剑桥风险研究中心-高通胀世界金融灾难压力测试（英）_市场营销策划_重点报告202301202_doc.docx

CentreforRiskStudiesp三UniversityofPCAMBRIDGEJudgeBusinessSchlCambridgeCentreforRiskStudiesUniversityofCambridgeJudgeBusinessSchoolTrumpingtonStreetCambridge,CB21AGUnitedKingdomries.riskj6s.cam.ac.ukhttp，WWW.risk.jbs.cam.ac.ukDecember2015TheCambridgeCentreforRiskStudiesacknowledgesthegeneroussupportprovidedforthisresearchbythefollowingorganisations:TheviewscontainedinthisreportareentirelythoseoftheresearchteamoftheCambridgeCentreforRiskStudies,anddonotimplyanyendorsementoftheseviewsbytheorganisationssupportingtheresearch.Thisreportdescribesahypotheticalscenariodevelopedasastresstestforriskmanagementpurposes.Itdoesnotconstituteaprediction.TheCambridgeCentreforRiskStudiesdevelopshypotheticalscenariosforuseinimprovingbusinessresiliencetoshocks.Thesearecontingencyscenariosusedfor*what-ifstudiesanddonotconstituteforecastsofwhatislikelytohappen.FoodandOilPriceSpiralStressTestScenarioHighInflationWorldContents1ExecutiveSummary42 DefiningtheScenario83 HighInflationasaFinancialCatastrophe124 DefiningtheScenario155 TheScenario176 MacroeconomicAnalysis197 ImpactonInvestmentPortfolio258 MitigationandConclusions329 Bibliography33FoodandOilPriceSpiralStressTestScenarioHighInflationWorld1ExecutiveSummaryInthefollowingreport,wepresentanarrativeofhowglobalinflationarypressureoverseveralyearsimpactstheworldeconomyandfinancialmarkets.Thisprovidesabasisforaglobalenterprisetotestitsoperationalandstrategicmodel,asasteptowardimprovingitsresilience.Scenariosmoregenerallycanbeusedtocoverthespectrumofextremeshocks,suchasthoseproposedintheCambridgeTaxonomyofThreats,whichencompassesfiveclassesofbusinessrisk.1HighInflationWorldScenarioThisscenarioemdsionscostshocksinresponsetoshrinkingglobaloilsuppliesand,simultaneously,disruptionstocropproductionthatleadtoglobalfoodshortages.Theseinflationarydriverspersistovermanymonths,causinginternationaleconomicandhumanitarianpressures.Theeconomicimpact,expressedaslostglobalGrossDomesticProductoverfiveyears,comparedwiththeprojectrateofgrowth("GDPRiSk")，isbetween$4.9,$8and$10.9trillion,dependingontheseverityofthecommoditypriceshock.TheGreatRecessionof2007-2011,comparatively,sawalossof$20trillionin2015dollarestimates.Inthisperspective,althoughtheHighInflationWorldScenarioinflictssevereeconomicloss,thecatastrophedoesnotpreventtherecoveryoftheglobaleconomyovertime.HighInflationasaFinancialCrisisScenarioselectionInflationistiedtotherelationshipbetweenaggregatesupplyanddemand.Cost-pushdescribesasupplyshortage,e.g.,duetoadisruptioninproductionofacommodity.Demand-pulldescribesincreasingdemand,perhapsresultingfromalooseningofcredit.Inbothcases,inflationofcommoditypricesoccurs.TheHighInflationWorldScenarioisacost-pushsituationdrivenbyrelativescarcityofbothoilandagriculturalcommodities.Thefinalimpactofthesepricehikesdependsheavilyonthelevelofexposureacountryhastoeachcommodity.,CambridgeCentreforRiskStudies,uATaxonomyofThreatsforComplexRiskManagement*,2014Nonetheless,thedirectimpactofaglobalhighinflationisthecorrespondingincreaseinUnemPlOymentrates,albeitvaryingseverity,acrossmajoreconomies.VariantsofthescenarioWecalibratethreevariantsofthescenariousingdifferentlevelsofinflationforfoodandenergyprices.InourstandardscenarioSi,commoditypricesjumpbetween180and210%ofthepre-existingpricelevels,withpricespeakingaround15monthsaftertheinitialshock.ScenariovariantS2andextremevariantXiaresimilartothestandardscenario,butthecommoditypriceincreasesareraisedupto280and440%,respectively.ThescaleoflossinflictedbytheHighInflationWorldScenariohasbeencalibratedtocorrespondapproximatelytoaneventthathappensaboutonceacenturyonaverage,a1-in-100yearevent.Twoindicatorsthatmaygiveasenseofthelikelihoodofacatastrophescenariooccurringareitsimpactonequityreturnsandgrowthrates,whichareexpectedtobenegativeasaresultofcatastrophe.InthecaseoftheHighInflationWorldScenario,however,ouranalysisdoesnotshowextremebehaviourineitherofthesecategories.US(UK)equitiesoverthelasttwohundredyears