Royal Institute of Technology DEPARTMENT OF INFRASTRUCTURE Division of Building and Real Estate Economics Master of Science Thesis, # 175 ___________________________________________________________________________

SECONDARY MORTGAGE MARKET:

COMMERCIAL MORTGAGE- BACKED SECURITIES AS A CENTRAL SOURCE OF REAL ESTATE FINANCE- The Current Marketplace

Author: Frederick Delali Atutonu

Supervisor: Mats Wilhelmsson Co- Supervisor: Göran Råckle (Swedbank)

Stockholm, December 2002.

PREFACE This Study on CMBS as a central source of Real Estate Finance derives its momentum from my working experience in the bank, my interest in mortgages, and my inquisitiveness to study the question of the ability to reduce capital requirements, which is a hindrance to the development of the commercial real estate market, thus, calling for assessment of the current marketplace of CMBS. The study focuses on potential investors such as institutional / portfolio investors, Banks, etc. It presents an overview of the CMBS process with a sample deal to aid investors in making basic investment decision before embarking on CMBS investments. The study, however, is not intended to be a detailed process guide that outlines every task to be performed. My sincere thanks goes to the almighty God, my wife-Fortune, my son-Kirk, my parents, and my brothers for their love and encouragement, and also to my late sister (Stella Atutonu) who was my inspirer. My Supervisors deserve heartfelt thanks. First, to Mats Wilhelmson, Assistant Professor, for his guidance, interesting discussions, patience and constructive remarks. Second, to Göran Råckle, Head of Real Estate Department, ForenningsSparbanken (Swedbank), for making it possible for me (financially) to attend the Commercial Mortgage Securities Association (CMSA), Investors conference in Miami-Florida. I am also grateful to Sally Gordon, Senior Analyst- Moody’s Investors Services, New York; Erin Starford, Director- FITCH Rating Agency, Chicago; Dottie Cunningham, Chief Executive Officer-CMSA, New York; Gail Lee, Managing Director- Credit Suisse First Boston, New York; and Thomas F. Wratten, Principal Commercial Acceptance, LLC. The rest are: Pär Tysk, vice president- Handelsbanken Trading, Sweden; Per Sintring, corporate analyst- Handelsbanken Trading, Sweden. Finally, I express my sincere thanks to the department of Infrastructure and Real Estate Economics for supporting my trip to Miami for the CMSA conference. Thank You. Frederick D. Atutonu.

2

Master of Science Thesis Title: Secondary Mortgage Market: Commercial Mortgage-Backed Securities as a

Central Source of Real Estate Finance- The Current Marketplace. Author: Frederick Delali Atutonu Department: Department of Infrastructure Division of Building and Real Estate Economics University: Royal Institute of Technology (KTH)

Stockholm, Sweden. Report Reference: Master of Science Thesis, No. 175. Supervisor: Mats Wilhelmsson (KTH) Co-Supervisor: Göran Råckle (Swedbank) Key Words: Commercial, Mortgage, Market, Securitization, Securities, and Secondary. ___________________________________________________________________________ ABSTRACT This study is designed to serve investing institutions or investors who hold mortgage portfolios and/or for whom entrance into the Commercial Mortgage Backed Securities (CMBS) market is a possibility. The report is intended as a basic summary of some of the necessary steps and questions that investors and investing institution can address internally prior to embarking on a CMBS transaction. The study also proposes to aid investors including portfolio lenders towards investing in commercial mortgage-backed securities. The study was conducted at the Royal Institute of Technology (KTH), Stockholm, Sweden in consultation with three Swedish Banks- ForeningsSparbanken (SwedBank), Handelsbanken Trading, and SEB debt capital markets; and also participation in the Commercial Mortgage Securities Association, Investors conference held in Miami-Florida, January 2002. The study findings are based on analyses of interviews, literature review, and discussions with Professionals within the CMBS industry from the US, UK, Germany and Sweden. The results of the study present a generally encouraging picture on the CMBS innovation Scenario as truly a key source of real estate finance. The study informs that CMBS furnish conventional lenders with a broader array of real estate finance vehicles. Whether capital is needed for real estate acquisition or development the secondary market has allowed lenders to consolidate and sell their debt. Some benefits include increased liquidity, the ability to hedge against cyclical credit crunches, borrower and collateral diversification and the stabilization of commercial property. Results also show that The European CMBS has taken off and despite some added complexities in the market, growth is still expected to be dynamic over the next year. Analysis also shows strong potential for CMBS structures in the Nordic market. Countries including France, Belgium and Sweden are the likely new entrants soon.

3

TABLE OF CONTENTS

___________________________________________________________________________

PREFACE ABSTRACT

1 ARCHITECTURUE OF THE PROJECT WORK........................................................... 6

1.1 ADVENT OF THE NEW PARADIGM (PROBLEM DESCRIPTION) ..................... 6

1.2 WHY THE STUDY?......................................................................................................... 6

1.3(A) TARGET GROUP....................................................................................................... 7

1.3(B) CASE STUDY-FOCUS ............................................................................................... 7

1.4 THE STUDY ROAD MAP............................................................................................... 8

1.5 THE METHODOLOGY .................................................................................................. 9

1.6 OUTPUT ANTICIPATED (END PRODUCT OF STUDY) ......................................... 9

1.7 DISPOSITION (STRUCTURE OF THE REPORT)...................................................... 9

2 THE CMBS FOCUS OF THE STUDY............................................................................ 10 2.1 WHAT IS A CMBS IN THE US CONTEXT? ......................................................................... 10 2.2 A DESCRIPTION OF CMBS ................................................................................................ 10 2.3 A TYPICAL COMMERCIAL MORTGAGE............................................................................. 11 2.4 WHO ARE THE PLAYERS? .................................................................................................. 12 2.5 BUILD-A-BOND ................................................................................................................... 14 2.6 DEBT (LOAN) SERVICING AND ADVANCING FOR CMBS. ................................................. 15 3 THE COMMERCIAL MORTGAGE-BACKED SECURITIES (CMBS) PROCESS........ 17 3.1 WHERE THE MONEY GOES ............................................................................................... 17 3.2 TRANSACTION TIME TABLE .............................................................................................. 19 3.3 BASIC CMBS STRUCTURE, $100M, 5-YEAR, FIXED RATE ............................................... 20 3.4 WHY BUY CMBS ............................................................................................................... 21 3.4.1 RELATIVE VALUE INVESTING........................................................................................... 21 3.4.2 SATISFYING ASSET ALLOCATION TO REAL ESTATE DEBT................................................ 21 3.4.3 DIFFERENT KINDS OF RISK, PARTICULARLY COMPARED TO SINGLE- FAMILY MBS AND CORPORATE BONDS.................................................................................................................... 21 3.5 WHO BUYS CMBS ............................................................................................................. 22

4

3.5.1 WHO BUYS WHAT, BY RATING ......................................................................................... 22 3.5.2 WHO BUYS WHAT, BY STRUCTURE ................................................................................... 22 3.6 CMBS ISSUANCE ............................................................................................................... 23 4 DEFAULTS AND PREPAYMENTS IN THE PRICING OF CMBS................................. 27 4.1 Commercial Loan Evaluation............................................................................................. 27 4.1.1 Property Underwriting and Assessment variables .......................................................... 28 4.1.2 Determinants of CMBS subordination- Rating Agency Model Mechanics.................... 31 4.1.3 Declining Levels of CMBS Subordination, Implications for Investors. ......................... 34 4.2 Basics of Commercial Real Estate Collateral .................................................................... 36 4.3 CMBS Relative Value and Pricing Methodology .............................................................. 37 4.4 Impact of CMBS Prepayment Variables............................................................................ 39 4.5 Current CMBS Pricing Conventions.................................................................................. 41 4.5.1 A Swipe at Swaps............................................................................................................ 41 4.5.2 Theory ............................................................................................................................. 41 4.5.3 The Way Forward............................................................................................................ 42 4.5.4 What Should Be Done? ................................................................................................... 42 4.6 A Review of CMBS Defaults............................................................................................. 43 4.7 The Impact of Defaults on CMBS Transactions ................................................................ 44 5 EUROPEAN CMBS- THE MARKET TAKES OFF- IS SWEDEN PLAYING PART? ... 45 5.1 2001 Review....................................................................................................................... 45 5.2 2002 Outlook- Market Trend from Continental European Sources ................................... 46 5.3 The Role of Sweden ........................................................................................................... 46 6. SUMMARY AND CONCLUSION..................................................................................... 48 REFERENCES......................................................................................................................... 50 APPENDIX .............................................................................................................................. 53

5

CHAPTER 1 1. ARCHITECTURUE OF THE PROJECT WORK 1.1 Advent of the new Paradigm (problem Description) A key distinguishing factor in real estate markets is the inadequacy of real estate finance. This problem is even more acute with commercial real estate industry due to the lack of standardization in the commercial mortgage securitization process. Nonetheless, investment decisions require information about the profitability of investment opportunities, not forgetting the risks that those investments carry. The ability to reduce capital requirements, which is a hinderance to the development of the commercial real estate market, means that internal securitization of mortgages must be generated, developed and maintained. With some countries like Australia, Canada, England, Germany Italy and Japan, haven entered into the mortgage securitization process, signifies the potential that the commercial mortgage backed securities (CMBS) market holds for real estate capital requirements. Experiences from the US shows that due to innovations within the Secondary Mortgage and Securitization markets, it has become evident that the opportunity exists to create additional value in the commercial mortgage portfolios of traditional lenders and investors if efficient processes and standardization can be fully achieved. In order to attract and sustain foreign, direct investments, which is key to economic performance, there is the need to increase the output of performance measures. Against this background, the study is to investigate the current marketplace including the roles of default and prepayment in CMBS pricing, towards yielding positive returns, and make recommendations to aid investors. 1.2 Why the study? Aim The aim of this research is to study and evaluate the Commercial Mortgage-Backed Securitization (CMBS) process based on the experiences of the U. S. A., as a benchmark for institutional / portfolio investors. However, it is not intended to be a detailed process guide that outlines every task to be performed. Objective The CMBS process, which evolved from the U.S.A. as a key source for commercial real estate finance, has the potential of being responded to by investors from other countries. This research proposes to:

a) Analyse the conditions that necessitate the need for commercial mortgage-backed securitization process based on the U.S. model.

b) Evaluate and assess the importance of the CMBS with regards to the secondary mortgage market.

c) Evaluate the roles of default and prepayments in the pricing of CMBS.

6

1.3(a) Target Group Potential investors such as institutional / portfolio investors, Banks, etc. are perceived to be the livewire of any efficient economy. It therefore requires that research is carried out to constantly promote their ability to look for new possibilities, generate new ideas, think effectively, achieve effective networking and hence, make efficient investment decisions. Investors have been targeted for this thesis because they make very vital contributions to the driving of both national and Global Economies. 1.3(b) Case Study-Focus The study will be conducted based on the experiences of the U. S. market. The United States is selected because it represents a typical model country and presently has the largest, if not best, information source for commercial mortgage-backed securities. As a guide the study will propose to aid investors including portfolio lenders towards investing in commercial mortgage-backed securities. The choice of the study area would provide an excellent possibility for adaptation by other countries lacking the necessary capital requirements for the commercial real estate market.

7

1.4 The Study Road Map

Data collection, Surveys and observations

Interviews and Discussions

Serminars, observation and writing, additional literature studies.

Research findings, assessment and recommendations.

Setting the Ground: Literature Search, Contacting Actors, Attending Conference.

