Operational and integrated risk management FRM

FRM PART II BOOK 3: OPERATIONALOPERATIONALAND INTEGRATED RISK MANAGEMENT 34:CapitalAllocation andPerformanceMeasurement 35:RangeofPracticesandIssues in EconomicCapital Modeling 36:Assess

1 of 2

FRM PART II BOOK 3: OPERATIONAL

OPERATIONALAND INTEGRATED RISK MANAGEMENT

34:CapitalAllocation andPerformanceMeasurement

35:RangeofPracticesandIssues in EconomicCapital Modeling

36:Assessingdie Quality of Risk Measures

37: LiquidityandLeverage

38:EstimatingLiquidityRisks

39:ModelRisk

4t):EnterpriseRiskManagement: TheoryandPractice

4l: AReviewof theKey IssuesinOperational Risk Capital Modeling

42: Challengesand PitfallsinMeasuringOperationalRisk from Loss Data

43:The Failure MechanicsofDealer Banks

44: Principles for the Sound Management of Operational Risk

45:ObservationsonDevelopmentsin RiskAppetiteFrameworks

andITInfrastructure

46:StressTesting Banks

47:BaselII:InternationalConvergenceof Capital Measurement

andCapita]Standards

48:Basel III: A Global RegulatoryFrameworkfor More Resilient Banks and

BankingSystems

49:Basel III: International FrameworkforLiquidityRisk

Measurement,Standards,andMonitoring

50:Revisions tothe Basel II Market Risk Framework

51:OperationalRisk-Supervisory Guidelinestor theAdvancedMeasurement

Approaches

52:A Comparative Assessmentof Basel II/IIIandSolvency II

SELF-TEST: OPERATIONALAND INTEGRATED RISK MANAGEMENT

PAST FRM EXAM QUESTIONS

FORMULAS

INDEX

13254l52

72

8795105

117

126

137

151161

172

200

219235

243

258

272 278 302

305

e-READING ASSIGNMENTS AND

Thefallowingmaterialis a reviewoftheOperationalandIntegratedRiskManagement

principlesdesignedtoaddress the AIMstatements setforth bytheGlobalAssociationofRisk

Professionals,

READING ASSIGNMENTS

MichelCrouhy,DanGalai and Robert Mark,Risk Management(New York:

McGraw-Hill, 2001)

34. “Capital AllocationandPerformanceMeasurement/Chapter14

35 “RangeofPracticesand Issuesin EconomicCapitalFrameworks/ (Basel Committee on

BankingSupervision Publication,March 2009).

Allan Mala, Financial RiskManagement:Models,History, andInstitutions(Hoboken,NJ:

JohnWiley& Sons,2011)

36 “AssessingtheQualityofRiskMeasures/Chapter1 1

37- “Liquidity andLeverage,”Chapter12

Kevin Dowd,Measuring MarketRisk,2nd Edition(West Sussex,England:JohnWiley&

39 “ModelRisk/ Chapter16

40.Brian Noccoand RendStulz, “EnterpriseRisk Management: Theory andPractice/

Journal of AppliedCorporateFinance Id,No.4(2006): 8-20. (page95)

4l MoChaudhury, “AReviewof dieKeyIssuesinOperationalRiskCapital Modeling,” The

Journal ofOperationalRisk,Volume5/Number3,Fall2010: pp.37-66 (page105)

42.EricCope,GiulioMignola,GianJucaAntoniniand Roberto Ugoccioni,“Challengesand

Pitfallsin MeasuringOperational Risk fromLoss Data,” TheJournal of OperationalRisk,

Volume4/Number4,Winter2009/10: pp 3—27

43- Darrell Duffie, 2010.“Failure MechanicsofDealerBanks/ Journal ofEconomic

Perspectives24:1,51—72

(page117)

(page126)

44.“Principlesfor dieSound Managementof Operational Risk,” (Basel Committee on

BankingSupervision Publication,June2011). (page137)

Book 3

Reading Assignments and AIM Statements

45 “Observationson Developmentsin RiskAppetite Frameworks and IT Infrastructure,”

SeniorSupervisorsGroup, December 2010

46 TilSchuermann."StressTestingBanks,*April 2012.

47.“Basel International ConvergenceofCapitalMeasurementandCapitalStandards:

A Revised Framework—ComprehensiveVersion,”(BaselCommittee onBanking

Supervision Publication,June 2006)

{page151)

(page161)

(page172)

48 “BaselHI: AGlobal Regulatory Framework forMoreResilient BanksandBanking

Systems—RevisedVersion,”(Basel CommitteeonBanking Supervision Publication,

49.“BaselHI:InternationalFrameworkfor Liquidity RiskMeasurement,Standardsand

Monitoring,*(Basel Committee on BankingSupervision Publication,

50 “Revisions to the Basel II Market Risk Framework—Updatedasof31December 2010,”

(Basel Committee onBankingSupervision Publication,February2011). (page235)

51 “Operational Risk—Supervisory Guidelines for dieAdvancedMeasurement

Approaches,” (BaselCommitteeonBankingSupervision Publication,

52.NadineGatzert,Hannah"Wesker, "A ComparativeAssessmentof BaselII/IIIand

SolvencyII,”WorkingPaper, Friedrich-Alestander-University ofErlangen-Nuremherg,

Version:October 2011. (page258)

Book3ReadingAssignments and AIM Statements

AIM STATEMENTS

34. Capital Allocation and PerformanceMeasurement

Candidates,after completingthisreading, should beableto:

1.Describe the RARQC(risk-adjusted return on capital) methodologyanddescribe

someof the potentialbenefitsofits use.(page13)

2 Define, compareandcontrasteconomic andregulatorycapital, (page13)

3. Computeand interpret theRAROCfora loan or loanportfolio,and useRAROC

tocomparebusinessunit performance,{page 14)

4. Explain how capitalisattributed tomarket, credit,and operational risk,(page15)

5 Calculate thecapital chargefor market risk and creditrisk,{page15)

6 Explain the difficultiesencountered inattributingeconomiccapital tooperational

risk, (page15)

7 Descrihe theLoanEquivalent Approach and use it tocalculateRAROCcapital

{page17)

H. Explain how thesecond-generation RAROCapproachesimproveeconomiccapital

allocationdecisions,(page18)

9 Computetheadjusted RAROCfora project todetermineitsviahility.{page18)

35.RangeofPracticesandIssuesin EconomicCapitalModeling

Candidates,after completingthisreading,shouldbeableto:

1 Within die economiccapital implementationftameworkdescrihe thechallenges

that appear in:

• Definingriskmeasures

• Riskaggregation

• Validationof models

* Dependency modelingin creditrisk

• Evaluatingcounterpartycredit risk

• Assessinginterest rateriskin thehankingbook

(page25)

2 Describe theBIS recommendationsthatsupervisorsshould consider tomake

effective useof risk measuresnotdesigned for regulatory purposes,(page35)

3 Descrihe theconstraintsimposedand the opportunities offeredbyeconomiccapital

within die followingareas:

• Credit portfoliomanagement

• Risk based pricing

• Customer profitability analysis

• Managementincentives

(page36)

36.AssessingtheQualityof Risk Measures

Candidates,after completingthisreading,should heable to:

1 Describewaysthaterrors can heintroducedinto models,(page41)

2. Descrihe the typesofhorizon,computational andmodeling decisions which could

resultinvariabilityof VaRestimates, (page 42)

3 Identifychallengesrelated tomapping of risk factorsto positions inmakingVaR

calculations,(page 43)

4 Explain howimproper mappingcan understatespecific risks suchas basisor

Book 3

Reading Assignments and AIM Statements

5- Identifyreasonsfor the failure of dielong-equity tranche, short-mezzaninecredit

tradein 2005 and describe how such modelingerrors couldhave beenavoided

(page44)

6 Identifythe two majordefectsin modelassumptionswhichled to the

underestimationof systematic risk for residential mortgage backed securities

(RMBS) during the200&-2009financialdownturn, (page46)

37 LiquidityandLeverage

Candidates,aftercompleting this reading, shouldhe ableto:

1. Defineanddifferentiate betweensourcesofliquidityrisk, includingtransactions

liquidityrisk,balance sheet/funding liquidity risk andsystemicrisk, (page52}

2 Summarizetheprocesshy whichafractional-reserve bankengagesin asset liability

management, (page 53)

3 Describeissues related tosystematicfunding liquidityriskwith respect co LBOs,

merger arbitrage hedgefunds, andconvertiblearbitrage hedgefunds,(page54}

4 Explain specific liquidity issues facedhymoney marketmutualfunds, (page54)

5 Describe theeconomicsof the collateral market and explain themechanicsof the

followingtransactions using collateral:marginlending,repos, securitieslending,and totalreturnswaps, (page 55)

6. Calculateafirm'sleverageratio,describe theformulafor theleverageeffect,and

explain tire relationship hetweenleverage andafirms return oilequity,(page57)

7 Computeafirm’sleverageand construct a firm’s balance sheet given thefollowing

typesoftransactions: purchasinglongequity positionsonmargin,enteringinto

short sales,andtradingin derivatives,(page59)

B Identifythe main sources oftransactionsliquidityrisk, (page63)

9 Calculate tire expected transactions cost and the99 percent spread risk factor fora

transaction, (page64)

