Embed Size (px)
Citation preview
Financial Instruments: ImpairmentAdapting to change
A new measurement philosophy
The change from the incurred to the expected loss methodology for measuring impairment represents a fundamental shift:
•Thedriversofimpairment(namely,creditlosses)willhavetobeassessedoverthefullexpectedlifeofaninstrument.Previouslythetimeframesformeasurementweremuchshorter,asonlyincurredlosseswereconsidered.
•Italsoincreasestheneedtomeasureaccuratelytheexpectedlifeofaninstrument(incorporatingpaymentbehaviour),asthismaymateriallyaffecttheperiodoverwhichaninstrumentisexposedtocreditrisk.
•Thiswillalmostcertainlyrequireafundamentalchangetotheoperatingmodel.
•Theadoptionofthe“openportfolioapproach”facilitateseasyapplicationoftheparameterstothebalancesheetwithouttheneedforgranularimpairmentcalculations.
•Itmakestheestimationofimpairmentmoresubjectiveparticularlyrelatingtoestimatesofhowcashflowsarelikelytorespondtotheeconomiccycle.
•Theincreaseinsubjectivityregardinglongertermcreditestimatesmakestheneedforarobustgovernanceprocess essential.
Tracking credit states through a transition matrix
Overitslife,afinancialinstrumentwillevolveoveranynumberofstates:
•Thisevolutioniscapturedinatransitionmatrix.
•Thetransitionorrollratematrixbelowisderivedasapivottablefromthemasterdatasuite.Itshowsthatifallthecustomersstartt0 in the Current state,83%remaininthatcategoryatt1,while2%haveprepaid,and10%and5%havemovedto30and60daysinarrearsrespectively.
•Thechangeindistributionint2isdeterminedbymultiplyingthenewt1 distributionthroughtherollratematrix,andsoon.
t0 t1 t2
Current
Default
Prepaid
Start\End Current Early Settlement 30 days in arrears 60 days in arrears Default Total
Current 83% 2% 10% 5% 0% 100%
Early Settlement 0% 100% 0% 0% 0% 100%
30 days in arrears 20% 0% 30% 45% 5% 100%
60 days in arrears 15% 5% 20% 40% 20% 100%
Default 0% 0% 0% 0% 100% 100%
The building blocks
Key inputs into the impairment calculation
Probability of default Loss given default Prepayment
Thedistributionbetweenstateswithinatransitionmatrixisusedtoderivetheprobabilityofdefault(PD)oftheportfolioforeachtime,t.
PDstypicallyhaveashape.Theexamplebelowistypicalofhomeloans.PDspeakafter18monthsto2years,andthenfallawaysharply.Thedecreaseusually stems from the fact that the housevaluehasincreasedatthispointtosuchanextentthattheborrowercansellhimselfoutoftrouble.
Reducingthedefaultratewillreducetheamount of the impairment.
Onceindefault,theeffectivenessoftheentity’scollectionsandrecoveriesprocessdeterminestherecoveryprofileandultimatelossassociatedwiththeinstrument.
Collectionsandrecoveriesperformanceisusuallyexpressedintermsofarecoveriescurve,reflectingtheamountcollectedineachperiod,d,sincedefault.Thelossgivendefault(LGD)represents1–(thevalueofrecoveriesdiscountedtothepointofdefault).
Accelerating the collection of arrears or increasingtheamountofcollectionswillreduce the amount of the impairment.
A change in prepayment assumptions does not directly impact the amount ofimpairment.Howeveritindirectlyinfluencesimpairmentthroughlengthening or shortening the period overwhichaninstrumentisexposedtocreditrisk.
Allotherthingsbeingequal,aportfoliowhosedurationshortensthrough an increase in prepayments shouldreflectareductioninimpairments,andviceversa.
