purpose of conveying messages between the two systems.

PCN – Personal Communications Network – high capacity digital cellularnetworks (Orange and One2One are the current UK PCN operators).

PSTN – Public Switched Telephony Network – the telecommunicationsnetworks of the major operators, on which calls can be made to all customers ofthe PSTN.

PTO – Public Telecommunications Operator – network operators providingservices to the public with powers granted by the Secretary of State for Trade andIndustry under the Telecommunications Act 1984 to enable them to install theirsystems on public and private land, property, etc.

Ramsey Pricing – a method by which firms can recover fixed and common costswhich allows maximisation of economic welfare.

Second Generation (2G) – 2G means spectrum within the 880–915 MHz, 925–960 MHz, 1710–1785 MHz or 1805–1880 MHz bands.

Significant Market Power (SMP) – the SMP test is set out in various EuropeanDirectives, notably the Interconnection directives. It is used by the NationalRegulatory Authority (in the UK, Oftel) to identify those operators who must meetadditional obligations under the relevant directives.

SIM – Subscriber Identity Module. A small smart card type device that has detailsof the mobile subscriber including public telephone number and the numbersrequired by the network to recognise and authenticate the subscriber.

SMS (Short Messaging Service) – facility to send text messages of up to 160alphanumeric characters between compatible devices.

Stand alone costs – the costs to a single product firm of providing a service. Thestand-alone costs of a service exceed the incremental costs to a multi-product firmif there are economies of scope.

Terminating network – the network to which a customer who receives a call isdirectly connected.

Terminating operator – the operator on whose network the call terminates.

236

TETRA – Terrestrial Trunked Radio, a modern digital Private Mobile Radiotechnology.

Third Generation (3G) mobile systems – 3G mobile communications system willprovide an enhanced range of multimedia services (eg high speed Internetaccess). 3G networks are expected to enter service in 2003/04 using radiospectrum in the 2GHz bands.

UMTS – Universal Mobile Telecommunications System – so called 3rdgeneration mobile communications system which will provide enhanced range ofmultimedia services (eg video, high speed Internet access).

Voice over IP (VoIP) – the conveying of voice messages over Internet Protocol.