10. Calculatetheliquidity-adjustedVaRfora position toheliquidatedover a number

oftradingdays, (page65)

11 Define characteristics used to measure market liquidity, including tightness,depth

andresiliency, (page66)

12. Explain thechallenges posed hy liquidityconstraints ouhedgefundsduring times

of financial distress, with anemphasison handling redemptions, (page66)

38.EstimatingLiquidityRisksCandidates,aftercompleting thisreading,should he ableto:

1. Defineliquidityrisk anddescrihefactors thatinfluenceliquidity,(page72)

2 Discussthe hid-ask spreadas a measureof liquidity,(page72)

3 Defineexogenous andendogenous liquidity, (page73)

4 Descrihe thechallengesofestimatingliquidity-adjustedVaR (LVaR).(page73)

5 DescribeandcalculateLVaR usingdieConstantSpread approach and the

ExogenousSpread approach,(page74)

6 Describe Endogenous Priceapproaches to LVaR, its motivation andlimitations.(page78)

7 Explain dierelationshiphetweenliquidationstrategies, transaction costsand marketprice impact, (page79)

8. Describeliquidityat risk (LaR) and describe thefactors that affect futurecashflows.(page BO)

9. Explain theroleof liquidityin crisis situationsanddescribeapproachestoestimatingcrisisliquidity risk, (pageSi)

Reading Assignments and AIM Statements

39.ModelRisk

Candidates,aftercompletingthisreading,should beahleto:

t Define modelrisk;identifyanddescribesourcesof model risk*(page87)

2 Descrihe the challengesinvolvedwith quantifying model risk* (page88)

3. Describe methodsfor estimating modelrisk,givenanunknown componentfroma

financialmodel* (page88)

4 Identifywaysriskmanagerscan protect against modelrisk* (page90)

5 Summarize theroleofsenior managers inmanaging model risk,(page90)

6 Descrihe procedures forvettingandreviewingamodel*(page91)

7 Explain the function ofan independentriskoversight(IRO) unit* (page91)

40.EnterpriseRisk Management:Theoryand Practice

Candidates,aftercompletingthisreading,should be ahle to:

1. Defineenterpriseriskmanagement(ERM).(page95)

2. Explainhow implementingERM practicesand policiescreateshareholdervalue

bothat themacroand dremicro level,(page 95)

3 Explain howan ERM programcan beused todeterminedie rightamountof risk

(page97)

4 Descrihe the development and implementation ofanERMsystem,(pagje97)

5 Explain therelationshipbetween economicvalueandaccounting performance

(page 98)

6. Describe the roleof andissueswithcorrelation inrisk aggregation, {page98)

7* Distinguish between regulatory andeconomiccapital* (page99)

8 Explain the useofeconomiccapitalinthecorporatedecisionmaking process

(page99)

41.AReviewof the Key Issues in OperationalRisk Capital Modeling

Candidates,aftercompletingthisreading,shouldhe ahleto:

]* Descrihe the loss distribution approach tomeasuring operational risk* (page105)

2. Identifyissues relatedto external andinternaloperational los*sdatasets, (page1(J7)

3 Explain how frequency andseverity distributionsof operational lossesareobtained*

42.Challengesand Pitfalls in Measuring Operational Risk from Loss Data

Candidates,aftercompletingthisreading,should be ahle to:

I Describe thenatureofoperationallossdistributions,(page118)

2. Explain theconsequencesof working with heavy tailedlossdata* (page1 19)

3 Determine theamountof datarequired to estimate percentilesoflossdistributions

(page119)

4. Descrihe methodsofextrapolating beyond thedata, (page120)

5 Explain the lossdistributionapproachtomodeling operational risk losses

(page121)

6. Explain thechallengesin validatingcapital models,(page122)

Reading Assignments and ATM Statements

43 TheFailureMechanicsofDealer Banks

Candidates,after completing thisreading,shouldbeahleto:

1 Descrihe the majorfunctionsoflargedealer banks andexplain thefirm-specificandsystemicriskfactors attendant to each, (page126)

2. Descrihe thestructureof the major markets inwhichlargedealer banksoperate.

{page128)

3. Explain howdiseconomiesofscopein riskmanagementandcorporategovernance

mayarise inlargedealerbanks,(page129)

4 Identify factors thatcan precipitateor acceleratealiquiditycrisis at adealer hank

and what prudentriskmanagement steps can be taken tomitigatethese risks.(page130)

5 Comparealiquidity1'crisis at adealerbank to atraditional hank run (page132)

6. Descrihe policymeasures thatcouldalleviatesomeofthe firm-specificand systemic

risks related tolarge dealerbanks,(page133)

44 Principlesfor the SoundManagementof Operational RiskCandidates,aftercompletingthis reading,should beahle to:

1 Descrihe the three ‘'linesof defense” in the Basel modelforoperationalriskgovernance.(page137)

2. Define anddescribediecorporateoperationalrisk function(CORF)and compare

andcontrastthestructureand responsibilities of dieCORFatsmallerand larger

hanks, (page138)

3 Summarizethe eleven fundamental principles of operational riskmanagement as

suggested bythe Baselcommittee,{page138)

4 Evaluatethe roleof theBoardof Directors aswellas senior management in

implementinganeffectiveoperationalriskstructure per theBasel committee

recommendations, (page 139)

5 Descrihe the elementsofaframeworkforoperationalrisk management,including

documentationrequirements,(page142)

6. Identifyexamplesof tools whichcan he usedtoidentifyandassessoperationalrisk

(page 143)

7 Descrihefeatures ofaneffective control environment and identify specificcontrolswhichshouldbein place toaddressoperationalrisk, (page144)

8. Evaluate theBasel committee’ssuggestions formanaging technologyriskand

outsourcing risk,(page144)

45.Observations onDevelopments inRiskAppetiteFrameworks andITInfrastructure

Candidates,aftercompletingthisreading,shouldbeahleto:

1 Describe theconceptofa risk appetite framework(RAF),idendfythe elementsofa

RAFandexplain die benefitsto afirmofhavingawell developedRAF.(page151)

2 Descrihe best practicesfor afirm’s Chief Risk Officer(CRO),Chief Executive

Officer (CEO) and BoardofDirectors inthedevelopmentandimplementationof

aneffective riskappetiteframework,(page152)

3 Explain theroleofa RAFin managing die risk ofindividual businesslineswithina

firm, (page153)

4 Identifymetrics whichcan hemonitoredas partofaneffectiveRAPanddescribe

the classesof metrics tohecommunicated to variousmanagers within the firm

{page154)

©2013Kaplan, Inc,

Page8

Book3ReadingAssignments and AIM Statements

5 Explain die henefitsto afirmfrom havinga robuserisk data infrastructure,

anddescribe keyelementsofaneffective IT riskmanagement policyat a firm,

46 Stress TestingBanks

Candidates,aftercompletingthisreading,should be ableto:

1, Explain the differencesin thefeatures andscope ofstress testsbefore andafter the

SupervisoryCapitalAssessmentProgram (SCAR). (page162)

2. Describe die problemofcoherence in modelingriskfactorsduringthestresstesting

of banks,(page163)

3 Describe thechallengesinmapping from broadermacroeconomicfactors tospecific

intermediate riskfactors in modelinglosses, (page 164)

4 Explain diechallengesinmodelingabanks balance sheetover a stress test horizon

period (page164)

5 Compareandcontrast the2009SCAPstress test, the2011 and2012CCAR, and

the20 11 EBAIrishandEBA Europeanstress tests in their methodologies and key

findings, (page 165)

47 BaselII:InternationalConvergence ofCapitalMeasurement andCapital Standards

Candidates,after completing diis reading, should beableto:

I Describe the keyelementsof die threepillarsof Basel II:

* Minimumcapital requirements

3 Describe the major risk categoriescoveredby die BaselIIAccord,(page179)

4 Describe andcontrastdie major elements of the three options available for the

calculationof credit risk:

• Standardised Approach

* Foundation IRBApproach

* AdvancedIRBApproach

(page181)

5 Describeand contrastdie majorelementsof the threeoptionsavailablefor the

calculationofoperationalrisk:

• BasicIndicatorApproach

* StandardisedApproach

• AdvancedMeasurementApproach

(page190)

6. Describeand contrastdie major elements—includinga descriptionof the risks

covered—of thetwooptionsavailablefor the calculation of market risk:

* StandardisedMeasurementMethod

* InternalModelsApproach

Reading Assignments and AIM Statements

7 Definein thecontextcd'Basel II andcalculatewhere appropriate:

* Capitalratio

* Capitalcharge

* Riskweightsand risk-weightedassets

* Tier 1capital andits components

* Tier2capitalandits components

* Tier3 capital andits components

* Probability of default(PD)

* Loss given default (LCD)

* Exposureatdefault (EAD)

1. Describereasons for thechangesimplementedthrough the Basel III framework

(page200)

2 Describechanges to the regulatorycapital framework,including changesto:

* Themeasurement, treatment,andcalculationofTier 1, Tier 2,andTier 3

capital

* Risk coverage,theuseofstress tests,thetreatmentofcounter-partyrisk with

creditvaluationsadjustmentsthe use ofexternalratings, and the use ofleverage

ratios

(page202)

3 Explainchangesdesigned todampen theprocyclical amplificationof financial

shocks andto promotecountercyclicalbuffers,(page211)

4 Describechangesintended toimprove diehandlingof systemic risk (page212)

5 Describechangesintended toimprovediemanagement of liquidity risk includingliquiditycoverage rados, netstahlefundingratios, anddieuseof monitoringmetrics, (page213)

49.Basel III: InternationalFrameworkfor LiquidityRiskMeasurement,Standards, and

Monitoring

Candidates,aftercompletingdiisreading,should he ableto:

1. Defineanddescribe the minimumliquidity1'coverageratio, (page220)

2. Defineanddescribe thenetstahlefundingratio, (page223)

3 Define and describe pracdcal applicationsof prescribed liquidity monitoring tools,

including:

* Contractual maturity mismatch

* Concentrationof funding

* Available unencumberedassets

* Liquidity coverageratiohy significant currency

* Market related monitoring tools

(page 227)

©2013Kaplan, Inc.