Prob
abili
ty o
f de
faul
t
Duration / years on book (t)
0.20%
0.40%
0.60%
0.80%
1.00%
1.20%
1 2 3 4 5 6 7 8 9 10
Cum
ulat
ive
reco
very
rat
e
Months in default (d)0.0%
20.0%
40.0%
60.0%
80.0%
100.0%
% o
f or
igin
al b
alan
ce
Duration / years on book (t)
0.0%
20.0%
40.0%
60.0%
80.0%
100.0%
Contractual Behavioural profile 1 Behavioural profile 2
Prob
abili
ty o
f de
faul
t
Duration / years on book (t)
0.20%
0.40%
0.60%
0.80%
1.00%
1.20%
1 2 3 4 5 6 7 8 9 10
Cum
ulat
ive
reco
very
rat
e
Months in default (d)0.0%
20.0%
40.0%
60.0%
80.0%
100.0%
% o
f or
igin
al b
alan
ce
Duration / years on book (t)
0.0%
20.0%
40.0%
60.0%
80.0%
100.0%
Contractual Behavioural profile 1 Behavioural profile 2
Prob
abili
ty o
f de
faul
t
Duration / years on book (t)
0.20%
0.40%
0.60%
0.80%
1.00%
1.20%
1 2 3 4 5 6 7 8 9 10
Cum
ulat
ive
reco
very
rat
e
Months in default (d)0.0%
20.0%
40.0%
60.0%
80.0%
100.0%
% o
f or
igin
al b
alan
ce
Duration / years on book (t)
0.0%
20.0%
40.0%
60.0%
80.0%
100.0%
Contractual Behavioural profile 1 Behavioural profile 2
N
Durational PD Prepayment Durational LGD
Assumptions for each creditstate,Se.g.current,30,60,90days and time since origination,tforthePDandtimesincedefault,dfortheLGD
Assumption derivation
Data at individual account level
Historical account information
Where:N = contractual
maturity
H= thehorizon,whichis shorter than the remaining life,representingthe“foreseeablefuture”
t = time since origination
S = credit state
d = time since default
Forthebadbook,thePD=1,andthustheEListhecurrentbalancemultipliedbytheLGD.Forloansthathavebeenindefaultforsometime,adurationalLGD
dismoreappropriate,toreflectpost-defaultrecoveriesthathavealreadybeencollectedandreflectedinthebalanceoutstanding.
The impairment framework for instruments that are assessed collectively
Allowance account estimation
EL (S,t) WAA WAL
WAA WAL
Time since origination, t
Credit state, S
Balance
Data aggregated into cohorts
Effectivematurity
Age
m
i
Bal i
Bal portfolio
m
i
Bal i
Bal portfolio
EL LIFE (S, t) = Bal t x PD t x LGD
EL H (S, t) = Bal t x PD t x LGD
EL BAD (S, d) = LGD d x Bal d
ww
t
t
t + H
m= numberofexposures
in the portfolio
Key questions for management
•Doweunderstandtherequirementsofthestandard?
•Whatarethekeyareasofjudgement?
•Whatarethemainoperating model changeslikelytobe?
•Giventhatwewillrevisitourmodels,howcanweharnesschangetoimproveourmodellingefficiency,ortoprovidevalue-addoutofourmodellingprocesses?
•Whatenhancementstoourgovernance model arerequired?
- Shouldweinstituteaformalmodelvalidation processakintothewaywevalidatemodelsforregulatorycapitalpurposes?
- Shouldweinstituteaformalassumptions committeetoapprovetheassumptionsthemselves,specificallyanyeconomiccycleadjustments?
•Howshouldwecommunicatethechangeanditsimpact,bothinternallyandexternally?
LGD and the alignment with Basel II parameters
Mean
Valuation
Volatility
Stress tests
Tails
confidence interval
Whilstaccountingseekstoaddressthemean,orexpectedvalue,theBaselIIcapitalrequirementseekstoaddressvolatility,orunexpectedloss.
ThebankshouldconsiderusingtheexpectedLGDforaccounting,ratherthantheBaselIIdownturnLGD.
Introducing discounting
Thedecisiontodiscountiselective.DiscountingmaybenefitthebankwherethetermstructureofELislong.Wherelossesemergeearly,thecostbenefitconsiderationsmeanthatdiscountingmaynotdeliverclearbenefit
EL LIFE (S, t) = (1+i)t Bal t x PD t x LGD x (1+i)-t-1
w N
t
Risk free rate ≤ i ≤ effective interest rate
Bal t of a revolving portfolio
Determiningtheprojectedbalanceofanopenportfolioofrevolvingfacilitiesiscomplex,asprojectionsoffuturebalance,Balt,shouldavoidinclusionofnewaccounts,andfuturedrawsonexistingfacilities.Theirinclusionwilloverstatetheallowanceaccount.