Fig. 1 : The study road map

8

1.5 The Methodology The study unfolds itself through the following activities: 1. Comprehension of the dynamics of the CMBS current marketplace as in the U.S. The process involves:

• Study and literature review; • Analysis of responses to discussions and interviews; • Validation of responses to discussions and personal interview with Directors and

Managers in Miami, Florida during the annual CMBS Investors Conference held on January, 2002;

• Overall assessment of CMBS performance through analyses of the current marketplace.

2. Examination of the roles and mandate of some CMBS investing institutions, rating agencies, servicing and origination institutions (including Banks). 3. Propose development of comprehensive and interactive guidelines on managing CMBS for the benefit of investors. 1.6 Output Anticipated (End product of study) The study when finished should reflect the current marketplace of commercial mortgage-backed securities. The result should provide a guideline for investors to access and invest in CMBS. The end product will be a master thesis with practical use. It will also serve as a source of information to aid investors including portfolio lenders in making CMBS investment decisions. Possibly, the ideas in the study could be transferred onto other countries with similar economic conditions. 1.7 Disposition (Structure of the report) This report embarks on an ambitious task of studying the current marketplace of CMBS, using the US as a study case. The first three chapters lay the ground work on the objective, scope, and the process of CMBS, with an introduction to the key participants, a timeline of the CMBS process, and an actual sample deal, which can serve as a “road map,” thus, clarifying the intricacies of a CMBS deal. Chapter four discusses the role of defaults and prepayments regards the pricing of CMBS, followed by the state of CMBS in Europe including the prospects for Sweden and the Nordic area. Chapter six concludes with a summary by the author wrapping up the research and findings.

9

CHAPTER 2 2. THE CMBS FOCUS OF THE STUDY 2.1 What is a CMBS in the US context? Thomas F. Wratten indicated in one of his papers that there has always been the need to create liquidity and secondary market in Commercial Real Estate (CRE) investments. This liquidity need gave birth to private sector capital market, and Government Agent initiatives, which brought about instruments (investments) like Mortgage Backed Bonds, Real Estate Investment Trusts (REITs), Industrial Revenue Bonds, and Mortgage participations that followed the REITs. Multifamily and small business lending also became part of the capital market. A motive of these efforts is to allow small investors participate in more extensive commercial Real Estate investment activities. Today, the evolution of commercial real estate finance from its illiquid age gave way to the most developed format for liquidity called Portfolio Securitization, which is but one segment of the emerging secondary marketplaces. This securitization process gave birth to the Commercial Mortgage Backed Securities (CMBS) innovation. As compared to single-family securities market, the evolution of commercial real estate securities’ market is slow due to its complex nature, and this took about 8years to reach efficiency, according a survey by Salomon Smith Barney, New York. My view from the just ended Commercial Mortgage Backed Securities Association (CMSA) conference in Miami shows that the current process of securitization with time, will serve as the basis for which CMBS will develop to the fullest. 2.2 A description of CMBS A commercial mortgage backed security (CMBS), which may take different structures and legal forms, is a bond backed by cash flow of mortgages or pool of mortgages on commercial real estate. There is quite a similarity between the structure of a CMBS (not underlying loans) and that of its single-family, residential counterparts. On a CMBS transaction, principal and interest payments on underlying loans are passed through to certificate holders, after deduction of servicing expenses, in a sequential payment order. The CMBS bond’s payment stream depends on the cash flow produced by the mortgage pool and how the mortgages are refinanced at their balloon maturity. Contrary to an agency residential security mortgage, CMBS have no built-in implied Government guarantee for payment of principal and interest to the investor thus, making them similar to non-agency residential mortgage securities. Interest Only (IO) CMBS, which are bundled and sold from interest payments, have a likeness to residential securities. High prepayment penalty (or defeasance mechanism) imposed on CMBS, and for the fact that underlying mortgages are usually locked out from prepayment for a period of the loan term, a

10

cash flow of CMBS IO certificate is more certain than its residential IO counterparts. To explain further the CMBS by comparing it to residential securities, in terms of recourse to the borrower, they have no recourse to the borrower unlike residential securities, which does. CMBS are usually non- callable for its life. Prepayments usually allowed through defeasance (prepayment penalty) whilst residential securities are pre-payable at par without penalty. CMBS unlike residential securities, are secured by income-producing assets (office buildings, multi-family apartments, hotel, healthcare, retail property, mobile home parks MHP, industrial). They also have a characteristic bullet structure. i.e. Typically ten year balloon payments with a 25 to 30 year amortization schedule basis, whereas residential securities are secured by single-family residential properties with a fully amortizing structure- typically 15 or 30 years. 2.3 A typical Commercial Mortgage Commercial mortgages are in numerous ways different from single-family residential loans. Commercial mortgages are secured by income-producing properties, such as those properties mentioned earlier. Mortgages on these properties have shorter stated maturities than their residential counterparts. Commercial Mortgages are larger sized loans that might either be non-amortizing or partially amortizing, and thus, has significant principal balance (“Balloon”) on maturity. They may also be subject to greater environmental risk (particularly, industrial and/or warehouse) than single-family properties. A mortgage loan may offer prepayment flexibility through three methods tailored to maintain or to compensate the collateral pools or the investors for lost payment respectively:

a) Yield Maintenance- clears or make the lender whole for the loss of an above coupon on a net present value basis. The effect being that there is generally no economic incentive for the borrower to refinance;

b) Defeasance- makes the borrower pledge to the holder of the mortgage, Treasury Securities in the case of the US, whose cash flows equal or exceeds that of the mortgage;

c) Seldom, a Declining Fee proportional to the remaining balance say, a “6-5-4-3-2-1” schedule refers to a penalty of 6% of the outstanding loan balance in the beginning year of the penalty period, 5% during the second year etc.

An owner may voluntarily prepay and incur the yield maintenance or defeasance cost should they have a compelling necessity to sell. The reason for these prepayment restrictions was to create securities with low loan prepayment expectations thus, maximizing cash flow certainty. A study by Solomon Smith Barney shows that during 1998 virtually all (greater than 99%) loan issuance were balloon payment due after 10years with all the loans having prepayment restrictions. It was also found out that since 1998, most CMBS prepayment restrictions have relied much on defeasance with recent CMBS pools having loans locked out until three months prior to maturity, permitting only mortgage prepayments through defeasance. Prepayment restrictions usually end about three to six months prior to the balloon date (free or open period). The aim is to allow the borrower time to refinance the loan and, hence, make the balloon payment.

11

Commercial mortgages usually require the borrower to fund an escrow account with one month’s payment of dept service, as well as ongoing reserves for real estate taxes and property insurance. 2.4 Who are the players? -- Securities

Trustee

Borrower Lender Issuer Trust Investor

Trading

Appraisal Financial Statement

Engineering Report

Master Servicer/Special servicer

Fig. 2: Shows who the players are. Trustee The trustee represents the trust. i.e. any person or entity that holds the legal title to and manages the collateral for the benefit of all tranches (class holders) of certificates. This may be a trust Accountant, Investment Bank, Servicer or Rating Agency etc. Some of the trustee’s responsibilities include: Acting as supervisor to the Master Servicer and Special Servicer; Holding the mortgage collateral; Making sure that the Servicers act according to the Pooling and Servicing Agreement; Passing all collected funds by the Master

12

Servicer to the certificate holders; Appointing a new Servicer in the event of violating terms of the Pooling and Servicing Agreement. Master Servicer For the benefit of and on behalf of bond (certificate) holders, the Master Servicer is required to service mortgage loans, which collateralise the CMBS. Some basic responsibilities are: Collecting and transferring mortgage payments to the trustee; Advancing late payments (if any) to the trustee; Passing on-to the Special Servicer non-performing loans; And providing certificate holders with Mortgage Performance Reports. Special Servicer A Special Sevicer usually is responsible for delinquency problems (loan payments past at least, 30days due for payment), default loans and its foreclosure process (“work-out”). Conditions determining this category of Servicers include: Issuers and sellers who retain the First Loss piece (unrated tranche); Investing in B piece in return for Special Servicers’ rights; And appointing as result of asset management expertise. Borrower These are individuals and institutions that borrow money from the lender. In other words, they receive funds in the form of loans with an obligation to repay principal with interest. Lender This may be a Commercial Bank or Mortgage Bank that lends money to the borrower. Issuer As the mortgages are transferred from the mortgage banker to the issuer, e.g., Investment bank, then a true sale has occurred. The issuer then subjects the mortgages to a structuring process through appraisal, issuing financial and engineering reports. The process shows various experiments of combinations (pools) of mortgages and security classes, and coming out with the highest price for a CMBS based upon capital market forces. Trust This is the legal mechanism through which any person or entity, for example, Trust Accountant, Investment Bank etc., administers the assets and their use to the benefit of certificate holders. Investors

13

Investors such as Life Insurance Companies, Commercial Banks, Pension Funds are mostly the largest investors in the investment grade bonds (tranches). In the case of below investment grade tranches, Rule 144(a) 2 guides the placement of securities, and as such, are required to be sold to Qualified Institutional Buyers (QIB). For this category, the largest investment comes from prominent Real Estate Investment Funds and CMBS servicing companies. This is because they possess the real estate sophisticated tools needed to aid them properly underwrite the real estate risk hidden in this tranches. 2.5 Build-a-bond Collateral Pool New Securities Last Lowest Loss Risk Loss Credit Position Risk First Highest Loss Risk

$170m Credit Inv. Grade Support CMBS. Level. Aaa/AAA 20%-32%Aa/AA 18%-24%A/A 14%-19%Baa2/BBB 10%-14%Baa3/BBB- 7%-12% $22m. Non IG1 CMBS Ba/BB 4%-7% B/B 2%-4% $8m Unrated CMBS

$200m Pool of Mortgages

Fig.3: A hypothetical structure: CMBS credit tranching with subordination levels. As can be seen from figure 3 above, commercial mortgage securities are structured with credit support (enhancement) levels to protect against shortfalls in cash flows. This credit enhancement usually is in the form of internal credit support through allocation of loan losses in reverse sequential order as shown above. ____________________ 1 IG- refers to Investment Grade. 2 See CMBS Pooling and Servicing Agreement- the Agreement that makes how each CMBS transaction cash flow is handled.

14

The triple-A class usually receives all loan prepayments including recoveries first. The credit support levels above show that for example, at the triple-A level, CMBS transactions have credit enhancement levels between 20%-32% (i.e. 80%-68% of the transaction was rated triple-A). Credit enhancement levels will be revisited in a later chapter. 2.6 Debt (loan) Servicing and Advancing for CMBS. A Master Servicer (required to service to mortgage loans collaterallizing a CMBS on behalf of and for the benefit of the certificate holder) collects and aggregates monthly debt service payments for CMBS, and remits on a stated monthly date, payments to the trustee who then makes monthly payments to certificate holders. Certificate holders are protected from possible short-term cash flow shortfalls by the Master Servicer, who must make bond Advances (principal and interest) to the trustee and pay property taxes and insurance payments so that such advances are recoverable from the underlying mortgage obligation. The Special Servicer, a separate entity from the Master Servicer evaluates whether to foreclose and liquidate the loan or to restructure it to enable its return to the Master Servicer. During what is called Appraisal Reduction, certain events may lead to a new appraisal. Usually if it so happens that the property’s value becomes less than 90% of the outstanding loan balance and servicer advances, the Special Servicer must reset the principal loan amount to the new appraisal value. The purpose of this Appraisal Reduction is to preserve capital for the senior rated certificates, as it might be unnecessary to confine advances to junior classes that seize to have economic interest in the loan. The Agreement that makes how each CMBS transaction’s cash flow is handled, i.e. every element of converting the mortgages’ cash flow to bond cash flow very mechanical is contained in a Pooling and Servicing Agreement between the Issuer, Servicer and Trustee.