Page10

Boot 3

Reading Assignments and AIM Statements

50 Revisions to die Basel II Market Risk Framework

Candidates,after completing thisreading, shouldbe ableto:

Descrihe theobjectivesfor revisingtheBaselII market risk framework,(page235)

2 Definedie capitalcharge for specific riskand general market risk,{page235}

3. Explain the relationshipregulators requirebetween market riskfactors used for

pricingversus chose usedforcalculatingValue-at-Riskandthe riskscaptured by the

Value-at-Risk model, (page236)

4 Explainand calculate diestressedValue-at-Risk measure anddiefrequencywhichit

must becalculated, (page237)

5 Explain and calculate die market riskcapital requirement,(page237)

6. Descrihe thequalitativedisclosuresfor the incremental riskcapital charge,{page

238)

7 Descrihe the quantitativedisclosuresfortrading portfolios under the internal

models approach, (page 238)

8. Descrihe theregulatory guidanceon prudent valuation of illiquid positions

(page238}

51 Operational Risk—SupervisoryGuidelinesfor theAdvancedMeasurement

Approaches

Candidates,aftercompletingdiisreading,should beable to:

1 Deline gross loss andnetlossandidentify which specificitemsshould beincluded

orexcludedin grosslosscomputationspet the Baselcommittee, (page244)

2. Descrihe the process andconsiderationssuggested bythe Basel committeefora

hank to useindeterminingaloss data threshold, (page245)

3 Describe thefour data elements which are required to compute abank’soperational

riskcapitalcharge per the BaselCommittee'sAMAframework,(page246)

4. Defineanoperational risk managementframework(ORMF)andanoperational

riskmeasurement system (ORMS) andexplain therelationshipbetween abank’s

ORMFanditsORMS (page247)

5 Describekey guidelinesfor verification andvalidationofabank'sORMF and

ORMS.(page248)

6. Descrihe keysupervisoryguidelinesfor the selection ofa reference date foran

internalloss,(page249)

7 Descrihe key guidelines for the selectionofabank'sOperationalRisk Categories

(ORCs).(pagje251)

8 Explain key guidelines formodelingthe distribution ofindividualORCs,including

theselectionofthresholds, necessaryadjustments,andselectionof statistical cools

andprobabilitydistributions,(page 252)

52.A ComparativeAssessmentof Basel1I/III amiSolvencyII

Candidates,aftercompleting thisreading, should be ahleto:

1 Contrastthe use of VaR parametersand confidenceintervals in theBaselII/IIIand

theSolvencyII frameworks,(page 258}

2 Explain thedifferencebetweenclassesof risks taken into accountin Basel 11/111and

SolvencyII.(page259)

3 Differentiate between solvency capital requirements (SCR) and minimumcapital

requirements (MCR), anddescribe the repercussions to an insurancecompany for

hreachingthe SCRandMCRunder theSolvencyIIframework,(page260}

1

ReadingAssignments and AIM Statements

5 Explain die difference between BaselIff Iff andtheSolvencyII frameworks widi

respect to: l) risk classes andcapitalrequirements,2) riskmeasureandcalibration,3) timeperspecdve,and4) valuation,(page262)

6. Compareandcontrast the Basel II/TITandSolvencyII frameworks with respect

toqualitadveriskmanagement aspects, includingthe internal risk management

process, governance, and supervision,(page263)

7 Describe the key differences between BaselII/IIIand Solvency II with respect to

publicdisclosure,(page266)

Page12 ©2013Kaplan, Inc.

The following is i review of the Operationÿ and Integrated Risk Management principles designed to address

the AIM statements set forth byGART®.This Lopk is also covered id:

PERFORMANCE MEASUREMENT

Topic34

EXAM FOCUS

This topic covers dieapplication of die risk-adjusted return on capital (RAROC}approach

to the allocation of economic capital The practical issues associated with the necessary

measurement of market, credit, and operational risks are discussed.The application of the

RAROCapproachtononloan productsisdescribed,andamodifiedversionof the tradidoual

RAROCapproachis presented.Forthe exam,understand therelationshipbetween economic

capitalandRAROCandknow how to compute RAROCandadjListedRAROC.

RISK-ADJUSTED RETURN ON CAPITAL

AJM34.1:Describe theRAROC (risk-adjusted return on capital) methodology

and describesomeof thepotential benefitsofits use.

The risk-adjustedreturn on capital(RAROC) measure isessential tosuccessful integrated

riskmanagement.Itsmainfunctionis to relate thereturn on capital to the riskinessoffirm

investments Themeasure promotes a consistentandunbiasedway to measureperformance

andprovides necessary informationto supportefficient risk-and-return decisions In the

AIMs tofollow,wewilldiscuss the useofeconomiccapital and how this capital levelis

incorporatedintothe risk-adjusted measure.Alsodiscussed is an adjustedRAROC measure

that providesa moreappropriate way toalign firm risks

ECONOMIC CAPITALAND RAROC

AJM34.2:Define,compareandcontrasteconomicand regulatory capital.

Economiccapital providesprotectionagainst risk (he.,unexpectedlosses).Itfurnishesan

institutionsvariousstakeholders withadegree of confidence that theirinvestedfundsare safe.

Itisimportant todistinguish betweeneconomiccapital andreserves Firms setasidereserves

in preparationfor expected losses.Economiccapitalisdesigned to provideacushionagainst

unexpectedlossesat aspecifiedconfidence level

The confidence levelatwhicheconomiccapital is set canbe viewedas theprobabilitythat

thefirmwill be ableto absorbunexpectedlossesover aspecified period- Forexample, ifa

banksets economiccapitalatdie 95%confidencelevel, diereis a95% chance thatactual

Cross Reference to GASP Assigned Reading—Croughy, Galai, & Mark, Chapter14

economiccapital None thatit is costprohibitiveforafinancial institution to operate atthe

100% confidence level

Theamountof regulatory capital that ahankisrequired to holdisdetermined by

regulatory guidelines,whicharedesigned to assure thereissufficientcapitalin thebanking

system.Theeconomiccapitalheld bymostfinancialinstitutionsexceedsthe requiredamountofregulatorycapital

Professor'sNote:Wewillexaminetheregulatorycapitalchargesforcredit, market,

andoperationalriskin the BaselReferenceReadings,

Economiccapital is important from the perspective of the firm’sstakeholders.Theamount

ofeconomiccapital thata firm holds and the allocation ofeconomiccapitalamongits

businesslineshave a profound effectonbusiness and overall firm performance

AIM34.3: Computeandinterpret the RAROCforaloan orloan portfolio,and

useRAROC tocompare businessunit performance.

The necessaryamountofeconomiccapitalis afunctionof credit risk, market risk, and

operating risk TheRAROCfora loan canbe defined asrisk-adjustedreturndivided byrisk-adjustedcapital.Asyou willsee later,economiccapitalcan beusedas a prosy for risk-

adjustedcapital

The relationship betweeneconomiccapital and RAROCcanbeexamined by consideringa

$10t)million loan with the followinginitial assumptions:

* Expected loss(EL)-lOObp

* Economiccapital required million

Unexpected lossat ahighconfidence levelfor this loan = [worst-caseloss(i.e.,VaR)

-expectedloss] x loanamount.

TheRAROCfor this loanis therisk-adjusted returndividedby risk-adjustedcapital,on.

revenues—expectedloss—expenses+return on economiccapitaldrtransfer priceRAROC=

economiccapital

Thenumerator in die RAROCequationincludes the gross revenuesfrom the loan less die

expectedloss andocherloan expenses.Economiccapitalisoften investedin high-qualityliquidsecurities, sodie return oninvestedeconomiccapitalmust beadded toexpected loan

revenues Also,anadjustmentismadefor relevant operating revenuesorexpensesassociated

writh theloan.

ProfessortNote:Onpreviousexams, therisk-adjustedreturn oncapitalhas been

measuredsimplyas:profitlriskcapital Bepreparedto useVaRas aproxy forrisk

capital ifa measureofeconomiccapitalis notprovided.

Cross Reference to CARPAssignedReading—Cioughy, Galai, & Mark,Chapter l4Example:ComputeRAROC

Given diefollowingassumptionson a$100 million loan, calculate theRAROC.

* Grossrevenue of$7million

• Interest expense of$5 million

* Returnoninvestedeconomiccapital of$350,000.

* Operatingcost (to thebusiness unit makingdieloan) of$250,00(1.

