IMF-马尔代夫：金融部门评估方案关于银行压力测试和气候风险分析的技术说明（英）-2024.1.docx

INTERNATIONA1.MONETARYFUNDIMFCountryReportNo.24/19MA1.DIVESFINANCIA1.SECTORASSESSMENTPROGRAMJanuary2024TECHNICA1.NOTEONBANKSTRESSTESTINGANDC1.IMATERISKANA1.YSISThispaperontheMaldiveswaspreparedbyastaffteamoftheInternationalMonetaryFundasbackgrounddocumentationfortheperiodicconsultationwiththemembercountry.ItisbasedontheinformationavailableatthetimeitwascompletedonDecember18,2023.CopiesofthisreportareavailabletothepublicfromInternationalMonetaryFundPublicationServicesPOBox92780Washington,D.C.20090Telephone:(202)623-7430Fax:(202)623-7201E-mail:Dublicationsimf.orgWeb:InternationalMonetaryFundWashington,D.C.©2024InternationalMonetaryFundDecember18,2023INTERNATIONA1.MONETARYFUNDMA1.DIVESFINANCIA1.SECTORASSESSMENTPROGRAMTECHNICA1.NOTEBANKSTRESSTESTINGANDC1.IMATERISKANA1.YSISPreparedByMonetaryandCapitalMarketsDepartmentThisTechnicalNotewaspreparedbyIMFstaffinthecontextoftheFinancialSectorAssessmentPrograminMaldives.ItcontainstechnicalanalysisanddetailedinformationunderpinningtheFSAP,sfindingsandrecommendations.FurtherinformationontheFSAPcanbefoundatCONTENTSGlossary4EXECUTIVESUMMARY6INTRODUCnON9A. MacrofinancialDevelopments9B. FinancialSystemStructure9C. BankingSystemCharacteristics10SYSTEMICRISKANA1.YSIS14A. ScopeandDataQuality14B. MacrofinancialStressTestScenarios15SO1.VENCYSTRESSTESTS17A. Top-DownStressTestMethodology17B. Top-DownStressTestsResults20C. Bottom-UpStressTestResults27Uquiditystresstests31A. Cashflow-based1.iquidityStressTests31B. 1.iquidityCoverageRatio33C. DepositConcentrationSensitivityAnalysis34D. HERSENSITIVITYANA1.YSES35A. InterestRateRisk35B. ForeignCurrencyRiskinBalanceSheets35C1.IMATERISKANA1.YSIS36A. PhysicalClimateRiskContextoftheMaldives36B. ClimateScenarios37C. Methodology38D. Results48E. Recommendations49References67FIGURES1. StructureoftheBankingSystem102. AssetAllocation113. AssetQuality124. 1.iquidityandFunding135. Capitalization146. ProjectedPathsofMacroeconomicVariablesinStressTestScenarios167. SatelliteModelProjections:AggregateNP1.Ratio218. AggregateCapitalizationinBaselineScenario229. AggregateCapitalizationinModerateScenario2310. AggregateCapitalizationinSevereScenario2411. StandaloneSovereignSensitivityAnalysisResults2712. SummaryBottom-UpStressTestResults:BaselineScenario2813. SummaryBottom-UpStressTestResults:ModerateScenario2914. SummaryBottom-UpStressTestResults:SevereScenario3015. ShareofBanksthatFailed1.argestFiveDepositorsOutflowsTest3416. MarketRiskSensitivityAnalysisResults3617. Sea1.evelRiseUnderDifferentClimateScenarios3818. PhysicalRiskAnalysisFramework3919. IslandsSubsetandClimateDataMatching4020. CoastalFloodEstimates4421. GeographicalExposure4522. DamageRatesEstimations4623. ResultsofClimateRiskAnalysis49TAB1.ES1. RecommendationsonStressTestingandClimateRiskAnalysis82. CreditRiskSatelliteModelEstimates183. Pre-ProvisionIncomeandRisk-WeightedAssetsGrowthPath194. SovereignSensitivityAnalysisAssumptions205. SummarySolvencyRiskResults256. Cashflow-basedStressTestAssumptions327. Cashflow-basedStressTestResults338. Summary1.iquidityStressTestResults349. SummaryMarketRiskSensitivityAnalysisResults3510. TVaR99ImpacttoCapitalStock47APPENDICESI.SelectedEconomicIndicators,2019-20285111.FinancialSoundnessIndicators,2019-202252I11.RiskAssessmentMatrix53IV. StressTestingMatrix55V. FSAPMacroVariablesfortheBaseline,ModerateandSevereScenarios59VI. Bottom-UpStressTest:InstructionsandAssumptions60VII. ClimateDataTreatmentProcess61VIII. Monte-CarloSimulationProcess62IX. 