15

Fig. 4: Senior/ Subordination Structure: 5-year Security First Next After 3.6yrs 1.4yrs 5.0yrs P + I I P + I I I P

A

B C

A

B

C

A

B

C This structure is sometimes referred to as the “waterfall” structure. It shows an example of 5year Fixed Rate Security with principal sequential payments, falling according to each class. During the first 3.6 years of this security, a fixed rate of principal plus interest is paid. During the next 1.4 years, another fixed rate (part principal and rest of interest) is paid, leaving rest of principal to be paid or refinanced at the end of five years. Principal and interest payments on underlying loans, after the deduction of servicing expenses, are passed through directly to the certificate holders in sequential payment order. Excess interest payments are also bundled and sold as interest only (IO) certificates, which has more certain cash flows compared to residential IOs. This is because the underlying commercial mortgages are usually locked out from prepayment for a portion of the loan term, and then the loans often have a high prepayment penalty or a defeasance mechanism. The bond’s payment stream depends on the cash flows produced by the mortgage pool and the possibility of refinancing the mortgages at their balloon maturity. CHAPTER 3.

16

3 THE COMMERCIAL MORTGAGE-BACKED SECURITIES (CMBS) PROCESS. 3.1 Where the Money Goes The preceding chapter mentioned the main players, including Trustee, servicer, Borrower, Mortgage Banker (Lender), Issuer, the Trust and Investors who are involved in the process of Commercial Mortgage Backed Security. The figure below shows us where, as result of the activities of these actors, the money goes. Assignment of Mortgage and Leases

Mortgage Loans Collateral Issuance Proceeds Debt Service + Debt Advances – Service Servicing Fees Debt Service + Draws on External Credit

Enhancement Issuance Proceeds

Borrower

Servicer

Collection Account

Lender

Trustee -Distribution

Account

Investors

Securities

Fig. 5: Where the Money goes.

17

Commercial mortgage loans that serve as the underlying collateral for the commercial mortgage backed securities may come from traditional portfolio holders who originate their own portfolio or from conduits who originate mortgages solely for securitisation (these loans may or may not be for the expectation of being securitised). The underwriting of such a loan assesses, among other things, real estate risk, existing lease terms, property condition, borrower quality, environmental liability to determine whether to originate a loan on the subject property, and tenant quality. The underwriter will then analyze, based on the anticipated cash flow whether the two main indicators, the debt service coverage ratio (DSCR) and the loan-to-value (LVR) are prudent and acceptable. This is because it is the quality of the underlying loans that will determine the necessary credit enhancement and the ultimate pricing of the certificates. These loans, after being made to the borrower are later serviced into a Servicer collection account. The master servicer has the responsibility of servicing this mortgage loan. He collects the mortgage payments and passes the funds onto the trustee. He also advance any late payments, among others, to the trustee. The fund is transferred to the trustee distribution account after considering all advances of principal and interest less servicing fees. On the trustee distribution account, which is simultaneous with closing-the period where all legal documents, such as the Pooling and Servicing Agreement and all necessary opinions are finalized, certificates of beneficial ownership is issued after receiving the requisite money from the portfolio holder or servicer. This certificates of beneficial ownership, which are the securities, are undivided interests in the trust which owns the collateral pool.

18

3.2 Transaction Time Table The transaction timetable below shows the duration at each stage of a CMBS deal, from start to end of the deal, with all the various stages of the deal and also key actors in it.

TRANSACTION TIMETABLE WEEK ACTIVITY 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 ResponsibilityInitial Analysis SL,IB Due Diligence Phase SL,IB Structuring Process SL,IB Rating Agency Review SL,IB,UC,SC Selection of Servicer & Trustee SL,IB Legal Documentation Private Offering IB,RA,TA,SV Public Offering IB,RA,TA,SV Pre-marketing of securities IB Marketing/Pricing Private Offering IB,CA,TA,UC Public Offering IB,CA,TA,UC Closing Private Offering ALL Public Offering ALL Key Participants SL Seller SV Servicer IB Investment Bank CA Certified Accountant RA Rating Agency UC Underwriter's Counsel TA Trust Accountant SC Seller's Counsel

Fig. 6 Transaction timetable

19

3.3 Basic CMBS structure, $100m, 5-year, Fixed Rate Class Size Rating Coupon Expected Subordination Life Class A $85m Aa2/AA 7.05% 3.6years 15%a

Class B $11m Ba2/BB 8.20% 5.0years 04%b Class C $4m Not Rated “First Loss Piece”

a Aa2- level stress could be calculated as follows: Foreclosure Frequency X Loss Severity = Loss Coverage 30% X 50% = 0.3 X 0.5 = 0.15 or 15% subordination b Ba2- level stress could be calculated as follows: Foreclosure Frequency X Loss Severity =

20% X 20% = 0.2 X 0.2 = 0.04 or 4% subordination These results show that in the event of loss, all classes below Aa2 will first absorb 15% of the loss before any loss is attributed to it. Also, all classes below Ba2 will first absorb 4% of the loss. It also means that these subordination levels or loss coverage is a protection against foreclosure.

20

3.4 Why buy CMBS 3.4.1 Relative Value Investing There is yield advantage in terms of relative value investing. For instance, it is cheaper to hold bonds than mortgages even at comparable yields. Whereas liquidity is good both in issue and in yield, it also has information premium built into the yield since default and prepayment information are especially lacking. 3.4.2 Satisfying Asset Allocation to Real Estate Debt Cost of Management in most cases is low and there is liquidity. Also, as mentioned by Sally Gordon, a Senior Analyst and Director at Moody’s Investors Services during the 2002 Commercial Mortgage Backed Securities Association (CMSA) Annual Conference held in Miami, Florida, the Risk Based Capital requirement for some mortgages is 3% while the requirement for some bonds is 0.3%. For some investors therefore, it is cheaper to hold bonds rather than mortgages, even at a comparable yield. 3.4.3 Different kinds of risk, particularly compared to single- family MBS and corporate bonds Comparing CMBS different risk profiles to Single Family MBS and Corporate bonds in particular, I found out that for CMBS there is more attention to credit risk than prepayment risk. This prepayments are mitigated by call protection in the form of lock outs. For Single Family MBS however, there is more attention to prepayment risk than to credit risk. They have an embedded call option since mortgagor can prepay at par. For Corporate bonds more attention is to credit risk but this risk is driven by different variables than CMBS. CMBS have extension risk, for instance, an attractive spread over a five (5)- year curve may no longer be as appealing when compared to six (6) or seven (7)- year time frame if security extends. However, there is no extension risk in the cases of single family MBS and corporate bonds. As a more useful predictor for Default Risk, Debt Service Coverage Ratio (DSCR), Loan to Value ratio (LTV), and Credit Analysis are the more useful predictors to CMBS, Single Family, and Corporate bonds respectively. Default risk is mitigated in the case of CMBS by various forms of credit enhancement vehicles whereas in single family MBS it is mitigated by government (quasi) guarantees in agency securities. Information available to holders of CMBS public tranches is only a fraction of that available on corporate bonds. For Single Family MBS and corporate bonds, information is widely disseminated. Thus, there is information premium built into CMBS. Whereas Rating Agencies are still developing for CMBS, expertise and staff experience, they have long and rich experience with single family MBS and corporate bonds.

21

3.5 Who Buys CMBS CMBS may be in various forms and placements ranging from public and private placements to fixed and floating rate and to rated and unrated subordinates. These attract investors as Insurance Companies, Money Managers, Banks, Thrifts, Corporations, Pension Funds, Investment Advisors, Municipalities, and Opportunistic Funds among others. 3.5.1 Who buys what, by Rating Aaa/AAA to Aa/AA class is particularly attractive to pension funds that only are permitted often to buy the highest-rated class of any offering (even if the highest rated class is A). A/A This class is often a small part of a total structure. Usually when there is Aa and Aaa above there are fewer buyers for this class. Insurers and money managers are in this category since they like extra yield. Baa/BBB is often appealing to insurance companies and other investors. If class is upgraded to Aa, such investors might sell because they often see little value in this intermediate area, meaning that they want Aaa for some portion of the portfolio and Baa for another. Ba/BB and B (below-.investment grade) are less suitable for any institution that has risk-based capital reserve requirements (e.g., insurance companies, Banks), because the cost of holding these securities is large. This class is better for higher-risk, longer-term bond funds that can wait for an upgrade. It is often appealing to firms that have good knowledge of and experience with real estate thus, capitalizing on the inefficiencies of pricing these classes. The non-rated class is suitable for private placements, which capitalize on an information arbitrage. There is typically confidential information that cannot be revealed to public buyers, but control can be exercised in sharing information with private buyers through non-disclosure agreements3. Buyers of first-loss pieces must know more about the underlying assets than can be publicly shared without breach of confidentiality, and must have that information earlier in the process in order to adequately review it, which would otherwise violate the disclosure rules of a public offering. Apart from some funds specializing in first-loss pieces, it is often taken over or purchased by special or master servicers. 3.5.2 Who buys what, by Structure Principal Only (PO) strips may be purchased by Mortgage Bankers to protect the value of servicing income against high prepayment rates. In a declining rate environment, POs provide higher total returns compared with other discount mortgage securities (e.g., discount corporate bonds). Investors in Discount Corporate Bonds might thus find POs to be attractive alternative investments. _________________ 3This was known through a discussion I had with Gail Lee, Managing Director, Credit Suisse First Boston at the January, 2002 CMSA conference in Miami.

22

Interest Only (IO) strips may be purchased by savings institutions with long-term mortgage loans to offset their duration risk. IOs are one of the few fixed income instruments that generally rise when market interest rates increase. It is less attractive to short-term investors but better for long-term and high yield investors who want yield and ignore intermediate volatility. Fixed Rate classes, which has a payment structure that do not change over the life of the loan are attracted to any bond buyer looking for longer duration. Floating Rate classes, which could have change in payment periods are more suitable for financial institutions such as Banks that have shorter-term, in most cases 5 years, liabilities that are also floating rate obligations. 3.6 CMBS Issuance Today’s CMBS market is supported by a broad array of institutional investors that hold CMBS to diversify their bond portfolios. New institutional investors continue to recognize the benefit of the underlying asset diversity and credit enhancement structure. This wider investor base has combined with strong underlying collateral fundamentals to create an asset class that has become a capital safe haven in times of economic distress. The market has evolved to the point where several hundred million CMBS certificates can be bought and sold in minutes. The liquidity in triple-A CMBS is similar to that of agencies and less like that of many other structured debt products. Much of the liquidity has been driven by the rating agency diversity guidelines and B-piece investor reviews, which have formed homogeneous pools and attracted more investors to CMBS. If the secondary CMBS market continues to experience the strong secondary market that has existed since 1999, I beleive that CMBS spreads can tighten further relative to the liquid agency debt product.

23

Fig. 7 CMBS Issuance The graph above shows that for the period between 1990 and 2001, there had been, on the average improvement in the issuance of both domestic and non-domestic US CMBS. This means then that investors are embracing CMBS.

24

Fig. 8 CMBS Issuance and Interest Rates

25

The graph above shows the relationship between CMBS issuance and Interest Rates. It confirms that there exists a relationship. It also depicts that on the average, when interest rates fall, CMBS issuance rises and vice versa.