* Expectedloss(EL)= lOObp

* Economiccapital required= $Smillion

Answer:

Firsr,calculate theexpectedloss EL= 0.01 x $100,000,000=$1,000,000.Then, apply

theRAROC equation:

$7-$1—$5-j-$0.35-$0.25

$8The 13.75%RAROCcomputed herecan he viewedas the required return on theequity

used to support rhe loan

CAPITAL ATTRIBUTEDTO MARKET, CREDIT, ANDOPERATIONAL RISK

AJM34.4: Explainhowcapital is attributedto market, credit,and operational risk

AIM34.5:Calculate thecapital chargefor market risk and credit risk

AJM34.6: Explainthedifficultiesencounteredinattributingeconomiccapital to

operationalrisk

Market riskisdie riskof lossas a resultofchanges in market riskfactors Credit riskis the

riskoflossassociated with changesin thefactors thataffect the credit qualityofail asset,

andoperational risk istheriskofloss due tofactorsthatlead tooperational failure, such as

humanerror,technology change,computercrashes, andregulatory changes Manybanks,

at thevery least, measureand manageriskin thesethree categories.Measurementof

risk-adjustedreturnsfor banks is aformidablechallenge

The majorsourceof market risk fora hank that arisesfrom interest rateriskiscalledgjtp

risk, die risk inherentin die mismatch betweenabank'sinterest-rate-sensidve assetsandits

interest-rate-sensitiveliabilities.Gaprisk and other types of market riskmustbe considered

when constructingrhe loss distribution used in RAROCcomputations

Cross Reference to CARP Assigned Reading—Croughy, Gal si, & Mark, Chapter14

CapitalAttributed toMarket Risk

RAROCcapitalallocationfor market risk involvesattributingRAROCcapitalin termsof

the amountof riskindie computation ofvalue at risk(VaR).RAROCmarket riskcharges

areoften madeondie basisofunusedmarket risk limitsand excesses over these limits.For

example,givenaVaRestimate at aspecified confidencelevel, die market riskcapital charge

can he expressedas:

market risk capital charge-F,(VaR)+ F2[max(VaR limit-VaR, ()}] +

F3[max(VaR —VaR limit,0)]

where:

F1 =a constant diatadjustsfor theday-to-dayeventrisknot capturedin the VaR model

F3 = multiplier usedtodetermine die charge for theunusedportionof theVaR limit

= multiplierusedto determinethe charge forexceedingthe VaRlimit

The way die formulais constructed,eidierF3[max(VaR limit- VaR, 0)] orFÿ[max(VaR

-VaR limit, 0)jisequal to zero.If theVaRlimit hasnotbeen exceeded,there is achargefor dieunusedportion of the limit,andthe thirdterm is zero.If the VaR limit has beenexceeded, thereis achargeforexceedingthe limit, and die second term is zero.

Example: Capitalcharge for marketrisk

Suppose theVaRlimit atdie 99%confidencelevelis $100,000,FL =2.00, F3 = 0.20,and

Fj= 4.00.Computethe market riskcapitalchargeif theVaRis:

• $80,000(the VaRlimit hasnotbeen exceeded).

• $150,000 (theVaRlimit has been exceeded)

Professor$Note:This methodofattributingcapitalissimilartothe Internal

Models Approach (IMA)discussedin the Baselreadings NotethatIMA isusedfor

allocatingregulatorycapitaltomarketrisks

©2013 Kaplan,Inc.

Pagelb

Cross Reference to GARPAssigned Reading-Cioughy; Galai, & Mark,Chapter\4CapitalAttributed toCredit Risk

Theprocess used to attributecapitaltocredit risk employsstandardizedcapitalfactors,

whichexpress theamountof credit riskas afunctionof rating andmaturity*At agiven

rating, thecapitalfactorincreases as maturityincreases.Similarly, at agiven maturity, the

capital factorincreases ascreditquality decreases* Thecredit capital charge for credit riskis

determinedas:

creditrisk capitalcharge=capital factor xmarket valueof position

Capital factors may beobtainedfrom the ratingagencies,or by usingproprietary external

models suchasKMV®and publicly available models suchasCreditMetrics®*

CapitalAttributed toOperationalRisk

Relative to market risk andcredit risk,operationalriskis themostdifficultto measure

While manysophisticatedbankshavedevelopedsound methodologiesforquantifying

market and creditrisks,operational risk measurement remains more an artdiana science,

Thisi*s amajor concernbecauseoperational risksoften represent an enormouspotentialloss

to afinancialinstitution

Success inapplyingVaRconcepts tooperational riskmeasurementhas been limited because

internal data pointsare usuallytoofewtobuild the necessary lossdistribution.Another

simpler procedure forallocatingoperational risk capitalis toassignoperational riskratings

to businesslinesor transactionsbasedondiefactors that lead tooperationallosses (i.e.,

people,processes, and technology)*Firm-wideoperationalriskcapitalcan then be allocated

to die individualbusinesses or products basedon a rankingof their operational risk ratings,

Professor'sNote:Topics 4land42will address theprocess ofgeneratingan

operationalloss distribution and calculating valueatriskfromthat distribution,

OperationalVaRcombinesexpectedandunexpectedlosses andhelpsthe bank

determine the levelofeconomiccapital

CAPITALFOR NONLOAN PRODUCTS

AIM34.7:Describe the LoanEquivalent Approach anduseittocalculateRAROC

capital

Tnaddition tostandardloans, manybanksoffer odierproductstowhich capital must

be allocated* Indoingso, it is useful to thinkof the risks of diese productsin termsof

their loan equivalencies The general approach forallocating RAROCcapital for nonloan

productsis tomultiply their loanequivalentvaluebythestandardizedcapital factors

discussedin the previoussection*

Topic 34

Cross Reference to GARP AssignedReading-Croughy,Gal id, & Mark,Chapter 14

FIRST-AND SECOND-GENERATION RAROC

AIM34.8: Explain how thesecond-generationRAROCapproaches improve

economiccapital allocationdecisions

AIM34.9: ComputetheadjustedRAROCforaproject todetermine itsviability

Manybanks use theRAROCapproach toallocateeconomiccapital by calculating the

RAROC fora businessunit or productandcomparingit to a preset RAROC hurdlerate.

Theargumentfor using this methodologyis dial onlyprojectsthat providea RAROC

above thehurdleratemakea positivecontributiontoshareholder wealth.Unfortunately

thisapproachcanactuallylead todecisions thatdecrease shareholderwealth

Under the traditional(first-generation) RAROCapproach, theRAROCfora business

iscompared tothe firm'scostofequity.If dieRAROCexceedsthecostofequity itis

concludedthat die businessaddsvalueto the firm The flawin thefirst-generadonRAROC

approachis itsassumpdon that thedefault probability of risky investmentsremains

constant.This assumption is inconsistent widia constantexpected return on the firm’sequitybecauseinorderfor theprobabilityofdefault to remain constant,the firm’s return

on equitymust change. Likewise,if diereturn on die firmsequityis to remain constant,dieprobabilityof default mustchange

Professor’sNote:Stateddifferently, thefirstgenerationRAROCpicksa constant return onequitytoevaluate whetherprojectsadd value.Asindicated\ thisapproach

assumes a constantprobabilityof default.However, asthe riskofthe businesschanges, theprobability of defaultwill change, so a constant return on equitydoes

not ensure a constantprobabilityof default.

Asecond-generation RAROC methodologyhasbeendeveloped to overcome the inherent

problem with the first-generation approach Themaingoal of thesecond-generationmediodologyis toalign die risk of the business widi the risk of the firm’s equity Under the

second-generationapproach,anadjustedRAROC (ARAROC) iscomputedas follows:

(RAROC -RF) ARAROC- 3E

where:

PE -systemadcriskof die firm'sequity

Rp = risk-freerateofreturn

AninvestmentwillincreaseshareholdervalueifARAROCexceeds die difference between

the market return,i?M,and the risk-freerate,whichisoften referred to as the market risk:

premium Thus,die decision ruleis to accepttheprojectifARAROC>RM — Rp.

©2013 Kaplan,Inc.

Page18

Cross Reference to CARPAssignedReading—Croughy, Galai, & Mark,Chapter 14Example:Adjusted RAROC

SupposeRAROC is 12%, the rish-ftieerate Is5%,the marketreturn is 11%,and the firing

equitybetais1.5.Use ARAROCtodetermine whether the projectshould he accepted

The projectshould berejected because theARAROCof4.7%isless than theexcess

return ondie market:11%—5% =6%

Cross Reference to CARP AssignedReading-Croughy, Galai, & Mark,Chapter l4

KEY CONCEPTS

AIM34.1The risk-adjusted return oncapital(RAROC) measure isessential tosuccessful integrated

riskmanagement.Itsmainfunctionis to relate thereturn oncapital to theriskinessof firm

investments.

AIM34.2

Economiccapital providesacushionagainst unexpected lossesata specifiedconfidence

level.Regulatory capital Isdeterminedhy regulatory guidelines to protect thehanking

system.