1.ossDistributions63X. ImpactofCoastalFloodsontheCapitalStock64XI. ImpactofPreviousEvents65XII. AdministrativeAtollsCodesandNames66GlossaryAFSAM1./CFTAR6BCPCARCDSDDEFSAPFSPNFCFXGCMGDPGEVGFCHDCHFTHTMIPCC1.C1.CR1.1.P1.1.R1.TVMBSM1.SAMMAMoEMoFMoTMRPSMSMENASANIMNOPNSFRNP1.NPVPPIRAMRCPAvailableforSaleAnti-Money1.aundering/CombatingtheFinancingofTerrorismSixthAssessmentReportBaselCorePrinciplesforEffectiveBankingSupervisionCapitalAdequacyRatioCreditDefaultSwapDomesticDebtExchangeFinancialSectorAssessmentProgramFinancialSectorPolicyNoteForeignCurrencyForeignExchangeGeneralCirculationModelGrossDomesticProductGeneralizedExtremeValueDistributionGlobalFinancialCrisisHousingDevelopmentCorporationHeldforTradeHeldtoMaturityIntergovernmentalPanelonClimateChange1.ocalCurrency1.iquidityCoverageRatio1.oan1.ossProvisioning(flow)1.oan1.ossReserves(stock)1.oan-to-ValueMaldivesBureauofStatisticsMaldives1.andandSurveyAuthorityMaldivesMonetaryAuthorityMinistryofEnvironment,ClimateChangeandTechnologyMinistryofFinanceMinistryofTourismMaldivesRetirementPensionSchemeMicro,SmallandMedium-SizedEnterprisesNationalAeronauticsandSpaceAdministrationNetInterestMarginNetOpenPositionNetStableFundingRatioNonperforming1.oanNetPresentValuePre-ProvisionIncomeRiskAssessmentMatrixRepresentativeConcentrationPathwayrhsRighthandsideROAReturnonAssetsROEReturnonEquityRWRiskWeightRWARisk-WeightedAssetsSOEState-OwnedEnterpriseSSPSharedSocioeconomicPathwaySTeMStressTestingMatrixTaVRTailValueatRiskUNUnitedNationsEXECUTIVESUMMARYThisTechnicalNotewaspreparedbyIvanGuerraandJavierUrunuela1.6pe乙withcontributionsfromYizhiXuandKiranSastry.AsystemicvulnerabilityanalysisandstresstestswereconductedaspartoftheMaldivesFSAP.ThevulnerabilityanalysisandstresstestswerebasedonquarterlyaggregatebalancesheetsupervisorydatafortheeightbanksinMaldivesasofDecember2022.IdentifiedvulnerabilitiesweresubjectedtohypotheticalextremebutplausiblescenariosthatwereinformedbytheRiskAssessmentMatrix.Risksanalyzedwerecreditrisk,liquidityriskandmarketrisk.Creditrisksmaterializedasnon-performingloansandpressureonpre-provisionincome,liquidityrisksasdepositoutflows,andmarketrisksaschangesininterestandexchangerates.AlthoughtheMaldives'economyhasreboundedstronglyfromthepandemic-inducedcontraction,macroandfinancialvulnerabilitiesremain.Fiscalandexternalvulnerabilitieshavebeenelevated,arisingfromhighpublicdebt,increasingfiscalexpenditureondebtserviceandpricesubsidies,andawideningcurrentaccountdeficit.Inaddition,continuedfinancialsupporttostate-ownedenterprises(SOEs)andapersistentFXshortageintheofficialmarketshavecontributedtoincreaseddomesticfiscalfinancingneedsandfurtherrationingonFXsupplytotheprivatesector.Relatedtothesemacrodevelopments,systemicfinancialvulnerabilitieshavebecomemoreprominent,whichincludeanintensifiedsovereign-banknexus,highdollarizationinteractingwithFXshortages,shadowbankingactivities,andweakliquiditymanagement.Againstthebackdropofthesemacro-financialdevelopments,theFSAPidentifiedanumberofsystemicvulnerabilities.ThemainmacrofinancialvulnerabilitystemsfromhighcentralgovernmentandSOEdebtthatisincreasinglyfinancedbybanksthroughincreasingholdingsofsovereignsecuritiesandsharplyrisinglendingtoSOEs.Prudentialandregulatorypolicieshavefurtherincentivizedtheaccumulationofsovereigndebtonbankbalancesheets,notablythroughthezero-riskweight(RW)ondomesticsovereignpaper,includingFX-denominatedissues.Moreover,thecurrenttrajectoryofpublicdebtservice,includinginforeigncurrency,combinedwithapossibledropinFXinflowspresentsachallengeformanagingofficialreservesandcouldpromptanexchangerangerealignment,affectingSOEsandcorporateswithcurrencymismatches.Financingofconsumerdurablesbyleasingcompanies,someofwhichareunregulated,usingwidespreadleaseandhirepurchaseprogramsleaverecurringhouseholdpaymentobligationsunderreported.Banksarealsoexposedtolargecorporateclients,asevidencedbyindividualbanksbeingclosetotheirsingleexposurelimits.1.astly,managementofsystemicliquidityneedsimprovement,reserverequirementswouldneedfinetuning,anddraftregulationaddressingissuescontributingtotheparallelFXmarketshouldbeadopted.ThestresstestsappliedtheusualFSAPrangefornon-comple×bankingsectors.Thequalityofsupervisorydataforstresstestingseemedadequateoverallalthoughmixedintermsofcoverageandgranularity.Thesolvencystresstestassumedthreemacrofinancialscenarios,increasinginseverity.Thesescenarioswerealsosharedwiththebanksalongwithinstructionstoconductabottom-upstresstest,inwhichbanksweregiventheprojectionsforallmacrovariablesandwereaskedtoapplytheirownstresstestingmethods. Forthetop-downsolvencystresstests,along-runrelationshipbetweennon-performingloansandmacroeconomicvariablesineachscenariowasestimatedusingquarterlybank-by-bankpaneldatafortheperiod2010-2022.Thus,thecreditriskmodelsprojectedNP1.ratiosforbanks/loanportfoliosinlocalandforeigncurrencyforeachstressscenario,andadditionalprovisioningneedswerecalculated.Apartfromprovisions,thepre-provisionincome,taxes,anddividendsofbankswerealsoprojectedtoarriveatafter-taxincomeandthereforechangesinbankcapital.Theprojectedcapitaladequacyratioswerethenobtainedapplyingaprojectionofrisk-weightedassets(RWA). Theseverescenar100fthesolvencystresstestwasaugmentedbyincorporatingasovereignrisksensitivityanalysis.Inaddition,acreditconcentrationsensitivityanalysiswasruntoaccountfortheexposuretolargecorporateclients.Furthermore,inanalternativecalculationofcapitaladequacy,thetestsalsoassumednon-zeroriskweightsonsovereignsecuritiesinforeigncurrency,therebyincreasingRWASubstantiallyandlowercapitaladequacyratios. Theliquiditystresstestsappliedthecashflows-basedmethodologyusinglongtermestimatesofoutflowsbydeposittype.Inaddition,theBaselIII1.iquidityCoverageRatio(1.CR)wasalsocalculated.Bothtestswereperformedinlocalandforeigncurrency. Othersensitivityanalysisaccountedforinterestraterisk,foreigncurrencyrisk,anddepositconcentrationrisk.Nointerconnectednessstresstestswererun,astheinterbankmarketisvirtuallynon-existent.Thestresstestresultsbroadlycorroboratedtheidentifiedvulnerabilitiesandquantifiedthem.Whilethebankingsystemseemstoberesilienttomacroeconomicshocks,itislesssotosovereignshocksand/orconcentrationrisk.Banks'solvencywasmostlyimpactedintheseverescenario,whichalsoincludedasovereigndomesticdebtexchange,andbythecreditconcentrationshocksimulatingthedefaultofthefivelargestexposures,withrequiredrecapitalizationsamountingtolessthan1percentofGDPineithercase.Inaddition,aBaselriskweightsadjustmentof100percentondomesticsecuritiesinFXalsohadaconsiderableimpactoncapitalbutnotenoughforanyrequiredrecapitalization.Intheliquiditystresstests,acoupleofbanksfaceddifficultiesbutonlyforspecifictimebuckets.Moreover,calculationofthe1.CRindicatedthatthebankingsectorwouldbecompliantwithBaselIII.However,depositconcentrationwasalsofoundtobearisk,withthebankingsystemshowingvulnerabilitiestowithdrawalsfromeachbank,sfivelargestdepositors.Marketriskconsistingofinterestraterepricingandforeigncurrencyriskwasfoundtobemoderate.Theclimateriskanalysisconsideredamicroapproachthatshocksbanks'immovableasset-relatedloansunderthreeclimatescenarios.Coastalfloodshazardwasconsideredintermsofsealevelriseandstormsurgewithfutureclimate;thelattermodeledthroughwindspeed.Theexposureincludedageographicaldisaggregationoftheeconomicactivitywithproxyvariables,butnotdifferentiatedbybanks.Usingthedamagefunctionsandelevation,themissionteamtransformedthecoastalflooddepthandatollexposureintoadamageratebyatoll;thesedamagerateswereusedtocomputetheinteractionbetweenatollsandcalculatetheaggregatelossesatthecountrylevel.Whilemid-centuryclimateeffectsonthebankingsystemwerefoundtobemild,theassessmentofend-centuryimpactsandinsurabilityissueswouldrequiremoregranulardata.Consideringthe99thpercentileofthecountrylossdistributionforthemid-century,theeffectsofclimate-relatedeventsonthebankingsystemwerefoundtobemild.However,theycouldbesignificantlyexacerbatedfortheendofthecentury,mainlyduetosealevelrise.Withrisingreinsurancepremiums,thecountrycouldbechallengedinthefuturebylimitedornoreinsuranceforclimate-relatedevents.Improvingthegranularityandcoverageofthedata,aswellasinitiatingwithclimateriskanalysiswouldallowabetterunderstandingoftheclimateimplicationsintheeconomyandfinancialsector.Whiletheanalysisleveragedglobalandlocaldatasources,thereisaneedforbetterdatagranularityforthecountryandfinancialsystem,whichwouldimprovetheassessmentoftheclimaterisk.Table1.Maldives:RecommendationsonStressTestingandClimateRiskAnalysisRecommendationResponsibleAuthorityTime*StressTestingImproveintegrityandgranularityofsupervisorydata,includingdatacompiledbytheCreditInformationBureau(CIB)MMASTDevelopmethodologiesforsolvency,liquidity,andmarketriskstresstestsandengagebanksinadialogueaboutstresstestproceduresandresults,includingbanks'ownstresstests.MMASTImplementscenario-basedsolvencystresstestsMMAMTImplementcashflow-basedliquiditystresstestsMMAMTGranularityofdatashouldbeimprovedtoidentifystablevs.less-stabledepositstowardcalculationoftheNetStableFundingRatioMMASTClimateRiskAnalysisImprovegranularityandcoverageofclimatedata,geographicalexposuresofthecountryandfinancialsystemaswellasclimate-relateddamages,andfosterintra-agencycollaborationtosupportaccesstoexistingdata.MoEfMBS,MMASTInitiateclimateriskanalysisincollaborationwithotheragenciestoassesstheeffectofactualandfutureclimateconditionsonthefinancialsectorandtheeconomy.MMAMT*ST:shortterm=1-2years;MT:mediumterm=3-5yearsZIINTRODUCTIONA. IVIacrofinanciaIDevelopments1. TheMaldives'economyhasreboundedstronglyfromthepandemic-inducedcontraction,thankstoarobustresumptionintourism,yetvulnerabilitiespersist.Afterdoubledigitgrowthin2022,realgrossdomesticproduct(GDP)growthisprojectedat7.2percentin2023basedonIMF'sApril2023WorldEconomicOutlook.However,fiscalandexternalvulnerabilitiesremainelevated,arisingfromhighpublicdebt,increasingfiscalexpenditureondebtserviceandpricesubsidies,andawideningcurrentaccountdeficit.Continuedfinancialsupporttostate-ownedenterprises(SOEs)lparticularlyforinvestmentprojectsoftheHousingDevelopmentCorporation(HDC),hasaddedtofiscalvulnerabilities.Asaresult,risingfiscalfinancingneedsarebeingmetbydomesticdebtissuanceandmonetaryfinancing,increasingthesovereignriskexposureofboththeMaldivesMonetaryAuthority(MMA)andthedomesticbankingsystem.AworseningshortageontheofficialForeignExchange(FX)marketreflectsimport-intensiveinvestment,pandemic-relatedincreaseinpublicspending,andFXrationingbybothMMAanddomesticbanks.TheFXshortagehasfurtherfueledalargeandwell-establishedparallelmarketthatprovidesmostoftheFXneedsofimporters,implyingcoststotheprivatesectorandlowertaxrevenue.2. Systemicvulnerabilitiesremainoutstanding,whichincludeanintensifiedsovereignbanknexus,persistentFXshortages,growingshadowbanking,andweakliquiditymanagement.ThemainmacrofinancialvulnerabilitystemsfromhighcentralgovernmentandSOEdebtthatisincreasinglyfinancedbybanksthroughgrowingholdingsofsovereignsecuritiesandsharplyrisingSOElending.Meanwhile,banks,appetiteforsovereigndebthasincreased,incentivebycurrentprudentialandregulatorypolicies,notablythroughthezero-riskweight(RW)ondomesticsovereignpaper,includingFX-denominatedissues.Moreover,thecurrenttrajectoryo