26

CHAPTER 4 4 DEFAULTS AND PREPAYMENTS IN THE PRICING OF CMBS 4.1 Commercial Loan Evaluation Separate discussions I had at the January 2002 CMSA annual conference in Miami with Gail Lee, Director of Credit Suisse First Boston and chairperson of CMSA Research Committee; Sally Gordon, Vice President and Senior Analyst at Moody’s Investors Services (rating agency), and also Erin Stafford, Associate Director of FITCH rating agency and Co-Chair of CMSA Investor Reporting & Securitization committee revealed that the rating agencies have developed analytical models that take much of the investigative onus off the investor. After extensive analysis, the agencies provide investors with a consistent measurement of the likelihood that cash flow from underlying mortgages will be sufficient to meet scheduled payments of principal and interest on the security at each certificate rating. At higher rating levels, mortgage cash flows are expected to hold up under increasingly severe economic conditions. Thus, for a given pool, progressively greater credit support is needed at higher rating levels. When the rating agencies initially developed their models the real estate market was emerging from a severe recession, the commercial mortgage market suffered from a lack of standardization, and there was a shortage of historical delinquency and foreclosure data. As a result, the rating agencies developed subordination models based on very conservative default and loss assumptions. The process used to rate commercial mortgage securities is, in some respects, similar to the residential mortgage rating process: the primary focus of the analysis in each case is on the credit quality of the underlying collateral; however, significant differences exist as has been mentioned earlier. Commercial mortgages are nonrecourse loans, so securities are backed only by income-producing properties whose profit-oriented owners’ motivation depends on different factors than those affecting a residential homeowner. Of primary concern in analyzing a commercial mortgage is the property’s underlying net operating income (NOI) and net cash flow (NCF), and all factors that influence them. Therefore, an analysis of the credit quality of a commercial loan requires a careful review of underlying tenants’ creditworthiness, the structure and term of underlying leases, the historical level of vacancies and rents on the property and on other properties in the region, and the real estate and economic cycle within the local market. All of this analysis feeds into a review of the mortgage’s loan-to-value ratio (LTV) and debt service coverage ratio (DSCR). In reviewing a property’s cash flow, a rating agency will typically reduce the property’s cash flow to the lower of the rental market or the contractual lease rent. A higher contractual rent may be accepted as cash flow if the underlying tenant is publicly rated and the term of the lease extends beyond the term of the mortgage loan. The property’s occupancy may also be adjusted downward to reflect surrounding market conditions on the assumption that even a fully occupied building will eventually lose tenants and stabilize at the market occupancy rate. The property’s operating costs are assessed at historic levels adjusted for inflation with any lower-than-market costs increased to reflect a third party’s cost in anticipation of the possibility the mortgagee may have to foreclose and operate the property. The NOI, which is derived from revenues less operating expenses, is then further adjusted by estimated annual market releasing commission and tenant inducement costs that reflect the building’s scheduled lease expiries and releasing costs for similar buildings. Finally, the agency

27

subtracts a capital expenditure allowance in anticipation of ongoing structural repair costs that will not be reimbursable from tenants. The rating agencies’ resulting NCF is usually 3%-5% less than the underwriter’s figure (indicated by Sally and Erin from Moody’s and FITCH respectively), but reflects an objective opinion of the property’s likely ongoing cash flow in a stressed environment. 4.1.1 Property Underwriting and Assessment variables Prior to reviewing specific rating agency methodology, we briefly discuss the loan underwriting variables considered in an analysis of commercial mortgage securities: ♥ Debt service coverage ratio ♥ Loan-to-value ratio ♥ Mortgage payment structure ♥ Mortgage amortization/term ♥ Tenancy ♥ Property location DSCR is calculated as follows: 1 The current NOI on a property divided by the actual mortgage debt service obligation.

This is the NOI DSCR, which is usually provided by the underwriter and reported on an ongoing basis by the servicer.

2 The estimated cash flow (after leasing costs and capital expenditures not recoverable

from the tenants) divided by the actual mortgage debt service obligation.

3 The cash flow after leasing cost and capital expenditures not recoverable from the tenants divided by a stressed refinance constant (a debt constant that is higher than the actual rate to account for potential unfavorable refinance conditions when the balloon comes due). This is referred to as a “stressed DSCR” and is used by rating agencies to perform their analysis. LTV refers to the ratio of the outstanding loan balance on a property to the estimated valuation of that property. LTV is assessed as an important indicator of potential Loss severity, because the analysis of how much will be lost upon default relies heavily on the property value relative to the outstanding debt on that property. State foreclosure laws can also affect loss severity. If a state’s foreclosure laws could delay the mortgage servicer’s property liquidation, the agency may toughen the LTV parameters for that state. The rating agencies treat DSCR as the best indicator of default probability, while LTV is viewed as the best indicator of loss severity.

28

Similar to the residential market, floating-rate structures in the commercial market inability to meet undefined or floating debt service payments in the future increases the risk that the borrower may default. The presence of a floating-rate lessens somewhat the reliance of the analysis on a DSCR. To compensate, rating agencies usually assess the loan's DSCR at a higher debt constant, creating a lower DSCR that requires higher credit enhancement levels. If the borrower has the benefit of an interest rate cap agreement, the rating agency may recognize the contract’s benefit by using the contract interest rate to calculate DSCR. Furthermore, buildings that are in a value enhancement stage, intensifying the default, secure many floating-rate commercial mortgages and refinance risks because cash flow may never stabilize. Overall, floating-rate mortgages are viewed as more risky by the agencies and carry significantly more credit enhancement. Whereas most residential mortgages fully amortize over a 15- or 30-year life, many domestic commercial mortgage structures either are partially amortizing or do not amortize at all (interest only). Most carry shorter terms than residential mortgages, and a large majority mature with an outstanding balloon payment. From a credit perspective, fully amortizing mortgages are the least risky. As a loan amortizes, principal is paid down, reducing the indebtedness of the borrower and reducing the risk of default (equity buildup is an important protector against default in the commercial market, as well as minimizing loss severity). A partially amortizing loan will have paid down only part of its principal by maturity, and a non-amortizing loan will have paid only interest and no principal by maturity, translating into little or no equity buildup. Partial or no-amortization loans will mature with a significant remaining principal balance — a “balloon payment” — that must be made. For example, after ten years of amortization a loan has usually only paid 15% of the principal, if the loan is on a 30-year amortization schedule. Typically, the borrower will try to refinance that balance into a new mortgage, creating refinancing risk in the credit equation. Various factors may impair the borrower’s ability to refinance the remaining principal at maturity:

1 An increase in interest rates 2 A decline in property value 3 More restrictive underwriting criteria

4 A relatively large number of vacancies 5 Tight credit conditions

All, or any of these factors, may inhibit lenders in the primary market from providing refinancing on the mortgage. Given this potential risk, the rating agencies focus on the coupon rate of the mortgage and analyze the ability of the property to support a new mortgage in the amount of the balloon at a potentially higher interest rate level. Moody’s and Fitch use a maximum refinance interest constant for the collateral, which is typically 9.25% for Moody’s and 8.88% to 11.33% for Fitch, depending on the property type and loan amortization to the balloon.

29

These refinance constants are based on historic rates and reflect a worst-case DSCR. The resulting stressed DSCR is usually 0.10-0.15 times lower than the underwriter’s DSCR. For reference, a typical CMBS transaction may have a NOI DSCR of 1.35- 1.45, a CF DSCR of 1.20-1.40, and a “stressed DSCR” by an agency such as Fitch of 1.05-1.20. The rating agencies also look at the expiration dates for major leases on the property relative to the maturity date of the mortgages to help determine vacancy potential and, thus, the releasing risk exposure that will be evident on the balloon date. If many leases expire near the loan maturity date, the loan will usually be structured to accrue a sufficient releasing reserve to mitigate the risk. To the extent the rating agencies perceive lease rollover risk, they will increase their probability of default assessment and sometimes decrease their specific estimated loan recovery, leading to higher subordination levels. The rating agencies also take into account the structure of the loans to evaluate the level of credit support necessary to protect against balloon risk. In many cases, the loans may have an early permitted repayment date of ten years, after which the interest rate steps up and all cash flows are captured in a lockbox controlled by the lender to amortize principal (an anticipated repayment date (ARD) loan). The ARD loan structure provides a built-in loan restructuring, because after the anticipated loan repayment date, the lender receives all the property cash flow without having to foreclose. The rating agencies also evaluate the flexibility of the servicer in situations where refinancing risk is present. The question of whether a workout will be attempted is an important one — in terms of whether a default will occur, and if it does, what the loss severity will be. Traditionally, commercial lenders have maximized recoveries when afforded flexibility in these loan workout situations. In some cases, they have extended the term of the mortgage; in other cases, they have reduced the monthly payments on the mortgage. While the servicer’s ability to extend loans may lead to a maximum loan recovery, it also affects the certificate’s average life and has an impact on the investor’s yield if they paid a premium or a discount for their bonds. Sometimes a forced foreclosure and cash sale of property will result in much higher losses than a workout negotiation whereby the current borrower is allowed to continue under new terms. In addition to a potentially low sales price, the expenses incurred in a foreclosure are generally higher than those associated with a lender/borrower cooperative loan workout, particularly if the borrower uses bankruptcy as a delaying method. Even in cases where the underlying property does not support the outstanding loan balance, a reduction of the debt claim and avoidance of the above-mentioned expenses may result in higher recovery levels. Therefore, the ability of the servicer to manage a workout is a critical factor in minimizing potential loan losses. Such ability will be a function of servicer experience, knowledge of the local markets and of the pertinent property type, and finally, workout volume capacity. As previously mentioned, most recent CMBS transactions give the special servicer the ability to modify loans subject to a standard of care to maximize expected loan net present value. Therefore, rating agencies assess the special servicer’s abilities and may adjust the credit enhancement levels by 0% to 1% based on the evaluation. Finally, the viability of an income-producing property depends on its tenants and the property location’s ability to attract replacement tenants. The loss of a major tenant can cause insolvency. Most recent commercial mortgage defaults have been caused by either lease

30

expiries or leases not being affirmed by a bankrupt entity1. Leases rolling over to lower market rents can create a lower DSCR and lead to defaults. Thus, rating agency models favor multi-tenanted buildings and, in addition, will look to the creditworthiness of the individual tenants, particularly the anchor tenant in a transaction backed by mortgages on shopping centers. Single-tenanted buildings are usually penalized with higher subordination levels to ensure that there is sufficient loss recovery should the individual tenant vacate and cause a default and loan liquidation. Tenant concentration may be somewhat offset by a strong property location and the resulting ability to attract new replacement tenants at low re-tenanting costs. 4.1.2 Determinants of CMBS subordination- Rating Agency Model Mechanics The number of mortgages in the pool and the relative size of the larger loans as a percentage of the pool are major determinants of agency subordination. This is the key reason that CMBS pools are classified by loan size. Smaller pools and pools with loan or borrower concentrations merit closer scrutiny, because they represent greater risk and, thus, require greater credit support. As loan sizes increase, the agencies’ loan size analysis can create an “add-on” to subordination of 0% to 4% for a diverse pool, 3% to 10% for a fusion pool and anywhere from 10 to 20% for a large loan pool or single asset transaction. The increase in credit enhancement generally steps up to ensure the default of a single loan concentration will not affect the rating of the pool. This lumpiness requires subordination approach, and ensures that in the event of a random default of a large loan, the loan recovery will be sufficient to pay down the upper-rated classes and maintain the original relative subordination and credit rating. On more diverse pools the rating agencies rely less on the pool’s ability to withstand an individual loan’s default and focus more on the pool’s anticipated cumulative lifetime defaults being similar to the experience in the early-1990s default studies4. Thus, the loan pool’s size diversity decides whether the rating agencies use a property-specific approach (with the ratings based on each property’s characteristics) or an actuarial analysis. The actuarial approach relies on an evaluation of credit characteristics from a sample of the overall pool. An analysis of pool credit quality through aggregate loan characteristics generally will be sufficient if the pool was originated with uniform underwriting standards, contains a sufficiently large number of loans, and the distribution of the loan balance is not widely skewed. To the extent any of these conditions is lacking, the rating agency may rely more on a loan-by-loan analysis. _____________ 4 Update to commercial mortgage defaults, Mark P. Snyderman, The Real Estate Finance Journal, Vol. 10, No.1, Summer 1994.