Therearevariousestablishedmethodstomodelrevolvingfacilities.Thebankcouldmodela“stable”and“volatile”component,andmakedifferentrun-offassumptionsforeachcomponent,whichisanapproachcommonlyusedtomodeltheliquiditytermstructureofrevolvingdeposits.
Thepolicythatmanagementelectsislikelytohaveasignificantimpactonthemodellingresult.
The rationale for change
•Followingthecrisis,theincurredlossmodelhasbeencriticisedforcontributingto procyclicality. Under theincurredlossmodel,impairmentlevelstendedtobelowestjustbeforethecreditcycle’sturn.Aslossesmounted,significantincreasesinimpairmentlevelswererequired.Thisputadditionalpressureonbanks’financialresourcesattheirweakestpoint.Theproposedexpectedlossapproachwillrequiretherecognitionof the lifetime expected loss (EL LIFE)ofanadvanceonatime proportional basis.
•Toensurethatportfolioswherelossesemergeearlyintheirlifecycleareadequatelyimpairedinadvanceofthelossbeingrealised,theimpairmentlevelissubjecttoaminimum floor.Thefloorissetatthelevelofthelossexpectedtoemergeovertheforeseeable future.Inthenotationthatfollows,wehavedescribedthisperiod as the Horizon, H.
•Foranopenportfolio,theproposedimpairmentlevelisrelativelystable as new loans replace old. Impairmentlevelswilllikelychangeduetochangeinportfoliosize,changeincreditqualityorchangeinagingand/orpaymentbehaviour.
•ImpairmentlevelsfortheperformingbookwilllikelyincreaseandareexpectedtoalignmorecloselywiththeExpectedLoss(EL)underBaselII.BanksthatapplytheAdvancedIRBforcreditriskshouldexperiencecapital relief,throughasignificantreductioninthedeductionfromeligiblecapitaloftheshortfallbetweenimpairmentandtheBaselIIEL.Notethattheinitialchangeinimpairmentlevelswilllikelygothroughbalancesheetrestatementratherthanthroughtheincomestatement.
The impairment balance
For the “good book”
SegmentationCalculationsofparametersandtheallowanceaccountwilltypicallybesegmentedtoreflectdifferentcreditstatuses,S,(e.g.current,arrears,default)andtimesinceorigination,t.
Frequency Creditandbehaviouralassumptionsshouldberefreshedateachreportingdate.
Time proportionality
Timeproportionalityisreflectedbytheratiooftheweightedaverageageoftheportfolio(WAA)toitsweightedaveragelife(WAL).Forastableportfolio,theincomestatementchargeforimpairmentwillbedrivenbychangesincreditparametersandportfoliomix,ratherthanbytimeeffects.
Allowance account = max EL LIFE (S, t) , EL H (S, t)
WAA
WAL
x
Time proportional amount Floor
For the “bad book”
Allowance account = EL BAD (S, d)
•ThefullamountoftheELonthebadbookisrecognisedintheallowanceaccountimmediately.
•TheELBADtypicallyreflectsthedurationalLGD,whichdependsonthelengthoftimetheadvancehasalreadybeenindefault,d.
Overview of the supplement to ED/2009/12
EL LIFE
EL H
H, the “foreseeable future”
WAL, weighted average life
t
EL LIFE , EL H and WAL
TheELLIFEandtheportfoliobalanceatanytimesinceorigination,t,aredependentontheportfolio’sbehaviouralprofile.
Thebehaviouralprofile,WAAandWALshouldbere-estimatedperiodically.
The open portfolio
•Asaconcessiontothecomplexityofcalculatingimpairmentatagranularlevel,theproposalhasintroducedtheconcept of the open portfolio.
•Creditparameterscanbeappliedtotheportfoliobalanceratherthantoindividualadvances.Toensureanaccurateestimate,theportfolioshouldbesegmentedbycreditstateandtimesinceorigination.