Theamountand allocationofeconomiccapitalsignificantlyaffects firm performance

AIM34.3Risk-adjusted return oncapital(RAROC) =

revenues-expected loss— expenses+return on economiccapital±transferprice

economiccapital

AIM34.4

RAROCcharges for market riskareoften madeonthe basisof unused market risk limits

andexcesses over theselimits.Attributing capital tocredit riskemploysstandardized

capitalfactors, which express theamountof credit risk as afunctionof ratingandmaturity

Attributingcapital tooperational riskassigns operational riskratingstobusinessfines or

transactionsbasedon thefactors diatleadtooperationallosses

AIM34.5

Given aVaRestimate at aspecifiedconfidencelevel, market risk capitalcharge

-FL(VaR)+ F [max(VaR limit- VaR,0)] + F3[max(VaR-VaR limit, 0}].

Creditcapitalcharge = capital factorx market valueofposition

Operational riskcapital maybe allocated usingarankingsystem.

AIM34.6

While manysophisticatedbanks havedevelopedsoundmethodologies forquantifying

marketand creditrisks,operational riskmeasurement remains more an art than a science This is a majorconcern becauseoperationalrisksoftenrepresent anenormouspotentialloss

to afinancialinstitution.

©2013Kaplan,Inc.

Page20

With thefirst-generationRAROCapproach,abusiness addsvalue tothefirm if the

RAROCofaproject exceeds tire firm'scostof equity The flaw in thefirst-generation

RAROCapproach isitsassumption that the probability of default remains constant.

Cross Reference to GASP Assigned Reading—Croughy, Galai, & Mark, Chapter14

CONCEPT CHECKERS

Whatisdie capitalchargeattributed tocreditriskforaloanvaluedat$1 million

andacapital factorequalto3-24%?Assumediat die loaninterest rate is5% and

that dieloan maturesinthreeyears

A. #32,400,

B #318,417

C #986,617

D* #1,032,400

Assume thataloan has thefollowingcharacteristics:

* Gross revenue isexpectedto be#3.0 million

• Interest expenseis$3.0million

* Expected return on the#6.0 million ineconomiccapitalis#175,000

• Expectedlosson the loanis#250,000

• Othercostsassociatedwithmakingtheloanequal#1,0 million

Whatisdiecapitalchargefor market risk if:

• Theappropriateadjustment factorforthe day-to-dayevent risk diatis not

captured inVaRis1.75

• The multiplier usedtodetermine the charge for theunusedpordonof theVaR

limitis 0*20.

• Themultiplierused todetermine thechargeforexceedingthe VaRlimitis2*75

• TheVaRlimitis#750,000

Self-TestQuestions: # 1(page272)

PastFRM Exam Questions:# 1ÿ4(page278)

Cross Reference to GARP AssignedReading-Croughy, Galai, & Mark, Chapter l4

CONCEPT CHECKER ANSWERS

1 A capitalcharge=capitalfactor x market value of position=0,0324 x $1,000,000-$32,400

(revenue— EL— cxpcnscs+ return on economiccapital=L transfer price)

3 B Unexpectedloss is the loss at aspecifiedconfidence level (vmrst-ca.se loss) minusexpected

loss In this ease, unexpectedloss=200—50=150 basis points.

4 C The market riskcapital chargeis:

F,(VaR) +Fornax(VaRlimit-VaR, 0)] + [ max(VaR-VaR limit, 0)1

where:

Fj =a constant thatadjustsfor theday-to-dayevent risk notcapturedin the VaR model

=themultiplierused todeterminethechargefor the unused portion of the VaRlimit

Fj =themultiplierused todeterminethechargefor exceeding the VaR limit

Thus, thecapital chargefor market risk is;

=systematic risk of the firm’s equity

Rj = risk-free rate of return

The fuUrtwblg W 1 review of tJie Opctiuioiull and Integrated Sist \f iciÿelneliL printiplea designed Lo iddicSs

the AIM stiiemeitLS tcLfurlt l)y GARP®, Thisuipitis alst> covered id:

ECONOMIC CAPITAL MODELING

Topic 35

EXAM FOCUS

This topic requiresanunderstandingof many riskmanagement conceptsdiatyouhavealready

coveredatFRMPartI, aswellasinearlier readingsintheFRMPartITcurriculum.Specifically,

this topicexpandson theconceptofeconomiccapital, which isdiecapital required toabsorb

unexpectedlossesforagiventimehorizon and confidence interval For theexam,payattention

to the terminologyandattempt tointegrate diis materialto the sections pertainingtomarket

risk andcredit riskso as toreinforce your understanding

ECONOMIC CAPITAL IMPLEM EN TATION FRAMEWORK

AIM35.1:Within theeconomiccapital implementation framework describethe

challenges thatappearin:

+ Definingriskmeasures

Risk aggregation

+ Validationof models

+ Dependency modelingin credit risk

Evaluatingcounterpartycredit risk

Assessinginterest rate riskin thehankinghook

For thisAIM, itwould hehelpful torecall die properties ofa coherentriskmeasurefrom

thePartIcurriculum Theproperties areasfollows:

I. Monotonicity: A portfolio withgreaterfuturereturnswill likely have less risk

2 Subadditivity:Theriskofaportfolio is at mostequal tothe riskof theassetswithin the

portfolio

3 Positivehomogeneity:Thesizeofaportfoliowill impact thesizeofitsrisk

4 Translationinvariance:The risk ofa portfolioisdependenton theassetswithin die

portfolio

DefiningRisk Measures

Itis notalwaysapparent how riskshould bequantifiedforagivenbank, especiallywhen

Topic 35

Cross Reference to GARP AssignedReading—Basel Committee on Banking Supervision

should be:intuitive,stable, easy to compute,easyto understand, coherent,andinterprecable

in economic terms.In addition, die risk decomposition process muse hesimpleand

meaningful foragiven riskmeasure.

Standarddeviation, valueat risk(VaR),expected shortfall(ES), as wellas spectral

(i.e.,coherent) anddistorted risk measures could be considered, each with theirrespective pros andcons.Obviously,

necessaryelements in measuring risk In practice, VaR andELSare the mostcommonly

used measures.The followingsection is asummary ofchallenges encounteredwhen

consideringtheappropriatenessof each risk measure

would perfectlyconsider all of the

no one measure

Standarddeviation

* Notstablebecauseitdependsonassumptionsaboutthe lossdistribution

• Notcoherent becauseitviolatesthe monotonicitycondition.

* Simple, but notverymeaningfulin the riskdecomposition process

VaR(the mostcommonlyused measure)

• Notstablebecauseitdependsonassumpdonsaboutthe lossdistribution

* Notcoherentbecauseitviolatesthesuhaddidvitycondidon (couldcauseproblemsin

interna]capitalallocationandlimitsetdng forsub-portfolios)

Expectedshortfall

• Mayor maynot be stable,dependingon the loss distrihudon

• Not easytointerpret,anddielinkto thebank’sdesired target radngis notclear

Spectraland distorted risk measures

• Notintuitive noreasilyunderstood (andrarely used in practice)

* Mayor maynot be stable,dependingon the lossdistrihudon

Indefiningor usingsuchrisk measures,banksoftenconsiderseveralof themandfor

different purposes Forexample,absolute risk andcapital allocation within the bank

are mostcommonly measured usingVaR,but increasingly, the latter isbeingmeasured

usingES.The VaR measureof absolute risktendsto be easier tocommunicate tosenior

management than ES,hutES is a morestablemeasure thanVaR forallocatingtotalportfolio capital The challenge for die bankis todetermineifandwhenone or theocher,

orboth,should beused.

Amongst thecommonly used measures tocalculateeconomiccapital, regulatorsdonothave

aclear preference forone overanother Ifdifferent risk measures areimplemented byabank

for external versus internal purposes, then theremustbealogicalconnection between die

two risk measures Forregulators,merely comparingabank’s internaland regulatorycapital

amounts is insufficientwhendeterminingthe underlying risks initsportfolio.Therefore,such ataskpresents ananalytical challenge to regulators

©2013Kaplan,Tnc.

Page26

Cross Reference to GARP Assigned Reading—Basel Committee oil Banking Supervision

RiskAggregation

Riskaggregationinvolves identifying the individual risktypesand makingcertain choices

inaggregatingthose risktypes.Classification byrisk types (market,credit,operational,and

business) maybeapproximateand proneto error. Forexample, the definitions of risktypes

may differacrossbanksorwithinagivenbank, whichcomplicates theaggregation process

Even dioughone or moreofdiepreviouslymentioned four risk typesmay be foundatdie

same timewithin agiven bank portfolio, die portfoliowilloften be represented byone

risk typefor thebanks classifications purposes.Suchasimplistic distinction mayresultin

inaccurate measurementsof the risktypesanddiis may bias the aggregation process

Most batiksbeginbyaggregatingriskintosilosby risk-typeacrosstheentirebank Other

banksprefer usingbusiness unitsilos,while others combine both approaches Thereis no

one unanimously accepted niediod,aseachapproachhasitsspecific advantages

Beforerisk types can beaggregatedinto asinglemeasure, theymust beexpressedin

comparableunits.Therearedireeitems toconsider: risk metric,confidence level, and time

horizon

1 Riskmetric: Relieson themetrics usedin the quantification of different risktypes.

Must consider whether the metricsatisfies thesubadditivitycondition

2. Confidencelevel: Loss distributionsfordifferenttypesof riskareassumedto

have differentshapes, whichimpliesdifferencesin confidence intervals.Thelack

ofconsistencyinchoosingconfidencelevelscreatesadditional complexityin the

aggregationprocess

3 Timehorizon:Choosing the riskmeasurement timehorizonis oneof themost

challengingtasksinriskmeasurement.Forexample, combining riskmeasures that have

been determinedusing different timehorizonscreatesprohlemsirrespective ofactual

measurement methods used Specifically, there will beinaccuratecomparisonsbetween

risk types.

