7/30/2019 Geneva Association Systemic Risk in Insurance Report March2010
1/129
Systemic Risk in InsuranceAn analysis of insurance and nancial stability
Special Report of The Geneva Association Systemic Risk Working Group
March2010
7/30/2019 Geneva Association Systemic Risk in Insurance Report March2010
2/129
The Geneva Association
(The International Associationfor the Study of Insurance Economics)
The Geneva Association is the leading international insurance think tank for strategically important insuranceand risk management issues.
The Geneva Association identies fundamental trends and strategic issues where insurance plays a substantial
role or which inuence the insurance sector. Through the development of research programmes, regular
publications and the organisation of international meetings, The Geneva Association serves as a catalyst for
progress in the understanding of risk and insurance matters and acts as an information creator and disseminator.
It is the leading voice of the largest insurance groups worldwide in the dialogue with international institutions.
In parallel, it advancesin economic and cultural termsthe development and application of risk management
and the understanding of uncertainty in the modern economy.
The Geneva Association membership comprises a statutory maximum of 80 Chief Executive Ofcers (CEOs)
from the worlds top (re)insurance companies. It organises international expert networks and manages discussionplatforms for senior insurance executives and specialists as well as policy-makers, regulators and multilateral
organisations. The Geneva Associations annual General Assembly is the most prestigious gathering of leading
insurance CEOs worldwide.
Established in 1973, The Geneva Association, ofcially the International Association for the Study of Insurance
Economics, is based in Geneva, Switzerland and is a non-prot organisation funded by its members.
President:Dr Nikolaus von Bomhard, Chairman of the Management Board, Munich Re Group, Munich.
Vice Presidents: Mr Michael Diekmann, Chairman of the Management Board, Allianz SE, Munich;
Mr Kunio Ishihara, Chairman of the Board, Tokio Marine & Nichido Fire Insurance Co., Tokyo;
Mr John Strangfeld, Chairman and CEO, Prudential Financial, Inc., Newark.
Members of the Board: Dr Carlo Acutis, Vice President, Vittoria Assicurazioni S.p.A., Turin;
Mr Antoine Bernheim, President, Assicurazioni Generali S.p.A., Trieste; Ms Christine Bosse, CEO, TrygVesta
Group, Ballerup; Mr Henri de Castries, Chairman of the Management Board and CEO, AXA Group, Paris;
Mr Patrickde Larragoiti Lucas, President, Sul America Seguros, Rio de Janeiro; Prof. Denis Kessler, Chairman
and CEO, SCOR, Paris; Dr Stefan Lippe, CEO, Swiss Re Group, Zurich; Mr Jos Manuel Martinez, Chairman,
MAPFRE SA, Madrid; Mr Andrew Moss, CEO, Aviva plc, London; Mr James J. Schiro, CEO, Zurich Financial
Services, Zurich; Mr Donald A. Stewart, CEO, Sun Life Financial Inc., Toronto; Mr PatrickThiele, President
and CEO, Partner Re Insurance Co., Pembroke; Mr Mark Tucker, Group Chief Executive, Prudential plc,
London; Dr Richard Ward, CEO, Lloyds, London.
Secretary General and Managing Director:Mr Patrick M. Liedtke (Head of Insurance and Finance), Geneva.
Vice Secretaries General: Prof. Jan Monkiewicz (Head of Progresand LiaisonEastern Europe), Warsaw;Mr WalterStahel (Risk Management), Geneva.
Heads of Research Programmes: Dr Christophe Courbage (Health & Ageing and Insurance Economics)
Geneva; Ms Genevive Reday-Mulvey (Four Pillars) Geneva.
Head of Communications:Mr Anthony Kennaway,Geneva.
Special Ofcers: Mr Katsuo Matsushita (LiaisonJapan & East Asia), Yokohama; Dr Bruno Porro (Chairman
of Chief Risk Ofcers Network), Zurich; Mr Gordon Stewart, (LiaisonEastern North America), New York.
Chairman of the Scientic Advisory Council:Prof. Harold Skipper, Georgia State University, Atlanta.
Former Presidents of The Geneva Association:Mr Raymond Barre, Paris (1973-1976); Mr Fabio Padoa,
Trieste (1976-1983); Mr Julius Neave, London (1983-1986); Prof. Dr Dr e.h. ReimerSchmidt, Aachen (1986-
1990); Sir Brian Corby, London (1990-1993); Drs. Jan H. Holsboer, Amsterdam (1993-1999); Mr WalterKielholz, Zurich (1999-2003); Mr Henri de Castries, Paris (2003-2008); Mr Martin J. Sullivan, New York
(2008); Mr Jacques Aigrain, Zurich (2008-2009).