Allowance account (S, t) = EL% LIFE (S, t) x Bal (S, t) x WAA ÷ WAL
•Thetimeproportionfactor,WAA/WAL,attemptstoreplicatetheuseoftheeffectiveinterestrate(EIR),inordertomatchimpairmentwithinterestincomefromlendingactivities.Theuseoftheportfolio-levelcharacteristicscapturedintheWAA/WALfactormeansthatthebankdoesnotneedtocalculateandcapturetheEIRatagranularlevel.Thissignificantlyreducesoperationalcomplexityandisreferredtoasadecoupledapproach in the literature.
“Good” versus “bad”
The“good”and“bad”definitionsshouldfollowthebank’sinternalriskmanagementprocesses.Alternatively,whenthevalueofexpectedlossesisexpectedtoexceedthecreditriskpremium,thenaloanshouldbeclassifiedasbad.
Thedefinitiondirectlyimpactsthetiming of loss recognition(immediateversustimeproportional).
Wesupportalignment with regulatory “default” definitiontomaximisealignmentwithBaselIIELmeasures,andtoavoidunintendedcapitalconsequences.
The horizon, H
Thefloorrepresentsthe“foreseeablefuture”overwhichlossestimatesarereliablyestimableandmaynotbelessthantwelvemonths.Differinghorizonsmaynegativelyaffectcomparabilitybetweenbanks.
Pro
babi
lity
of d
efau
lt
PD LIP
The Exposure Draft in practice
Wehavecreatedahypotheticalstableportfoliotoassesstheimpactofthetimeproportionalallowanceaccount(basedonthelifetimeEL)versusthefloor(basedonthefullELovertheHorizon,H).Forthepurposesofthisillustration,wehaveassumedthatHisoneyear.Thefollowinggraphsreflecttheresultsofavarietyofassumptionsabouttheportfolio’slossemergencepattern.
PDt significantly front-end loaded (early emergence)
PDt significantly back-end loaded (late emergence)
t
In all instances, the floor exceeds the time proportional allowance account. This effect is even more pronounced where H > one year.
This indicates that the allowance account may be approximated by ELH in many circumstances.
Quantitative impact
ForthePerformingportfolio,theproposalsrepresentasignificantincreaseoverthecurrentIBNRprovision,andwillconstitutethebulkoftheadditionalimpairmentrequired.
FortheArrearsportfolio,mostbankscurrentlyprovideonthebasisoftheELLIP.
ThismeansthattheyprovidefortheabnormallyhighlossesexpectedovertheLIP period.BorrowerswhohavenotdefaultedwithintheLIPperiodtypicallyreturntothePerforming portfolio.
Wehavesuggestedaconservativeadd-ontotheELLIP to reflectlossesovertheHorizonforadvancesthatsurvivetheLIPperiod.
Treatmentofthedefaultportfolioremainsrelativelyunchangedalthoughsomeflexibilityaroundthediscountratehasbeenintroduced.
.
Performing portfolio = ELH
(Performing)
Arrears portfolio = EL LIP (Arrears) + ELH
(Performing)
Default portfolio = EL LIFE (Default)
time
Operating model consequences
No need for roll rates / transition matrix
PD HandPDLIPcanbemeasureddirectly,withouttheneedtocalculatethePDforeachduration,t.
No need for prepayment modelling
OnthebasisoftheabovemethodofPDestimation,andthefactthattheWAAandWALfactorsareignored,thereisnoneedtomodelprepayments.
No need for discounting
ThehorizonsHandLIPtendtobeshort,sothediscountingbenefitissignificantlylessthandiscountingoverthefulllife.
Impact of a change in estimate
TheuseofthefloorcontinuestoholdwhereachangeinELestimateoccurs.
Catherine Stretton
Partner-CapitalMarkets
+27(0)844447033
Pravin Burra
Director-CapitalMarkets
+27(0)823361515
©2011Deloitte&Touche.Allrightsreserved.MemberofDeloitteToucheTohmatsuLimited
DesignedandproducedbyCreativeSolutionsatDeloitte,Johannesburg.(801989/mar)