Acommon belie!is thatcombiningtwoportfolios will result inlower risk perinvestment

unit inthe combined portfolioversusdieweightedaverage of thetwo separateportfolios

Howreyer, when weconsider risk aggregationsacrossdifferent portfoliosor business

units,such a belief doesnot hold up with VaR becauseitdoesnotnecessarily satisfy the

subadditivitycondition.Also, theremaybeafalseassumptionthat covariancealways fully

takesinto accountthedependenciesbetween risks.Specifically, therecouldbetimeswhere

the riskinteractions are suchthat dieresultingcombinationsrepresenthigher,notlower,

risk Thesepointshighlightanadditional challengein thecomputationofrisk.

Therearefive commonly used aggregationmethodologies.Thefollowingis a brief

descriptionof them,as wellas diechallengesassociated with using them.

1. Simple summation

* Addingtogetherindividual capitalcomponents.

Cross Reference to GARP Assigned Reading-Basel Committee on BankingSupervision

• Doesnotdifferentiate between risktypesand therefore assumesequal weighting

Also, doesnot takeinto accountthe underlyinginteractionsbetween risktypes

orfor differences in thewaythe risktypesmaycreatediversification henefits.Inaddition,complicationsarising from using different confidence levelsareignored

2* Constantdiversification

• LSame processassimple summation except thatitsuhtracts afixed diversificadon

percentagefrom the overallamount.

• Similar challengesassimplesummation.

3 Variance-covariance matrix

• Summarizestheinterdependenciesacrossrisk typesandprovidesaflexibleframeworkfor recognizing diversification benefits

• Estimatesof inter-riskcorrelations (abank-specificcharacterisdc)aredifficultand

costlytoobtain, and the matrixdoesnotadequatelycapture non-linearitiesand

• Themostdemanding methodintermsof requiredinputs Also, therearehigh

information technologydemands,dieprocessis timeconsuming, andit may provide

afalsesenseof security

Thevariance-covarianceapproachiscommonlyused bybanks Frequendy, however,

bank-specificdatais notavailableor is ofpoorquality.Asaresult, theitemsin the covariance matrix arecompletedonthe basisofexpert judgment.On a related note,banks

variance-often usea“conservative”variance-covariance matrix where thecorrelations are reported to

beapproximateand biased upward.Inorder to reducetheneed forexpert judgment, hanks

mayend uplimitingthedimensionalityof the matrixandaggregating risk categoriessothat thereareonlya few ofthem, not recognizing that such aggregationsembedcorrelation

assumptions.Clearly, adisadvantageofsuch a practice is thateachcategorybecomes lesshomogenousand therefore,morechallenging toquantify

Onepotentialdisadvantageof themoresophisticated methodologiesis that theyoften lead

to greaterconfidencein theaccuracyof theoutput.Itis important toconsider robustness

checks andestimatesofspecification andmeasurement error so as to prevent misleading

results

Validationof Models

Validation is the‘'proof thatamodel worksasintended.Asan example, whileitis a useful

tool to test a model’s risk sensitivity, it isless usefulfor testing the accuracy ofhigh quantiles

inaloss distribution

©2013 Kaplan,Inc.

Page 28

Cross Reference to GARP AssignedReading—Basel Committee on Banking Supervision

The validationofeconomiccapitalmodelsdiffersfrom die valuation ofanIRB

(internal-ratings based) model because theoutputofeconomiccapital modelsis adistribution radier

than asingle predicted forecast against which actualoutcomesmaybe compared.Also,

economiccapital modelsarequitesimilar toVaR modelsdespitethelonger time horizons,

higherconfidencelevels,andgreaterlackof data

Thereare sixqualitativevalidation processestoconsider.Thefollowingis abrief description

of them,aswellas thechallengesassociatedwithusingthem (whereapplicable)

1 Usetest

* Ifabankuses its measurement systemsfor internal purposes,then regulatorscould

place morerelianceon dieoutputsfor regulatory capital

* Thechallengeisfor regulatorstoobtainadetailedunderstandingofwhich models

propertiesarebeingusedand whichare not.

2. Qualitativereview

* Mustexaminedocumentationanddevelopmentwork,have discussions with the

model’s developers,test and derivealgorithms,and compare with other practices and

known information

* Thechallengeis toensure that the model worksin theoryand takesinto accountdie

correctriskdrivers.Also, confirmationof the accuracy of the mathematics bellind

themodel isnecessary

3 Systemsimplementation

* Torexample, user acceptance testing andcheckingofcodeshould bedonepriorto

implementation to ensureimplementation of the modelisdone properly

4 Management oversight

* Itisnecessary tohave involvementofsenior management inexamining theoutput

datafrom themodeland knowinghow to usethe datatomakebusinessdecisions

* Thechallengeisensuringthat senior management is awareof how the model isused

and how the modeloutputs areinterpreted

5 Dataquality checks

* Processes to ensurecompleteness, accuracy,andrelevanceof data usedindie

model.Examplesinclude:qualitativereview,identifyingerrors, and verificationof

transactiondata

6 Examinationofassumptions—sensitivity testing

• Assumptionsinclude:correlations, recoveryrates,andshapeof taildistributions

Theprocessinvolvesreviewingtheassumptionsandexaminingtheimpacton model

outputs.

Thereare also sixquantitativevalidation processes to consider Thefollowingis a brief

descriptionof them,as wellas thechallengesassociated with usingthem (whereapplicable)

1 Validationofinputsandparameters

• Validating inputparametersforeconomiccapital models requires validation of those

parameters notincludedin the IRB approach,such ascorrelations

Topic 35

Cross Reference to GARP AssignedReading—Basel Committee on BankingSupervision

• Thechallengeisthatchecking modelinputsis not likelytobefullyeffective becauseeverymodelishasedonunderlyingassumptions Therefore,the morecomplex the

model, themorelikelythere will be modelerror.Simply examininginputparameters

willnot preventthe problem

2 Model replication

• Attempts toreplicatethe model resultsobtained bythe bank

• Thechallengeis that die processisrarely enoughtovalidatemodels andin practice,

thereislittle evidenceofit beingused by banks Specifically, replicationsimplyby

re-runninga setofalgorithmstoproducethesame setofresultsis notconsidered

enough modelvalidation

3 Benchmarking andhypothetical portfolio testing

• The processiscommonly used and involvesdeterminingwhether the model

producesresultscomparable to astandard modelorcomparing modelson a setof

reference portfolios

• Thechallengeisthat die processcanonlycompareonemodelagainstanother andmay provide little comfortthatthemodelreflects ‘‘reality.*' All that theprocessis

abletodois providebroad comparisonsconfirmingthat input parameters or model

outputs arebroadlycomparable

Obviously, there will be risk measurement systemswhoseoutputs cannotbe

in terpreted thisway Also,backtestingis not yet amajor partof banks' validation

practices foreconomicpurposes

5- Profit and loss attribudon

• Involves regular analysis of profit and loss—comparison betweencausesofactual

profitandloss versus the model's risk drivers

• The challengeisthat theprocessis notwidely usedexcept for market risk pricing

models

6 Stresstesting

• bivalves stressing the model andcomparingmodeloutputs to stresslosses

Overall,although thesevalidationprocesses may behighlyeffectivein areassuch as risksensitivity, theymaynot be effectiveinareassuchasoverall absoluteaccuracy

Additionally, thereisdifficultyinvalidating theconceptualsoundnessofacapital model

Thedevelopmentofa model almostalwaysrequires assumptionsto be made.However,

someof die assumptions maynot be testable,so itcouldbeimpossibletobeabsolutely

certainof the conceptualsoundnessofamodel.Even thoughthe underlying points mayappear reasonableandlogical, diat maynot be diecase in practice

Fromaregulator’sperspective,some industryvalidation practicesareweak, especiallyfor total capital adequacy of the bank and die overall calibration of models.Sucha

©2013 Kaplan,Inc.