31

To determine the loss coverage required on a commercial mortgage pool, each rating agency reviews a large sample (40%-60%) of the individual loans to assess the pool’s cash flow underwriting and asset valuation by property type. This assessment of cash flow underwriting by property type is extrapolated to the entire pool to determine the DSCR and LTV characteristics. This leverage analysis is the core of the agencies’ assessment — they feed the DSCR and LTV into a “base pool” loss matrix that produces a foreclosure frequency and loss potential of the loans at various rating levels. Thus, for a given LTV and DSCR, the rating agency’s base matrix provides a foreclosure frequency and principal loss value. The rating agencies view LTV as the key determinant of loss severity following loan default. To estimate collateral values used in the calculation of LTV, the rating agencies capitalize5 their estimated cash flow at yields that reflect property type, property-specific characteristics, and local market conditions. The rating agencies assess fixed and variable liquidation/workout expenses in the loss severity calculation. For smaller loans, fixed expenses can have a significant impact on loss severity. State foreclosure laws also affect loss severity. Accrued interest and deteriorating property performance will increase losses; the longer it takes for the mortgage servicer to gain control over the property and to liquidate it, the higher those losses will be. In Figure (9) is presented a generic example of how a rating agency matrix might translate debt service to an actual subordination level. For simplicity, the example is based on an early Fitch default and loss table, which has since evolved considerably. ___________________________________________________________________________ Figure 9. Loan Subordination, Using Default and Loss Matrix and Loan with 1.15 Times coverage ___________________________________________________________________________ Loan Default × loss A Level DSCR Prob. Prob.= Loss A level Class Initial +Add-On Final 0.10 80% 40% 32.00% Class Enhance Gearing Subord.= Subord.= Level______ 0.50 65 40 26.00 AAA AA×1.25 22.4% 6.3% 28.7% 0.80 55 40 22.00 AA A×1.28 17.9 5.0 22.9 0.90 45 40 18.00 A 14.00% A×1 14.0 3.9 17.9 1.00 40 40 16.00 BBB A/1.39 10.1 2.8 12.9 1.15 35 40 14.00 BBB- BBB/1.125 9.0 2.5 11.5 1.25 32 40 12.80 BB BBB/1.9 5.3 1.5 6.8 1.34 28 40 11.20 B BBB/4.2 2.4 0.7 3.1 1.50 25 40 10.00 1.75 20 40 08.00 ___ Source: Illustration using Salomon Smith Barney assumptions and Fitch DSCR Default and Loss Matrix. ________________ 5 One of the simpler commercial real estate valuation methods is to divide a property’s expected cash flow by a required property yield. The yield commonly referred to as the “cap rate”.

32

Rating agencies supplement the DSCR analysis with a loan-to-value loss matrix, which accounts for the differences in property types by valuing the properties using property-specific capitalization rates. The loan in the example has a 1.15 times DSCR based on the rating agencies’ assessment of cash flow and a higher-than-actual property specific debt refinance constant. The Fitch table assigns the 1.15 times coverage ratio a 35% probability of default with an expected loss of 40%. Although this example assumes that each loan has a 40% loss severity, the agency would actually adjust the loss severity for LTV, loan amortization, and state foreclosure laws. In addition, the rating agencies capitalize their re-underwritten cash flow at a cap rate6 or property yield that they feel is sustainable for the specific property. After adjusting the cash flow and using a higher cap rate, the rating agencies’ LTVs typically will be 15%-20% higher than those of the underwriters. So, an underwriter’s LTV of 65%-70% is usually assessed as an 80%-88% LTV by a rating agency. Typical rating agency loan loss expectations range from 30% to 50%, depending mostly on the agency’s assessment of LTV. The Fitch default and loss variables were based on events during the previous recession, which Fitch considered an A-level recession (meaning any loan rated less than A leverage was expected to have defaulted). Thus, its matrix creates credit enhancement at the A credit level. Therefore, the loan in the example that has a 35% chance of default times a 40% expected loss requires a 14% credit enhancement at the single-A credit level. Fitch then gears that 14% credit enhancement up and down the credit classes reflect an appropriate level of relative credit risk protection at each level. To create their base default and loss matrixes, the rating agencies have had to make some generic assumptions about the underlying mortgage pool. To the extent that the pool varies from the generic assumptions (or concentrations based on the grouping analysis) then the subordination levels are increased or decreased by an additional subordination “add-on.” In the above example, the “add-ons” are added to the initial loan’s leverage credit enhancement to create the final loan’s credit enhancement levels. The subordination add-ons can be pool- or loan-specific features that differ from the characteristics considered in the default and loss matrix. If a loan deviates from the base pool characteristics (such as geographic concentration, underwriter quality, special servicer quality, property type, loan amortization, property quality, leasehold mortgages, secondary debt, environmental issues, and operating history), then the rating agencies adjust levels by changing the loan’s foreclosure frequency and/or principal loss. ___________________ 6 The capitalization Rate or Cap Rate is a ratio used to estimate the value of income peroducing properties.

33

4.1.3 Declining Levels of CMBS Subordination, Implications for Investors. But what are the implications of the declining levels for investors? From figure 10, even with the 6% decline in triple-A CMBS subordination levels, CMBS pools can still withstand a high frequency of defaults and loss severity prior to experiencing a loss. For example, as mentioned by Darrell Wheeler, a senior analyst at Salomon Smith Barney indicates that the credit support built into 2000 CMBS transactions is well in excess of that found in high-risk mortgage security products (sub-prime or home equity loans). As previously discussed, rating agencies use different expected loss rates, typically ranging from 30% to 50%, depending on their actual assessment of the underlying real estate. Therefore in a situation where there is stable real estate market conditions, and given strong structural features of conduit mortgages will make it less likely that borrowers will default, thus, rendering the 40% loss rate conservative. Specifically, I believe that a strong real estate market will make it less likely that borrowers will default, while professional servicing and the cash management structure of conduit loans should limit future foreclosure loan losses. ___________________________________________________________________________ Figure 10. CMBS- Implied Default Rates at 40% loan loss 2000 CMBS Divide Level Loan Defaults Class Credit by Expected Required for Class Enhancement / Loan Loss = Loss AAA 24.61% 40.0% 61.5% AA 20.10 40.0 50.2 A 15.65 40.0 39.1 BBB 10.99 40.0 27.5 BBB- 9.61 40.0 24.0 BB 5.33 40.0 13.3 B 2.60 40.0 6.5 Source: Salomon Smith Barney. The above figure is to illustrate that credit support built into 2000 CMBS transactions is well in excess of that found in high risk mortgage security products (sub-prime or home equity loans). This figure shows that an average 2000 mortgage pool experiencing a 40% loan loss on every default would require 39% of the loans to default before the single-A rated class lost any of its principal. Similarly, a 24% loan default rate is required for a triple-B-minus loss and a 61.5% loan default rate is required for a triple-A class loss. It is helpful to compare the implied default rates to those in the 1994 Snyderman commercial default study7. This study tracked 10,955 commercial mortgage loans, originated by eight insurance companies. The study observed a 13.8% default rate over a five-year period and projected an 18.3% lifetime default rate for the entire pool. ________________

7Update to commercial mortgage defaults, Mark P. Snyderman, The Real Estate Finance Journal,Vol. 10, No. 1, summer 1994.

34

The study showed lifetime default rates that ranged from 7.6% to 21.5%, depending on the period studied, the originator, and the lifetime projection method. Therefore, it is clear that the investment-grade implied default rates in Figure 10 above have a more than sufficient buffer when compared with the study’s worst case lifetime default rate of 21.5%. If a 40% expected loss is accepted as reasonable, then the projected default rates for triple-A through triple-B classes seem conservative relative to historical default data and projections for the next real estate recession. The realization that early credit enhancements models were tougher than current subordination calculations helps explain why most CMBS spreads have tightened significantly after only a couple years of seasoning. Investors looking at 2000 transactions should also take comfort from the fact that rating agencies have room to contract subordination levels further, possibly creating typical conduit CMBS levels with 20% subordination in the triple-A class. Therefore, analysts feel that double-B, single-B, and unrated subordination levels have declined as far as they can. However, double-B and single-B CMBS investors buy at a discount to par and typically receive a full return of their capital in the first few years of their investment.

35

4.2 Basics of Commercial Real Estate Collateral Mindful of the real estate market recession in 1991, Banks, real estate investors, and developers continue to watch for a real estate recession. Thus, creating a market in which speculation is rare. It cannot be projected when the next real estate recession will occur, but it is believed that whenever it does, it will be less severe than the previous one. Current vacancy rates suggest that if there were an occurrence of economic slowdown, it would only marginally affect occupancies and rents. It would take a severe recession to have a significant impact on the current market, which is enjoying relatively low vacancy levels. These low vacancy levels are a direct outcome of the market’s experience during the previous recession. At this stage in the economic cycle, it is expected that construction starts for office properties, which are in short supply, continue to rise, while starts for other property segments have already started to decline or level off. A good indicator to monitor potential office market supply is to calculate inventory under construction as a percentage of each market’s total inventory. It has been observed Morgan Stanley that a market that has a total vacancy rate and potential supply figure greater than 12% bears watching. Given current high cap rates and low vacancy rates, the next real estate recession is not expected to be as severe as the previous one. Nonetheless, at this stage, until the full extent of the current economic slowdown is established, investors’ cautious approach to retail and hotel exposures in CMBS transactions is prudent. Retail construction starts never declined significantly from late-1980 levels as indicated by Salomon Smith Barney. Rather, retail development moved from regional shopping centers to power centers. This retail construction has been supported by strong consumer spending (including at times a negative savings rate), which means the pace of retail sales may be unsustainable in an economic slowdown. Investors should actively monitor retail loans because slowing retail spending will likely increase retail delinquencies. Because it is impossible to avoid retail loans altogether, investors should consider employing a retail tenant concentration review. It is also important to realize that while retail loans may experience a high level of delinquency, retail loan recoveries are usually higher than recoveries for other commercial property types because the property space can have many alternative retail uses.