Prior to records from FTSE and S&P, We use surrogate stocks such as those from American railroad stock prices and other constructed indexes. We use similar surrogate data for estimating growth rates prior to the availability of standardised data. Our identification of%iles uses a normal curve fitting which is conservative in light of the fat tails associated with equity price distributions.haveexperiencedreturnratesbelow-24%(-13%)aboutonceintwentyyears,withreturnratesbelow-36%(-20%)signifying1-in-100events.Inourscenariovariants,thosereturnratesarebarelyeffectedotherthanintheextremeXivariantinwhichequityreturnratesare-8%intheUSand-4%intheUK.Nearzeroeconomicgrowthratesarefoundinourscenariosbutthesedon,tcomparetothehistoricalrecordforUS(UK)growthratesbeingbelow-7%(-3%),whichare1-in-20yearevents,orratesbelow-13%(-5%)whichhappenseverycentury.Thisisastresstest,notapredictionThisreportisoneofaseriesofstresstestscenariosthathavebeendevelopedbytheCentreforRiskStudiestoexploremanagementprocessesfordealingwithanextremeshock.Itdoesnotpredictacatastrophe.SeedsofshortageFarmingfailureTheScenarioassumesthatthemiddleoftheyearbringsWithitboutsofextremeweatheracrossthenorthernhemisphere:alongheatwaveinthePacificWest,floodsintheSub-Indiancontinent,heavyrainsintheAtlanticanddroughtinnorthernChina.Grainyieldsaresuretosuffer.Concurrently,apandemicsweepsthroughtheworld,spopulationofbees.Inadequatepollinationpreventstheworldwidedevelopmentofnuts,fruitandotheragriculturalproducts.HoldingtheStraitofHormuzHostageAmilitantgroupestablishesholdontheStraitofHormuzinthePersianGulf,effectivelyseizingcontrolof20%oftheworld,scrudeexports.ThegrouprestrictstheinternationalshipmentofcrudeoilthroughtheStraits,hikingthepriceofoiltoover$170perbarrel.Theimpactaffectstheinternationalmeatanddairyindustriessignificantly.Thecombinationofhighproductioncostsandefficiencylossesaffectaggregatedemandasacost-pushspiralemergesworldwide.GlobalstagflationAstheinternationalenergycrisiscontinuestheconsumerpriceindexspikesinmanynations,drivingdemandsfornationalwageincreases.Stagflationemergesacrosstheglobeascountriesthatimplementwagehikesexperienceanunemploymentspiral.Inanefforttocurtailworld¼ddestagflationattheheightofthecrisis,nationalcentralbanksgraduallyadjustinterestratesinordertosuppressconsumerspendingandrelieveeconomicpressure.Aftereighteenmonths,pricesbegintostabiliseandtherateofinflationdrops.GlobalGDPimpactTounderstandhowtheHighInflationWorldscenarioimpactstheglobaleconomyweusetheGlobalEconomicModel(GEM),OxfordEconomics,quarterly-linkedinternationaleconometricmodel.Priceshocksareapplieddirectlytoworldfoodandenergypricesovera15monthperiod,andthemodeladjustsendogenouslytoallocateinflationrateincreasesacrosstheworld.WeusetheGEMtoestimatethelossinglobalgrossdomesticproduct,cumulatedovera5yearperiod,whichisattributedtothisstresstestscenario.Wetermthisloss'GDPRiSk'.GPDRisk,expressedinrealtermsinUSdollars,rangesfromalossof$1)84.9trillionforSito$USio,9trillionintheXivariant.However,thisscenariodoesnotnecessarilyleadtoaglobalrecession,butinsteadslowsdowntheeconomicgrowthconsiderably.TheseimpactsaresignificantbutnotofthesamescaleastheGreatFinancialCrisis,from2008-2012,whoseGDPRiSkisaround$20trillionin2015dollars.FinancialmarketimpactWeestimatetheportfolioimpactsofthisscenariobymodellingtheoutputsfromOEMintoportfolioreturns,projectingmarketchangesandcashflowswhilekeeptheallocationpercentagesfixed.Wealsodefaultallcorporatebondsgiventhe2008defaultrates.Giventhattheconsumerpriceindex(CPI)wasdirectlyshockedinthemacroeconomicmodelling,weseethatthetotalportfolioreturnsinreal%aremoresignificantlyimpactedthaninnominaldollars.ThemaximumdownturnexperiencedfortheConservativeportfoliointheSivariantis-3.89%innominalor-9.69%inrealtermsandoccursinYr2Q4.Bestandworstperformances(withinequities)aretheUK(FTSE100),andJapan(N225);withinfixedincome,USandJapanesebonds,Withtheworstperformingportfoliostructurebeinghighfixedincome,at7.93%fortheSivariant.ForportfolioprotectionitisrecommendedthatequityallocationisshiftedawayfromJapantowardsUKandawayfromJapanfixedincometowardsUSfixedincome.RiskmanagementstrategiesScenariosasstresstestsThisscenarioisanillustrationoftherisksposedbysocialunresttriggeredbycatastrophicevent.TheHighInflationWorldscenarioisjustoneexampleofawiderangeofscenariosthatcouldoccur.Thisscenarioaimstoimproveorganizations)operationalriskmanagementplansaroundcontingencies,andstrategiesforsurvivingfinancialandcounterpartychallenges.Itpresentsacapitalstresstestforinsurerstoassesstheirabilitytomanageunderwritinglosseswhilealsosufferingmarketimpactsontheirinvestmentportfolios.SummaryofEffectsofHighInflationWorldScenarioandVariantsScenarioVariantS1S2X1VariantDescriptionStandardScenarioScenarioVariantExtremeVariantWorldenergypriceshock210%280%440%Worldfoodpriceshock180%250%310%Pricespiralduration5Qtrs5Qtrs5QtrsMaCrOeCOnOmiCIOSSeSGlobalrecessionseverity(MinimumqtrlygrowthrateglobalGDP)1.9%1.4%0.6%GlobalrecessiondurationNorecessionGDPRiSk$Tr(5yearlossofglobaloutput)$4.9Trillion$8.0Trillion$10.9TrillionGDP(三)Risk%(as%of5-yearbaselineGDP)1.7%2.2%2.6%PortfoliOImPaCtPerformanceatperiodofmaxdownturnHighFixedIncome-8%-10%-16%Conservative-4%-7%-14%Balanced-3%-6%-13%Aggressive-1%-4%-12%ASSetClaSSPerfOrmanCeYr1Qr4Yr3Qr4Yr1Qr4Yr3Qr4Yr1Qr4Yr3Qr4USEquities(W5000),%Change-20%4%-39%-36%-1%-1%UKEquities(FTSE100),%Change-72%-43%-73%-49%-3%18%USTreasuries2yrNotes,%Change0%3%0%5%-7%-16%USTreasuries10yrNotes,%Change2%15%2%17%-13%-22%Table1:SummaryimpactsoftheHighInflationWorldscenarioTrillionUS$GDPRiskacrossscenarios|S1S2xI藤MillennialUprisingSocialUnrestRisk1.64.68.1谡DollarDeposedDe-AmericanizationoftheFinancialSystemRisk1.91.6-1.6SybilLogicBombCyberCatastropheRisk4.57.415HighInflationWorldFoodandOilPriceSpiralRisk4.9810.9SaoPaoloInfluenzaVirusPandemicRisk71023EurozoneMeltdownSovereignDefaultRisk11.216.323.2GlobalPropertyCrashAssetBubbleCollapseRisk13.219.6China-JapanConflictGeopoliticalWarRisk1727322007-12GreatFinancialCrisis18GreatFinancialCrisisat201420Table2:GDP(g)RiskimpactoftheHighInflationWorldscenariocomparedwithpreviousCentreforRiskStudiesstresstestscenarios2FinancialCatastropheStressTestScenariosThisscenarioisanillustrationoftherisksposedbyaplausiblebutextremefinancialmarketbasedcatastrophe.Itrepresentsjustoneexampleofsuchacatastropheandisnotaprediction.Itisa"what?if,exercise,designedtoprovideastresstestforriskmanagementpurposesbyinstitutionsandinvestorswishingtoassesshowtheirsystemswouldfareunderextremecircumstances.ThisscenarioisoneofaseriesofstresstestscenariosdevelopedbytheCentreforRiskStudiestoexplorethemanagementprocessesfordealingwithanextremeshockevent.Itisoneoffourfinancialmarketcatastrophescenariosbeingmodelledunderthisworkpackageandincludesthefollowing: GlobalPropertyCrash:AssetBubbleCollapse; DollarDeposed:De-AmericanisationoftheGlobalFinancialSystem; EurozoneMeltdown:SovereignDefaultCrisis.Thescenariospresentaframeworkforunderstandinghowglobaleconomicandfinancialcollapsewillimpactregions,sectorsandbusinessesthroughoutthenebvorkedstructureoftheeconomy.Thesefinancialstresstestsaimtoimproveorganisations,operationalriskmanagementplanstoformcontingenciesandstrategiesforsurvivingandminimisingtheimpactsfrommarket-basedfinancialcatastrophe.Inparticular,thestresstestsallowinstitutionstomanageandbuildresiliencetodifferentformsofriskduringperiodsoffinancialstress.Theserisksinclude: financialandinvestmentriskstemmingfromacollapseinassetpricesacrossdifferentsectorsandregions; supplychainriskandtheabilityofaninstitutiontoeffectivelymanageitsinputrequirementsthroughitssupplychain,tomeetinternalproductionandoperationalrequirements; customerdemandriskandknowledgeforhowdemandmightshiftforgoodsandsendeesduringperiodsoflowinvestmentandconsumerspending; marketorsegmentationriskandanunderstandingofhowotherfirmswithinthesamesectorwillreactandperformduringperiodsoffinancialstressandhowthismayimpactonthebusiness; reputationalriskandtheprotectionofbrandimageforreactingappropriatelyandconfidentlyundercrisisconditions.Eachindividualscenariomayrevealsomeaspectsofpotentialvulnerabilityforanorganisation,buttheyareintendedtobeexploredasasuiteinordertoidentifywaysofimprovingoverallresiliencetounexpectedshocksthatarecomplexandhavemultifacetedimpacts.MarketcatastropheriskandfinancialcontagionTheGreatFinancialCrisisof2007-8notonlyrevealedtheextenttowhichtheglobalfinancialsystemisinterconnectedbuthowinterrelationshipsbetweencommercialbanks,investmentbanks,centralbanks,corporations,governments,andhouseholdscanultimatelyleadtosystemicinstability.Asglobalfinancialsystemsbecomeincreasinglyinterconnected,ashocktoonepartofthesystemhasthepotentialtosendacascadeofdefaultsthroughouttheentirenetwork.In2008,itwasonlythroughgovernmentinterventionintheformofextensivebailoutpackagesthatawidespreadcollapseoftheglobalfinancialsystemwasavoided.Newmodelsoftheglobalfinancialsystemareanessentialtoolforidentifyingandassessingpotentialrisksandvulnerabilitiesthatmayleadtoasystemicfinancialcrisis.Theliteratureidentifiesthreetypesofsystemicrisk:(i)build-upofwide-spreadimbalances,(ii)exogenousaggregateshocksand(iii)contagion(Sarlin,2013).Similarlyweworkwiththreeanalyticalmethodsthathelpdealwithdecisionsupport:(i)early-warningsystems,(ii)macrostress-testing,and(iii)contagionmodels.AllthreemethodsareactivelyunderresearchintheCentreforRiskStudiesandutilisedinthedevelopmentofthesestresstestscenarios.UnderstandingfinancialcatastrophethreatsThisscenarioexplorestheconsequencesofafinancialmarketcatastrophebyexaminingthenotional1-in-100possibilityforaHighInflationWorldScenarioandexamininghowtheshockwouldworkthroughthesystem.Foraprocessthattrulyassessesresiliencetomarketcatastrophe,vveneedtoconsiderhowdifferentmarket-basedcatastrophesoccurandthenpropagatetheseshocksthroughglobalfinancialandeconomicsystems.Thisexercisewouldideallyincludeathoroughanalysisforeachdifferenttypeofmarketcatastropheinadditiontothefourfinancialcatastrophesincludedinthissuiteofstresstests.Suchananalysiswouldalsoincludearangeofdifferentseveritiesandcharacteristicsforthesescenarioswouldoccurasaresultofthesedifferentfinancialandeconomiccrises.TheCambridgeRiskFrameworkattemptstocategorizeallpotentialcausesoffutureshocksintoaUniversalThreatTaxonomy.wWehavereviewedmorethanathousandyearsofhistoryinordertoidentifythedifferentcausesofdisruptiveevents,collatingotherdisastercataloguesandcategorizationstructures,andresearchingscientificconjectureandCounterfactualhypotheses,combinedwithafinalreviewprocess.Theresul