Page 30

Topic 35

Cross Reference to CARP AssignedReading—Basel Committee on Banking Supervision

validation projectischallengingbecauseitusually requitesevaluation of highquantiles of

lossdistributionsoverlongperiods oftime.In addition, therearedata scarcityproblems

plus technicaldifficulties,such as tailestimation.Therefore, it isimportantforsenior

managementand model users to understand dielimitationsofmodels and the risksofusing

models that havenotbeen fullyvalidated

DependencyModelinginCredit Risk

Modeling the dependencystructure betweenborrowers iscrucial,yetchallenging.Both

linearand nonlinear dependencyrelationships betweenobligorsneed to beconsidered

Ingeneral,dependenciescan bemodeled using: credit riskportfoliomodels, modelsusing

copulas,and models basedon theasymptoticsingle-riskfactor(ASRF) model.With die

ASRF approach,banks may use their ownestimates of correladonsor may usemuldple

systematic riskfactors toaddressconcentradons.Suchan approachwould result in

questioningthe method used to calibrate the correlationsand theways inwhich die bank

addressed the infinitegranularityandsingle-factorstructureof theASRF model

Therearemany issues toconsider regarding thechallengesincomingupwith reliable

dependencyassutnpdons usedincredit risk portfoliomodels.Regulatorsmay need to test

theaccuracy andstrengthof correladonestimates usedbybanks given theirheavyreliance

onmodel assumptions and thesignificantimpacton economiccapitalcalculations

In diepast, thevalidityof thefollowingassumptions havebeen questioned:(1) theASRF

Gaussiancopula approach,(2)the normaldistributionforthevariablesdrivingdefault,

(3) thestability of correlationsover time,and(4) the joint assumptions ofoorreedy

specified defaultprobabilitiesanddoubJy-stochasticprocesses,whichsuggest thatdefault

correlation issufficiently captured bycommonriskfactors

Doubts have beenraisedabout the ability ofsome modelsusingsuchassumptionsin terms

of theirabilitytoexplain the time-clusteringofdefaultsthatis seen in certainmarkets

Insufficientlyintegratingthe correlationbetween probability" ofdefault (PD)and loss

givendefault (LGD)in the models,coupled withinsufficiently modelingLGD variability,

mayleadtounderestimating thenecessaryeconomiccapital.Furthermore, it will create

challengesin identifyingthe different sources of correlationsand theclusteringofdefaults

andlosses

Ratingchangesaregreatlyimpacted bythe business cycleandareexplained bydifferent

modelsduringexpansionaryandrecessionaryperiods.Asaresult, the sample periodand

approach used tocalibrate thedependencystructurecould be importantinassessing

whether correlationestimates areoverestimatedor underestimated.Furthermore,some

modelsassumethat unobservableasset returnsmaybe approximated by changesinequity

prices but fail toconsider that therelationshipbetweenasset returnsand equity pricesare

unobservableandmaybe non-linear.Also, theuseofequity pricesto estimatecredit default

probability isproblematicbecause suchprices may include information chat isirrelevantfor

credit risk purposes Asaresult, usingequity' prices mayresult insomeinaccuracy in the

correlation estimates.

Cross Reference to CARP AssignedReading-Basel Committee on BankingSupervision

In contrast, whenbanks usearegulatory-typeapproach,the assumptions of such an

approach createother challenges for both banks and regulators:

• Correlation estimatesneedto be estimated,but there may be limited historical data

onwhichtobase the correlationestimates.Also,theassumptionsusedto generatethe

correlations maynot heconsistentwith theunderlyingassumptions of the Basel II credk

riskmodel

* A hanks useofthe BaselIIriskweight modelrequiresconcentration risktobe accounted

forbyother measuresand/ormanagementmethods Itwillalso require regulatorsto

evaluate suchmeasures/methods

Akey challengeto overcome isthe useofmisspecifiedorincorrecdycalibratedcorrelations

andthe useofanormaldistribution (whichdoes notreplicate thedetails of thedistribution

ofasset returns).This mayleadtolargeerrors in measuring portfoliocredit riskand

economiccapital

Evaluating CounterpartyCredit Risk

LSUCIIacaskis asignificant challenge becauseit requires:obtainingdatafrom multiple

systems,measuringexposures froman enormousnumberoftransactions (includingmany

that exhibit optionality)spanning awide range oftime periods,monitoringcollateraland

nettingarrangements,and categorizing exposuresacrossmany counterparties.As a result,

banks needtohave well-developedprocessesandtrainedstafftodealwith diesechallenges

Market-risk-relatedchallenges to counterpartyexposureatdefault(EAD)estimation.

• Counterparty credit exposure requiressimulationof marketriskfactors and the

revaluationofcounterpartypositions undersimulated riskfactor shocks, similar toVaR

models.Consider thefollowingtwochallengesdratoccurwhenattemptingto useVaRmodel technologyto measure counterparty creditexposure

* Market riskVaRmodelscombine allposidons inaportfoliointo asinglesimulation

Therefore, gains fromoneposidon mayfullyoffset dielosses inanother posidonin

thesamesimulation run However, counterpartycreditriskexposuremeasurement

does notallow nettingacrosscounterparties Asaresult, it isnecessary to compute

amounts atdie netdngsetlevel (on eachsetoftransactions that form the basis ofa

legally enforceablenetdngagreement), whichincreasescomputational complexity

* Market riskVaRcalculationsareusually performedforasingleshort-term holdingperiod.However, counterpartycredit exposuremeasurement must heperformedfor multiple holding periods into thefuture.Therefore,marketriskfactors need

tobesimulatedovermuchlonger dmeperiods than inVaR calculations,and the

revaluationof the potentialexposure inthefuturemustbedonefor theentire

portfolioat certain pointsinthefuture,

Credit-risk-relatedchallengestoPD andLGDestimation.

* Somematerial transactions areperformedwith counterparties with which the hankdoes

nothaveanyotherexposures Therefore,the bankmustcalculateaprobability ofdefault

(PD)and lossgivendefault (LGD)for thecounterparty and transaction.

* Forhedgefunds,diemeasurementchallengeoccurswhen there islittle informationprovidedonunderlyingfund volatility, leverage, or typesofinvestmentstrategies

employed

©2013 Kaplan,Inc.

Page 32

Topic 35

Cross Reference to GARP Assigned Reading—Basel Committee oil Banking Supervision

* Evenforcounterpartieswithwhichdie bank hasother creditexposures,die bank still

needs to calculate aspecificLGDfor die transaction.

Interactionbetween market risk and credit risk—wrong-way risk

• Identifyingandaccountingforwrong-wayrisk(exposuresthatarenegatively correlated

with die counterparty’s creditquality) is asignificant challenge because it requiresan

understanding of themarket riskfactors towhich thecounterparty isexposed.That

would bedifficulttodoin thecaseofahedgefund,for example, which would be less

transparent,

• Italsorequiresacomparisonof those factorsensitivities tothefactorsensitivitiesof the

hank’sownexposurestothecounterparty,

* Themagnitudeof wrong-way riskisdifficulttoquantify in an economiccapital model

since itrequiresalongtime horizonat ahighconfidence level

Operational-risk-relatedchallengesin managingcounterpartycredit risk

* Thechallengeis thatmanagingsuch risk requiresspecializedcomputer systems and

people Complicated transactions,suchasdailylimitmonitoring,marking-to-niarket,

collateralmanagement,andintraday liquidityand creditextensions, increase the riskof

measurement errors.

* The quantification of operational risksis asignificant challenge,especially whenit

pertainsto new orrapidlygrowing businesses, newproductsor processes,intraday

extensionsofcredit,and infrequendyoccurring hutsevere events.

Differencesinriskprofilesbetween marginedandnon-margtnedcounterparties

• Themodeling differencebetween thetwo typesof counterpartiesisprimarilyconcerned

with thefuture forecastingperiod Formarginedcounterparties, theforecasting periodis

short,andfor non-marginedcounterparties, it isusually muchlonger

• As aresultof thedifferencein timeperiods, theaggregationof riskbetween thesetwo

typesofcounterparties Isachallenge because theusual procedureis to use asingletime

periodfor all positions

Aggregationchallenges

* Ingeneral, thechallengesareincreased significandywhen moving frommeasuringcredit

riskofone counterparty tomeasuring credit risk of the firmingeneralforeconomic

capital purposes

• Whencounterpartieshave both derivativesand securities financingactivities,the

problemisespeciallychallengingbecausethe systems in placemaynot beabletohandle

suchaggregation

* Further aggregationchallengesexist when high-level credit riskmeasuresarerequired

to beaggregatedwithhigh-levelmarket risk andoperational risk measuresinorderto

calculateeconomiccapital

* Breaking downcounterpartycredit riskintodetailedcomponent parts (as isoften

donewith market risk)isanotherchallenge.Thesheercomputational complexitiesand

enormousamountsofdata required would generallybe costprohibitive toperformon a

frequenthasis Thechallengestillremainsfor many hanks duetooutdatedorineffective

computer systems.

Topic 35

Cross Reference to GARP AssignedReading—Basel Committee on Banking Supervision

Assessing Interest Rate Riskin theBankingBook

The computationchallengearisesfrom tirelongholdingperiodassumedforabank’sbalance sheetandthe need tomodel indeterminatecashHow'sonhodr theassetand liabilitysideduetodreembeddedoptionality ofmanyhanking hookitems.

Optionalityin thebankingbook

* A majormeasurementchallengeisfound with non-linear riskfromlong-term

fixed-incomeobligationswithembeddedoptionsfor theborrower to prepayandfrom

embedded optionsin non-maturity deposits

* In considering theassetsideof the hakncesheet, prepayment riskopuons

(Le.,mortgages,mortgage-backedsecurities,and consumerloans)are themainform ofembedded opuons Theprepaymentoptionresultsin uncertaincash flows and makes

interest rateriskmeasurement adifficult task

* Inconsideringtheliabilityside, thereare twoembedded optionsin non-maturity

deposits:(1) thehank hasanoptiontodetermine theinterest rate paid todepositors

and whentoamendthe rate,and(2) thedepositorhas the option towithdrawup to the

entirebalance widi no penalty Theinteractionbetween these LWOembeddedoptionscreatessignificantvaluationandinterest ratesensitivitymeasurement problems

* Sufficiendy modeling optionalityexposures requires verycomplexstochastic-path

evaluation techniques

Banks’pricingbehavior, \

* Thisfactorcontributestothechallengesin measuring theinterest rateriskofbanking

bookitems.Forexample, itwould requireamodel toanalyse the persistence of the many

differentnon-maturitybanking products,aswellas amodeltodetermine bank interest

rates diat considergeneral market conditions, customer relationships,bank commercialpower,and optimal commercial policies

* Determiningbankinterest rateswould requirethe pricingof credit risk The price

of credit riskapplied todifferent bankingproductscreates achallenge becauseit

would requireapricingrulethat links thecreditspreadtochangesin macroeconomic

conditions andinterest ratechanges.Also,it means thatin terest rate stress scenarios

should consider the dependence betweeninterest race andcredit riskfactors

The choiceofstress scenarios.