36

4.3 CMBS Relative Value and Pricing Methodology Commercial mortgage investment has traditionally been considered a high-risk activity requiring specific market expertise. Therefore, when CMBSs were introduced, only accounts with specific real estate experience took to the new product. As the market grew and more investors began to follow the CMBS asset class, its spreads tightened. The October 1998 bond market crisis surprised the developing CMBS market, as many commercial mortgage originators and issuers had been carrying large mortgage inventories, enabling them to assemble optimal CMBS securities and enjoy excess mortgage spread return while the mortgages were on their books. When all debt spreads suddenly widened, these issuers suffered significant inventory losses because many had only partially hedged their positions. The impact of the bond market crisis on the CMBS market was compounded by the repurchase positions that many issuers held from selling bonds to subordinate investment-grade investors. Issuers that had made an active secondary market had as much exposure to CMBSs as their credit mechanisms would permit, causing bid/offer spreads to widen in the secondary market. During late 1998 and the first quarter of 1999, issuers cleared their inventories and those that have remained in the market have diligently worked since then to clear their mortgage inventory on a quarterly basis. Repurchase financing leverage levels have also been scaled back significantly. Today, market makers carry smaller CMBSs. Versus treasuries, as indicated by Jacob and Phillips in 2001 CMBS midyear update (CMBS World- fall 2001 issue), CMBS has solidly outperformed since June of 2000. 5-year AAAs were the laggards, only tightening 16 basis points (bp) versus anywhere from 31 bp in BBB- bonds to 43 bp BBBs. This, they said partially reflects the strong rally in the 5-year Treasury note yields compared with 10-year, which dropped over 150 bp compared with only about 85 bpfor the 10-year note during the period under review. Many investors and analysts compare triple- A CMBS to single-A corporate bonds. We have never compared a CMBS class to anything other than the similarly rated corporate class, because down-rated comparisons imply the CMBS class should be trading wide of a lesser-rated security. Recent CMBS spreads, which have demonstrated strong performance versus similarly rated corporate bonds, seem to vindicate this relative-value credit approach. Given the credit protection provided by the senior/subordinate structure of CMBSs and their strong underlying collateral performance, these credits should prove as good as or better than similarly rated corporate credits. Specifically, the CMBS structure means that any CMBS downgrade usually only results in a one-notch9 downgrade for a certificate — from single-A to single-A-minus (incrementally) for example — because of the subordinate credit enhancement built into each transaction. By contrast, corporate downgrades have demonstrated that they can move through multiple classes (single-A corporate ratings can become double-B or single-B) in a short time frame. Of course, to perform a full relative-value analysis investors should still review alternative investment opportunities. ___________________ 9 Notching is a term used to explain the rise or fall in the credit support levels from one CMBS tranche to the other.

37

In addition, the CMBS market continues to demonstrate strong liquidity that should persist in an economic stress scenario as long as the underlying collateral fundamentals remain strong. In an economic slowdown we would expect that cash flow deterioration would be slow and, as Fitch highlighted in a recent study10, partially offset by the loan diversification in the CMBS pools. However, as CMBSs are a relatively new product, the market needs tangible evidence of how CMBS collateral delinquencies and losses perform versus other debt products in an economic stress scenario. Once CMBS have demonstrated the credit benefit of their subordinate structure, then their strong liquidity should ensure they trade on par with similarly rated corporate classes. ________________________ 10Comparing CMBS and Corporate Bond Defaults, Diane Lans and Janet Price, Fitch, November 17, 2000.

38

4.4 Impact of CMBS Prepayment Variables Restrictions on prepayments, along with the time and cost (typically 3%-5% of the loan balance) involved in refinancing a commercial loan, mean that CMBS investors are unlikely to experience the sudden refinancings that afflict investors in MBSs backed by residential loans. However, defaults and resulting liquidations of loans can prepay principal to the upper-rated classes and create losses to the most subordinate class. Therefore, many investors also supplement their spread analysis with analysis that takes into consideration the potential of early bond prepayment from voluntary or involuntary loan prepayments. Alternatively, if the loan defaults occur at loan maturity, then the CMBS certificate holder can find the mortgages being extended, increasing their average life. Because of their potential cash flow volatility, many CMBS investors use collateral cash flow services to evaluate the bonds under different prepayment, default, and loss scenarios. The three major CMBS cash flow-modeling services are Charter Research, Intex Solutions, and TREPP. Each of these services enables an investor to model different prepayment speeds, delinquencies, and losses to evaluate CMBS pricing and yield sensitivity. Because defaults have been random and unpredictable, many investors feel that general overall default and prepayment rates, while admittedly unrealistic, at least provide some baseline stress analysis. Investors usually use four common pricing scenarios: a. 0% CDR11 and 0% CPY12. Zero defaults, no prepayments during the loan lock-out and yield maintenance periods, followed by no prepayments after the loans’ yield maintenance periods (0% CPY). Although unrealistic, this measure is the most common starting point to reference CMBS pricing. b. 0% CDR and 100% CPY. Zero defaults, no prepayments during the lock-out and yield maintenance periods, followed by prepayment in full (an assumption usually labeled 100% CPY). This is a common scenario and is also used by investors as a base line. c. 0% CDR for 2 years, 1% CDR thereafter and 100% CPY. Zero defaults for the first two years, followed by a 1% constant default rate (1% CDR) at a 40% loss severity over a 12-month recovery period with no prepayments during the lock-out and yield maintenance periods, followed by prepayment in full. A default rate of 1% per year (or a CDR of 1%) implies a cumulative default rate of 9.5% after ten years, which analysts view as a likely scenario. Investors usually also vary the loss rate from 30% to 50%. d. 0% CDR for 2 years, 2% CDR thereafter and 100% CPY. Zero defaults for the first two years, followed by a 2% constant default rate (2% CDR) at a 40% loss severity over a 12-month recovery period with no prepayments during the lock-out and yield maintenance periods, followed by prepayment in full. A 2% CDR implies a cumulative default rate of 18% after ten years and represents a possible delinquency scenario. _____________________ 11 CDR refers to a constant default rate that is applied annually to the outstanding pool balance. 12 CPY refers to constant prepayment rate following the yield maintenance period.

39

These various pricing scenarios can have different impacts on each of the certificates within a CMBS structure. Figure 11 presents the WAL (Weighted Average Life) and yield of a five-year triple-A and a ten-year triple-A, single-A, triple-B, and IO certificate from a recently issued tier two CMBS transaction. Fig. 11 CMBS WAL and Price for Generic Tier 213 CMBS Issue Under Multiple Scenarios___ 5-yr. Triple –A 10-yr. Triple-A 10-yr. Single-A 10-yr. Triple-B

10-yr. IO Scenario WAL –yrs Yield-% WAL-yrs Yield-% WAL-yrs Yield-% WAL-yrs Yield-% Yield-% a) 0%CDR, 0% CPY 5.87 6.38 9.71 6.52 9.90 6.85 9.90 7.42 9.17 b) 0% CDR & 100% CPY 5.74 6.38 9.51 6.52 9.82 6.85 9.90 7.42 8.83 c) 0% CDR for 24Mos., 1% CDR, 100% CPY 5.58 6.38 9.48 6.52 9.82 6.85 9.97 7.42 8.38 d) 0% CDR for 24Mos., 2% CDR, 100% CPY 5.48 6.37 9.44 6.52 9.86 6.85 9.98 7.42 7.93 Source: Salomon Smith Barney. This example highlights many pricing characteristics of the various CMBS classes: -The early prepayments highlighted in scenario II shortened the certificate’s average life, but only slightly decreased the yield on the five-year triple-A class. The IO class was the only class that was significantly affected by the faster prepayment. -Increasing the delinquency rate did not have a significant impact on any of the regular certificates because no principal was lost even at the triple-B level. However, the IO is affected by increases in default rates, because the IO is prepaid earlier with no collection of prepayment penalty. In the example, a 2% increase in default rate decreases the IO yield to 7.93%. However, I view 2% CDR as a high pool stress, given that it resulted in a cumulative pool default rate of approximately 18% (data not shown here). Nonetheless, the IO return at 2% CDR of 7.93% is still significantly higher than an initial yield on the ten-year triple-A class, demonstrating the strong relative value of CMBS IOs. This analysis ignored the possibility of loan extensions, which should have resulted from loan defaults at maturity and increased the yield on the IO certificates. ____________________ 13 Tier or Tiering- is used in the computation of capital adequacy. Tier 1, which refers to core capital, is the sum of disclosed reserves as adjusted and equity capital. Tier 2- refers to reserves, general provisions, sudordinated long-term debt and hybrid debt equity instruments.

40

This analysis was intended to show that ten-year investment-grade yields are fairly insensitive to reasonable prepayment and default scenarios, while premium- or discount-priced five-year triple-A and IO yields are most sensitive to changes in prepayment speeds and defaults. The scenarios assume that every CMBS pool has the same probability of default and expected prepayment speeds. An alternative and more sophisticated analysis would have assumed basing prepayments upon expected underlying equity refinance “cash-out” objectives and defaults upon expected future Loan to Values (LTVs). All three of the major cash flow modeling companies provide data on the underlying properties’ most recent reported Net Operating Income (NOI), enabling investors to adjust future expected property cash flow based on property- or region-specific NOI growth assumptions. In addition, other third-party services can be tied into the three cash flow models, providing cash flow growth and future valuation parameters based on their analysis of property market conditions in each of the various regions. 4.5 Current CMBS Pricing Conventions As noticed by Peter Rubinstein*, bond trading used to be much easier. Virtually all fixed income securities were offered on a spread to treasuries bases, which facilitated comparisons across markets. Now methods of pricing are all over the map. 4.5.1 A Swipe at Swaps Was the shift to swaps14 good? Initially, it appeared so. Numerous regression studies were produced to show the high correlation between swap spreads, CMBS spreads, and many other fixed income spreads. The claim was made that swaps are a good proxy for credit risk and this became a major supporting argument for using swaps as a benchmark. But there were problems that were overlooked too as indicated by Robinstein. First, the correlations did not always hold. Some of the historic regression studies looked good because they skipped periods of time where swap spreads did not correlate well. That was said to have been a tip-off about the consistency of using swaps as a benchmark. He continued that this consistency problem has now showed up several times in real life. For example, Treasuries rallied after the World Trade Center attack, but this time round, swap spreads tightened, just the opposite of what would be expected if swaps were a good proxy for credit risk. In CMBS, spreads to swaps decoupled after the World Trade Center attack and widened significantly, yet they remained stable to Treasuries. 4.5.2 Theory Reports show that what is missing from the swaps studies is a theory that establishes a direct cause and effect linkage between swap spreads and CMBS (or other asset) spreads. It is said however, that the theory doesn’t exist, because the relationship between swaps and the rest of the fixed income market is not casual. _________________ * Peter Rubinstein, Ph.D., is Managing Director at Bear, Stearns & Co. Inc. 14 Swap spread is simply the difference between the yield on a government security, such as a US Treasury, and the fixed rate of a conventional fixed/floating swap with a similar maturity to the referenced bond.

41

What actually drives pricing in the markets beyond pure and simple supply and demand for money, which is captured in overall interest rates, are multiple sources of risk (systematic and unsystematic). At any point in time, either one of these risks can dominate in determining the ultimate price in the market. Systematic risk stems from macro level events like war and recessions. Since systematic risks cause all bonds to respond together, under pure systematic risk, swaps (which are credit sensitive) respond along with other bonds, and so perform well as both a benchmark and as a hedge. However, there is a twist under this risk. Under systematic risk, Treasury bonds usually break away from the rest of the market because, as Robinstein put it, they are the only (credit) risk-free alternative in the world. He continued that markets, on the other hand, also experience unsystematic risks, which are often the result of technical factors specific to one or just a few markets. The swaps market is subject to these specific, unsystematic risks just as much as any other asset, and sometimes the impact of unsystematic risks dominate in pricing, which explains why swaps sometimes fail as a benchmark and a hedge, and why sometimes swap spreads appear to behave erratically. It explains, for example, why, in the post World Trade Center attack, swap spreads tightened (instead of widening to reflect increased systematic credit risk as in 1998) as Treasuries rallied. Analysis have thus, shown that over the past few months, both treasury bonds and swaps sometimes fail to work as expected, both as benchmarks and as hedges. 4.5.3 The Way Forward The fact that swaps are merely a signal, and the fact that risk specific to the swap market can make them a poor benchmark and a poor hedge is a complete reversal of the thinking in 1998 when there was the conviction that swaps were the answer. As Rubinstein put it, the essence of the problem is that there is no one number that will always function as a stable benchmark, which means that we must compromise. The best we might be able to do is to identify the performance criteria that define a good benchmark, and then see how the various alternatives stack up, he continued. 4.5.4 What Should Be Done? Rubinstein suggests that new issue 10-year triple-A CMBS should trade in the secondary on an absolute dollar price basis, or perhaps on a dollar price behind some other, more liquid, assets. He says people are perfectly capable of recognizing yields and spreads on their own. Absolute yield levels would also work, although yield levels introduce the complication of making prepayment and default assumptions. The second best solution is to abandon swaps and price CMBS on a spread to treasury basis. Corporate bonds, agency debentures, and other major asset classes are still quoted this way, and CMBS structures and convexity are far more like corporate bonds and agency bonds than most ABS and Residential MBS. Further, Treasuries have the best liquidity, best transparency, broadest participation, probably the best protection against tampering, and probably the highest trading volumes. The biggest problem however, with using Treasuries as a basis, many will argue, is that they can, and sometimes have, failed when used for hedging purposes. So have swaps, and the cost of going in and out of swaps is much more expensive. Realistically, hedging strategies need to be independent of the pricing benchmark because