• The drawbacksof using simpleinterest rateshocks poseinterest rate measurement

challengesbecause the shocks:

+ Are not basedon probabilitiesand,therefore,aredifficultto integrateinto economic

capitalmodels basedonVaR

+ Arenot necessarilysensitive to the current rate or economic environment.

Do nottakeinto accountchangesin theslopeor curvatureof theyieldcurve

Do notallowforanintegratedanalysisofinterest rateandcredit riskson bankingbookitems.

Cross Reference to CARP Assigned Reading—Basel Committee on Banking LSupervision

BIS RECOMMENDATIONSFOR SUPERVISORS

AIM35-2: Describe theBIS recommendationsthat supervisorsshould considerto

make effectiveuseof riskmeasures notdesignedforregulatory purposes.

Thereare ten Bankfor International Setdements(BIS) recommendationstoconsider:

1 Useofeconomiccapitalmodels inassessingcapital adequacy.The bank should show

how such modelsare usedin checorporatedecision-makingprocessso as to assess

the model’simpactonwhich risks the hank chooses to accept.In addition, theboard

shouldhaveabasicunderstandingof die difference betweengross(stand alone) and net

(diversified) enterprise-wide riskinassessing the banksnet risktolerance

2 Senior management The economiccapitalprocessesabsolutelyrequireasignificant

commitmentfromsenior management.Theyshould understandits importancein die

corporate planningprocessand shouldensurethat thereis a stronginfrastructurein

placeto support the processes

3 Transparencyand integrationintodecision-making Economiccapital results need

tobeeasyto traceand understandinorder tohe useful.Careful attention mustbe

giventoobtainingreliableestimates on anabsolute basisinadditiontodevelopingthe

flexibilityto conductfirm-widestress testing

4 Risk identification Thisis thecrucial startingpoint in risk measurement.The risk

measurement processmustbe verythorough to ensure that the proper risk drivers,

positions, and exposuresaretakeninto account in measuringeconomiccapital.That

willensure that thereislittlevariancebetween inherent(actual)and measured risk.For

example,risks thataredifficult toquantifyshould be considered through sensitivity

analysis,stress testing, or scenarioanalysis

5 Riskmeasures. Nogiven risk measureis perfect, anda bankmust understand the

strengthsand weaknessesofitschosen riskmeasures.Noonerisk measure foreconomic

capita]is universally preferred

6. Riskaggregation.The reliability of the aggregation processisdeterminedby the

qualityof themeasurement riskcomponents, plustheinterrelationshipsbetween such

risks The aggregation process usuallyrequires consistency inthe riskmeasurement

parameters.The aggregation methodologiesused shouldmirror the batik’s business

compositionand risk profile

7 Validation Thevalidation processforeconomiccapital modelsmustbe thorough

andcorroboratingevidencefrom various tests must showdiat themodel‘‘works” as

intended In ocherwords,withinanagreeduponconfidence interval andtime period,

thecapitallevel determined mustheenough toabsorb the(unexpected) losses

Topic 35

Cross Reference to CARP AssignedReading-Basel Committee on Banking Supervision

S. Dependency modelingincreditrisk.Banks mustconsider the appropriateness of the

dependencystructuresusedwithin theircreditportfolio Specifically, credit models

need tobe assessedfor their limitations, andsuch limitations need tohe dealt widivia

appropriatesupplementary riskmanagement approaches,suchassensitivityor scenario

analysis

5 Counterparty credit risk There aretrade-offs tobeconsideredindecidingbetween

the available methods of measuringcounterpartycredit risk.Additionalmethods,such

as stress testing need tobe used tohelpcoverall exposures Measuring such riskis

complicated andchallenging.Specifically, the aggregationprocessneeds tobe vetted

prior to a hankhavinga bigpicture perspectiveofcounterpartycredit risk.

10. Interestrate risk in the bankingbook.Specifically, financial instruments with

embeddedoptionsneed tobe examined closelyinorderto control risk levels Certainly,

there are trade-offsbetweenusingearnings-basedversuseconomicvalue-based models

tomeasuringinterest race risk Forexample, theformer has aggregation problems

becauseodier risks aremeasured usingeconomicvalue.Also, usingeconomic

value-basedmodels couldbeinconsistentwith businesspractices

ECONOMIC CAPITAL CONSTRAINTSAND OPPORTUNITIES

AIM 35.3: Describe theconstraintsimposed andtheopportunitiesofferedby

economiccapitalwithin thefollowingareas:

* Credit portfolio management

* Risk based pricing+ Customer profitability’analysis

* Pricing decisionsarebasedonexpectedrisk-adjusted return oncapital(RAB.OC), so

dealswillberejected ifthey are lowerdianaspecificRAROC.The proposedinterest rate

isdeterminedbydieamountofeconomiccapitalallocated to thedeal

©2013Kaplan,Inc.

Page 36

Cross Reference to CARP AssignedReading—Basel Committee on Banking Supervision

• Pricing decisionsinclude: (1) costoffunding, (2) ejected loss,(3)allocated economic

capital,and(4)additional return required by shareholders.Therefore,aminimum

interest race isdeterminedthatwillincreaseshareholdervalue,

Opportunitiesoffered:

• Canbe used to maximize the bank’sprofitability.Forexample,some pricing decisions

may need tobeoverridden becausecertain customerrelationshipsare moreprofitable

(at alower prioe/interestrare) ordesirable froma reputationalpoint ofview.Ofcourse,

suchoverridesare not taken lightly andrequire uppermanagementapproval,aswell as

rigorous subsequent monitoring

Customer Profitability Analysis

Constraintsimposed:

* Theanalysisiscomplicatedin that many risks need tobeaggregatedat thecustomer

level

• Customers need to besegmentedin termsof ranges of(net) return per unit of risk; the

underlyinginformationisdifficult tomeasureand allocate

Opportunitiesoffered:

* Assuming that themeasurementobstacles have been overcome, dieanalysiscan beeasily

used todetermineunprofitableoronly slightly profitablecustomers.Suchcustomers

could he dropped andeconomiccapitalallocatedto diemoreprofitablecustomers.

* Economiccapitalis usedin maximizingthe risk-retnrn trade-off(throughrelative

risk-adjustedprofitabilityanalysisofcustomers).

ManagementIncentives

Constraintsimposed:

* Studiesshowthatcompensationschemesare a minorconsiderationin termsof the actual

usesofeconomiccapital measuresat the businessunitlevel

Opportunitiesoffered:

• It issuggested thatmanagement incentives is theissuethatmotivatesbank managersto

participatein the technical aspectsof the economiccapitalallocation process

Topic 35

Cross Reference to CARP AssignedReading-Basel Committee on Banking Supervision

KEY CONCEPTS

AIM35.1

Amultitudeofchallengesexist wiiJiin the economiccapitalframework drat involve:

(1)defining risk measures, (2) risk aggregation,[3}validation ofmodels, (4) dependency

modelingincredit risk,(5)evaluatingcounterpartycreditrisk,and(6)assessinginterest rate

riskin thebankingbook

AIM35.2

Thereare ten BISrecommendations that supervisors should consider tomake effectiveuse

of risk measures.

AIM35.3

A number ofspecificconstraintsimposedand opportunities offered byeconomiccapital

existwithin theareasofcredit portfoliomanagement, riskbased pricing,customer

profitability analysis,and management incentives.

©2013Kaplan,Inc.

Page 38

Cross Reference to GARP AssignedReading—Basel Committee on Banking Supervision

CONCEPT CHECKERS

Whichof thefollowing risk measures is dieleastcommonly usedmeasurein the

practiceof risk management?

A. Valueat risk

5 Standarddeviation

C. Expected shortfall*

D Spectral riskmeasures.

Whichof thefollowingaggregation methodologiesischaracterized bygreatdifficulty

in validatingparameterizationand buildinga jointdistribution?

A. Copulas

B Constan tdiversification

C Variance-covariance matrix.

D Fullmodeling/simulation

Whichof thefollowingmodel validation processesisspecificallycharacterizedby the

limitation thatitprovides little comfort that the model actually reflects reality?

A. Backtesting.

B Benchmarking

C Stresstesting

D Qualitativereview.

Whichof the followingcategoriesofBIS recommendations specifically referstodie

needtoconsider usingadditionalmethods,suchas stresstesting, to helpcoverall

exposures?

A Riskaggregation

B Counterpartycredit risk

C Dependency modelingincreditrisk

D Interestrateriskindie banking book

The use of which of thefollowingitems is meant morefor protecting against risk

deterioration?

A. Risk based pricing

B Managementincentives.

C Credit portfoliomanagement.

D Customer profitability analysis