42

empirically, no benchmark works well all of the time as a hedge. He says hedging is still an art as much as a science. 4.6 A Review of CMBS Defaults A good starting place to review CMBS defaults is the 1994 Snyderman commercial default study15 and a recent performance report of the public universe of CMBS data.16 The Snyderman study tracked 10,955 commercial mortgage loans, originated by eight insurance companies. The study observed a 13.8% default rate over a five-year period (equivalent to an annual rate of 2.6% CDR) and projected an 18.3% lifetime default rate for the entire pool (1.7% CDR). The study provided further expected lifetime default rates that ranged from 7.6% (1% CDR) to 21.5% (2% CDR), depending on the period studied, the originator, and the lifetime projection method. Salomon Smith Barney performance report shows that post-1993 CMBS collateral has accrued a 0.28% delinquency rate after two years, a 0.55% delinquency rate after three years and a 1.05% rate after four years. It is therefore felt that a reasonable and conservative range of default rates for scenario analysis is 1%-2% CDR. The assumption for the severity of loss from a resulting liquidation after default is critical to the analysis of CMBS cash flow. In a recession that causes loan defaults to increase, we would expect the loan loss to be more severe, likely shifting to around the 40% loan loss used by many investors. Also critical is the time until a loan defaults and the length of the recovery period after the default also affects CMBS cash flow. It is therefore suggested that analysis should allow for a 12-month to 24-month delay before defaults occur. The defaulted loan’s recovery period can also affect the underlying cash flows. Recent experience suggests a careful special servicer can take anywhere from 12 to 36 months to foreclose, work out, and liquidate a loan after a delinquency. In many instances, it is not uncommon for the special servicer to spend the first three to six months just assessing whether the delinquent borrower is acting in good faith to determine if foreclosure is necessary or whether a restructured loan might create the highest loan work out value. This discussion was intended to show that there is no specific expected default and loss scenario for analyzing CMBSs. This is why most CMBS structures are sequential pay, which insulates the longer- term investment-grade (BBB and above) classes from default prepayments, while exposing only the short triple-A, the IO, and the non-investment-grade classes to significant default prepayment risk __________________ 15 “Update to Commercial Mortgage Defaults,” Mark P. Snyderman, The Real Estate Finance Journal, Vol. 10, No. 1, Summer 1994. 16 SSB Issuer Performance Report, Salomon Smith Barney, October 18, 2000.

43

4.7 The Impact of Defaults on CMBS Transactions The impact that defaults can have on a transaction has caused investors to rank CMBS transactions based on perceived underwriter quality and transaction liquidity. Tiering CMBS transactions for liquidity makes sense because the events of 1998 removed some issuers from the market, creating orphan issues with no broker specifically committed to providing a market. Factoring in underwriter quality is meant to account for the concept of “credit culture,” which the market suggests is stronger in bank-based originators. Credit culture is the idea that some institutions have underwriting checks by third parties whose compensation is in no way tied to issuance volume. However, the rating agencies review the underwriter’s credit origination process and adjust credit enhancement levels to account for their analysis. The extra subordination built into some transactions may mean that investors are overcompensating for issuer quality if they also differentiate prices based on issuer. To date, loan defaults could be best described as random events. In general, defaults are higher in floating-rate transactions and on single-tenant retail properties. Delinquency rates are important, because any large delinquency immediately affects a CMBS transaction’s liquidity and causes CMBS spreads to widen. This spread widening is well acknowledged to be illogical, given that some defaults are originally projected for all CMBS transactions and are unlikely to create an investment-grade principal loss. However, given that spreads widen quickly on any bad, deal-specific news, investors have actively bid for transactions with low expected default rates, creating a small pricing difference between some issuers. Although highly subordinated transactions may experience more defaults, they are unlikely to remain on credit watch for long periods because the extra credit enhancement usually enables the rating agency to affirm levels quickly. In general, transactions with high credit subordination levels have extra default and loss protection and may weather a recession better than a lower-leverage transaction that has worse subordination levels. Specifically, a transaction with 1% extra subordination at the triple-A level can withstand 2.5% more loan defaults than a pool with lesser subordination (assuming a 40% loan loss rate). Analyses show that in 1999 the market actively bid tier one and tier two issues, while tier three transactions had limited liquidity. In 2000, the tiering price differential narrowed as investors realized that tier two and three transactions offered more default protection and greater relative value. As the CMBS market experiences more defaults, investors should grow more comfortable with defaults and further reduce any underwriter tiering differentiation in the market. Long-term CMBS investors will find value in the lower-tier CMBS issuers, which have greater subordination and wider spreads, while short-term CMBS investors that mark to market should stick with the tier one and two categories to minimize defaults and the resulting immediate spread widening caused by headline risk.

44

CHAPTER 5 5 EUROPEAN CMBS- THE MARKET TAKES OFF- IS SWEDEN PLAYING PART? 5.1 2001 Review Traditional CMBS transactions have encompassed many guises in Europe. They have principally, been supported by U.K. assets, ranging from Healthcare facilities, to Pubs, Offices and Shopping Centers. They have also encompassed theme parks, universities, and theaters. A review by Charles Gamm, Vice President and Senior Analyst at Moody’s shows that for all of 2001, there has been total CMBS European issuance of approximately €18.4 billion. This amount compares with a total €7.2 billion in 2000. The total for 2001 then translates into a growth of about 159% more than CMBS volumes in 2000. For this reason, Moody’s believes that a further growth of 40% to approximately €25.8 billion for 2002 is within reach. A number of significant events occurred in the European CMBS market during 2001. The market in the UK seemed to develop in a slightly different way to the market in continental Europe. In 2001, there was a mix of asset securitisations, actual and synthetic17 portfolio securitisations and sale and leaseback transactions. New participants have entered the market, particularly in continental Europe, while existing participants have re-entered the market through a mix of new issuance and tapping existing transactions. The UK continues to dominate in terms of collateral and debt issuance, but Moody’s has seen a continued increase in volume from the continental European market, with its use of synthetic securitisation structures. In the UK, the market saw a continuation of the activities of the past few years. Securitization transactions involved a mix of assets, as well as pools of commercial mortgages, originated specifically with the intention of securitising the loans in the pool. In the entire European market last year, 22 transactions were completed that ranged in size from €60.75 million to €2,889 million with an average size of €835.38 million. As stated, issuance in the United Kingdom still dominates the European market, but the number of continental European transactions has increased significantly over the past year, from five in 2000 to eight in 2001. Moody’s has also noted a number of transactions involving geographically diverse commercial properties in non-UK transactions, such as the global portfolio of hotel mortgages securitised in the Global Hotel One Limited transaction (the first global CMBS transaction) in 2001. It had properties from the USA, England, France, Italy, Denmark, Sweden, Belgium and The Netherlands. In addition, non-UK transactions were also originated from Sweden, Belgium, The Netherlands and France. In 2000, non-UK issuance amounted to 19.3% of the market, whereas this percentage increased to 26.1% in 2001. Moody’s believe that this trend of increasing issuance from continental European sources will continue into 2002. With the recent merger of three major German mortgage ___________________ 17 Synthetic CMBS- comprises guarantee structures whereby investors provide credit protection against the potential default of a pool of Loans.

45

Banks (Rheinhyp, Eurohypo and Deutsche Hypothekenbank) into one entity, Moody’s expects that by the end of 2002, this new entity will become a major participant in the European CMBS market. 5.2 2002 Outlook- Market Trend from Continental European Sources Analysis from Moody’s predicts that European CMBS will continue to expand for the whole of 2002 with volumes having a strong possibility of growing a further 40% to exceed €26 billion by the end of the year. This growth is to be realized through the expansion of the CMBS market from both continental and UK- based originators. The UK market will be dominated by a continuation of the trends that have been seen over the past year. Moody’s believes that true sale transactions will continue to dominate the market, and that repeat issuers will access the capital markets through newly formed securitisation structures. Moody’s further anticipates that the interest shown by German mortgage banks in accessing the public capital markets will continue to grow. The geographic diversity in the portfolios securitised to date will also continue with anticipation of further global pools of properties selected for securitisation. It is believed that more banks will continue to use the somewhat standardized synthetic securitisation structures in future. The use of true sale financing structures will also continue to expand throughout the continental European markets. Moody’s believes that other countries like France, Belgium, and Sweden will use securitisation as an alternative to more traditional bank financing in the coming year. 5.3 The Role of Sweden The focus on securistisation has dramatically increased within the Nordic region, and for that matter Sweden, in recent years with Handelsbanken Trading outlining new opportunities for Nordic real-estate companies to fund through CMBS, and also create advisory opportunities and a stronger balance sheet for banks. Separate interviews I had with Pär Tysk, Vice President- head of securitisation, Handelsbanken Trading/ Fixed Income, Per Sintring, corporate analyst- credit/financial research, Handelsbanken Trading/Fixed Income, Claes Norlen, President, Stadshypothek-subsidiary of Handelsbanken, and from SEB Merchant Banking-Debt Capital Market all point to the facts that past few years has seen dramatic increase in Sweden and in the Nordic region on asset-backed securities (ABS) and securitisation in general, with Handelsbanken playing a leading role. Corporates and banks have started to show real interest in the ABS market. During the 1990s, only a handful of Nordic borrowers seized the opportunities presented by the ABS market. The lion’s share of securitised transactions was based on assets stemming from real estate and residential mortgages with originators in both the public and private sectors. However, the turning point was during the year 2000. Analysis show that besides the increase and growth in volume, the main trend that is foreseen is the broadening of asset classes involved in securitised transactions with commercial mortgages as asset class with great potential that has yet to be explored within actual

46

transactions. Despite the growing use of commercial mortgage-backed securities (CMBS) in the European market, with its ability to deliver cost efficient funding to real-estate managers in several European countries, there is yet to evolve any major public CMBS transactions in the Nordic region. It is believed there are reasonable explanations for the hesitation of real-estate managers to enter into these complex transactions. However, it is also believed that this is about to change, considering the fact that there has been on September 2001, one European deal that was arranged by Handelsbanken Trading and SEB Merchant banking. One obvious reason why Nordic participants have yet to invest the time, effort and money needed for a CMBS transaction, is the availability of substantial and modestly priced bank lines. However, an increased focus on the correct pricing of credit risk, and an increased focus on the return on capital in Nordic banks, is about to change this picture. Another critical obstacle observed in Sweden, for instance, has been the Swedish legislation, which has hampered the growth of the Swedish ABS market in general for several years. Quite recently, the political pressure after the crises in Enron and World Com could be a threat. Touching on the Swedish Real Estate Market, this is what I found out from my interview. The value of all real estate in Sweden has been estimated at approximately €440 billion, of which 43% comprising commercial, industrial and multi-family properties, which could be regarded as securitisable. Over the past 10 years, the market has been characterised by low and stable inflation. The interview also show that direct yields are currently above funding levels with real estate companies generating relatively strong cashflows at modest leverage. Commercial rents have surged, but the cycle has now passed its peak. However, new construction is generally executed with high pre-let ratios and vacancies have remained fairly low. On the corporate front there is transformation with intense consolidation in the industry after some mergers and acquisitions. This event has created some large companies with the potential of entering into the marketplace for securitisations. From the facts presented above, I see a strong potential for CMBS structures in the Nordic market. However, the challenges are how to get the Rating Agencies to appreciate the structures being presented so ratings would be high. This is because investors want to see high ratings. Investors have also adapted a wait-and-see attitude by waiting to see what companies will enter the market first. Nonetheless, real estate companies would undoubtedly achieve a more diversified and, compared with increasingly expensive bilateral loans, cost efficient funding base. Since opportunities to take on Nordic credit risk in this sector have so far been scarce, as noted by one of the interviewees, I believe that investors will appreciate Nordic CMBS structures as an opportunity to obtain geographical diversification.

47

CHAPTER 6 6. SUMMARY AND CONCLUSION The advent of Commercial Mortgage Backed Securities (CMBS) and the Whole Loan Secondary Market has diversified methods of real estate lending. Traditionally, real estate lending was limited to commercial banks, thrifts, and insurance companies that provided new mortgage loan originations to local markets. Participating in the CMBS market has allowed lenders to not only spread underwriting risk but also mitigate the local economic market risk that was caused by serving only particular geographic areas. Other benefits include increased liquidity, the ability to hedge against cyclical credit crunches, borrower and collateral diversification and the stabilization of commercial property. CMBS, then, furnish conventional lenders with a broader array of real estate finance vehicles. Whether capital is needed for real estate acquisition or development the secondary market has allowed lenders to consolidate and sell their debt. Both private portfolio loans (large owners and developers) and institutional portfolio loans (insurance companies, banks, and pension funds) hold debt for which securitization is a viable option. I feel that investors should understand the following areas when reviewing CMBS:

1. Rating Agencies have come out with tools for measuring the likelihood that a particular commercial transaction will meet its schedule payments to investors in various economic scenarios. The focus of these methods ranges from financial leverage statistics, underwriting standards, properties types, and property tenancy. On the whole, the rating agencies’ credit enhancement models provide excess credit protection relative to the default and loss experience of the previous recession. It is expected that commercial credit enhancement levels will continue to decline even as the CMBS market experiences an economic slowdown and gains maturity.

2. Current CMBS collateral fundamentals are strong due to the relatively low level of

construction funding. Seasoned CMBS transactions have benefited most from the strong real estate market because the underlying properties have experienced significant appreciation since the loans were securitized, and yet the loans are precluded from prepayment by strong prepayment penalties. These conditions mean that it could take several years of low economic activity to significantly affect the credit quality of existing CMBSs.

3. Relative Value Analysis suggests that CMBSs provide significant excess return.

Considering CMBS market’s improved liquidity and given similar credit risk relative to other rated structured debt products, CMBS seem to have collateral structure and market characteristics that justify a tighter overall spreads (especially true for AAA classes including IOs).

The effects of the development of a CMBS market on the primary mortgage market will be considerable. The benefits should include increased liquidity; avoidance of cyclical credit crunches; increased geographic, borrower, and collateral diversification; evolving market standardization of underwriting; servicing and documentation; and the resulting stabilization of commercial property values. The disadvantages could include the risk that lenders may leave the real estate finance market or consolidate, the risk of poor underwriting and servicing, the loss of portfolio lender discipline, the loss of the personal lender/borrower

48

relationship, the unresponsiveness of investors, and the risk of governmental intervention in the market. The European CMBS market is not left out as it continues to demonstrate impressive growth. Given the different legal environments that still exist across the numerous European countries for funded CMBS transactions, the idea of a single Pan-European standardized structure, similar to what is seen in the US market, still seems far away. Despite this added complexity in the market, growth is still expected to be dynamic over the next year, with countries including France, Belgium and Sweden being the likely entrants soon There is also strong potential for CMBS structures in the Nordic market. Since opportunities to take on Nordic credit risk in this sector have so far been scarce, I believe that investors will appreciate Nordic CMBS structures as an opportunity to obtain geographical diversification.

49

REFERENCES PUBLISHED SOURCES Andrew S. Davidson, Michael D. Herskovitz, The Mortgage-Backed Securities Workbook: Hands On Analysis, Valuation and Strategies for Investment Decision-Making, Published by Irwin Professional Pub, April 1996. Chuck Ramsey, Frank J. Fabozzi, The Handbook of Non-agency Mortgage Backed Securities, Published by McGraw Hill, October 1999. Commercial Mortgage Securities Association (CMSA), CMBS World, Vol.3 Number 3, Fall 2001. Commercial Mortgage Securities Association (CMSA), CMBS World, Vol.3 Number 4, Winter 2002. Commercial Mortgage Securities Association (CMSA), CMBS World, Vol.4 Number 2, Summer 2002. Darrell Wheeler, Salomon Smith Barney, A Guide to Commercial Mortgage Backed Securities, January 2001. Darrell Wheeler, Strategy, Janet Showers, et al., A/B Note Structures in CMBS Transactions, Bond Market Roundup, February 18, 2000. Diane Lans and Janet Price, Fitch, Comparing CMBS and Corporate Bond Defaults, November 17, 2000. Frank Fabozzi (Editor), Handbook of Commercial Mortgage Backed Securities, Published by McGraw Hill, February 1999. Frank J. Fabozzi, Handbook of Structured Financial Products, Published by Frank J Fabozzi Assoc, October 1998. Frank J. Fabozzi, Mortgage and Mortgage-Backed Securities Markets, Published by Harvard Business School; May 1992. Frank J. Fabozzi (Editor), Trends in Commercial Mortgage Backed Securities, Published by McGraw Hill; August 1999. Ian Giddy, Asset Backed Securities- Corporate Finance Choices, 1999. John P. Harding and C.F. Sirmans, Commercial Mortgage Backed Securities: An Introduction for Professional Investors, Real Estate Finance Journal, Spring 1997, Pg. 43-51. Joseph C. Hu, Basics of Mortgage-Backed Securities, Published by McGraw Hill; September 1997.

50

Journal of Real Estate Research, Analysis of Yield Spreads on Commercial Mortgage- Backed Securities. 2002, vol. 23, issue 3, pages 235-252. Kau, J.B., D. Keenan, W.J. Muller III, and James F. Epperson, Pricing Commercial Mortgages and Their Mortgage Backed Securities, Journal of Real Estate Finance and Economics, 3 (1990), pp. 333-356. Mark P. Snyderman, Update to Commercial Mortgage Defaults, Real Estate Finance Journal, Vol.10, No. 1, Summer, 1994. PRECEPT CORPORATION, The Handbook of First Mortgage Underwriting, A Standard Method for the Commercial Real Estate Industry, published by McGraw Hill- 2002. Salomon Smith Barney, SSB Issuer Performance Report, October 18, 2000. Sally Gordon, Moody’s Investors Service, Introduction to Commercial Mortgage Backed Securities. Thomas F. Wratten, Introduction to Commercial Real Estate secondary and Securitization Market, October 1996. ELECTRONIC SOURCES www.dcrco.com www.fitchibca.com www.moodys.com www.gmac.com www.fitchratings.com www.cssacmbs.org www.csfb.com www.morganstanley.com www.ssb.com www.e-appraiser.net www.mccarystevens.com www.frankfabozzi.com

51

www.srz.com www.capitalconsortium.org www.mbaa.org www.standardpoors.com www.trepp.com

52

APPENDIX RATING AGENCY SUBORDINATION ADD-ONs Here is a description some vital subordination add-ons to CMBS pools to assist investors in reading rating agency presale reports. If any pool characteristic is highlighted as a negative in an agency presale, investors should know that this discussion means that the rating agency caught a specific item and increased subordination levels to account for the feature. In general, the more diversified and less dependent a pool is on any one entity, the less concentrated the credit exposure and the lower the subordination levels. Diversity Diversity is a measure of the loan concentration of the pool, which can affect subordination from 0% to 20%. Moody’s uses a Herfindahl index to measure the pool’s concentration and then scales its subordination add-on accordingly. The Herfindahl score is calculated by taking the inverse of the sum of the squares of each mortgage loan as percentage of the overall pool. Herfindahl Index = 1/∑ (loan/loan pool) 2 Investors in these concentrated pools are protected by the extra credit enhancement, but should strive to understand the larger underlying loans, should a default cause headline risk. Property Type (Loan Specific) Rating agencies believe that optimal credit consists of a pool perfectly diversified by property type. Given a less than perfectly diversified pool, the agencies make a credit assessment based on assumptions regarding the credit risk of each property category. They do this by assigning the property types different default probabilities or loss severity based on property-specific valuation property yields or DSCRs. Underwriting Quality This evaluation reviews whether the underwriter carefully reviewed the underlying property and its market and required appropriate loan structural features. For example, a loan with significant lease expiries may require an escrowed releasing reserve or even a cash trap in cases where a loan default is considered highly probable. Underwriter review also requires analysis of the underwriting criteria, appraisal standards, special hazard policies, and workout policies of the originator and servicer. The following areas will receive close focus on agencies’ evaluation:

53

1 Property appraisal standards and procedures. 2 Originator quality, based on historical portfolio performance.

3 Originator’s area of expertise, by mortgage purpose (for instance, refinance versus

new construction), property type and property location.

4 Servicing standards and policies, including those regarding balloon mortgages.

5 Approval procedures. Approval procedures that require third party non-business line credit reviews are seen as objective and receive subordination credit.

6 Documentation procedures and loan tracking systems.

In cases where all or some of the above data are unavailable, increased credit levels protect investors from potentially increased credit exposure. Rating agencies review the underwriter’s work and may adjust subordination levels by plus or minus 2% to reflect their opinion of the underwriter’s credit and loan structuring skills. Special Servicing The transaction’s master servicer is required usually to transfer servicing to the special servicer after a loan has been in default for 60 days. The special servicer is then required to deal with the defaulted loans to either restructure or recover the loan proceeds based on its assessment of the best method to maximize the net present value of loan proceeds. Rating agencies may also review each special servicer’s human resources and adjust subordination levels by 0% to 1.00% to reflect their opinion of the special servicer. The issuers realize that their issuance levels can suffer from having a poorly assessed servicer. Secondary Financing (Loan Specific) Many borrowers use secondary financing to leverage their property beyond limits imposed by rating agencies’ first mortgage guidelines. These secondary loans can take the following forms: Mezzanine Debt Mezzanine debt is the most common form of additional financing used by borrowers. A mezzanine loan is a loan secured by the equity interests of the first mortgage borrower, so it can only foreclose on the equity in the mortgage borrower, leaving the first mortgage unaffected. Another version of mezzanine financing is a preferred equity interest whereby the borrower issues preferred equity that earns a fixed rate of return and converts to common equity if interest rates are deferred. A/B Note Structures In the A/B notes structure, one loan is divided into a senior participation interest, which is deposited in the securitization vehicle, while the junior participation interest is privately placed outside of the securitization vehicle. In recent years issuers have used A/B notes to deliver larger loans, permitting the A-note to be shadow rated investment grade, improving the CMBS transaction’s overall credit levels, and attracting a broader investor base.

54

Any form of additional property leverage increases the property’s obligations, reducing the cash flow available to improve the property and, thus, increasing the first mortgage’s probability of default. To reflect this increased default risk, the agencies generally raise the first mortgage’s required subordination.

55