7/30/2019 Geneva Association Systemic Risk in Insurance Report March2010
3/129
Systemic Risk in InsuranceAn analysis of insurance and nancial stability
Special Report of The Geneva Association Systemic Risk Working Group
7/30/2019 Geneva Association Systemic Risk in Insurance Report March2010
4/129
The Geneva Association
53 Route de Malagnou,
CH-1208 Geneva,
SwitzerlandE-mail: [email protected]
Tel: +41-22-707 66 00
Fax: +41-22-736 75 36
www.genevaassociation.org
March 2010
The Geneva Association
Systemic Risk in InsuranceAn analysis of insurance and nancial stability
Published by The Geneva Association (The International Association for the Study of InsuranceEconomics)
7/30/2019 Geneva Association Systemic Risk in Insurance Report March2010
5/129
i
Contents
Foreword 1
Executivesummary 3
0. Introduction 5
0.1.Purposeofthisreport 5
0.2.Reportstructure 5
0.3.Theeconomicandsocialroleofinsurers 5
0.3.1.Providingprotectionforindividualsandtheirpossessions 6
0.3.2.Financingtheeconomythroughthepremiumsraised 6
0.4.Theimportanceofinsuranceinnumbers 7
0.5.Thedifferenttypesofinsurersbusinessmodels 8
1. Settingthescene 11
1.1.Thenancialcrisis:thedifferenteffectsonbanksandinsurers 11 1.1.1.Insurersvs.banks:capitalandcapacity 12
1.1.2.Insurersvs.banks:stablevolumeandpricing 13
1.2.Theeffectsofthecrisisontheinsuranceindustry 15
1.2.1.Insurerswithlimitedbankingactivities 15
1.2.2.Bank-insuranceconglomerates 16
1.2.3.Insurerswithwholesalebankingoperations:AIG 17
1.2.4.Monoliners 18
1.2.5.Conclusion 20
1.3.Whydidinsurersfarebetter? 20
1.4.Conclusion 22
2. FSBandIAISdenitionofsystemicrisk 23
2.1.Size 24
2.2.Interconnectedness 25
2.3.Substitutability 26
2.4.Timing 28
2.5.Contributingfactorstotheassessmentofsystemicimportance 29
2.6.Wind-upandrun-off:insuranceindustryexperience 29
2.7.ConclusionontheFSBandIAIScriteriaforsystemicrisk 31
3. Assessingsystemicrelevanceofinsurersriskactivities 33
3.1.Investmentmanagementactivities 35 3.1.1.Assetliabilitymanagementandstrategicassetallocation 35
7/30/2019 Geneva Association Systemic Risk in Insurance Report March2010
6/129
ii
Systemic Risk in InsuranceAn analysis of insurance and nancial stability
3.1.1.1.Exposurestoothernancialinstitutions 36
3.1.1.2.Investmentinequities 37
3.1.1.3.Investmentincallablebonds 38
3.1.1.4.Programmetrading 39
3.1.2.Derivativesactivitiesonnon-insurancebalancesheets 40
3.2.Liabilityoriginationactivities 40
3.2.1.Underwritingcatastrophicrisks 41
3.2.2.Underwritinglong-termrisks 44
3.2.3.Writingbusinesswithredemptionoptions 46
3.2.4.Writinglifeinsurancewithembeddedguarantees 48
3.3.Risktransferactivities 49
3.3.1.Hedgingwithderivatives 49
3.3.2.Reinsuranceandretrocession 50
3.3.3.Insurancelinkedsecuritiesandinsurancederivatives 53
3.4.Capital,fundingandliquiditymanagementactivities 55
3.4.1.Mis-managementofshort-termfundingraisedthroughcommercial
paperorsecuritieslending 55 3.4.2.Raisingdebtorequitycapital 58
3.5.Creditprotectionactivities 58
3.5.1.Creditinsurance 58
3.5.2.Financialguarantees 59
3.5.3.CDSwriting 61
3.6.Conclusion 63
4. TheImpactofregulatoryregimesonidentiedsystemicriskissues 65
4.1.Introduction 65
4.2.Assessmentofsystemicallyrelevantriskactivitiesunder
theEuropeanandU.S.regulatoryregimes 66
4.2.1.Derivativesactivityonnon-insurancebalancesheets 66
4.2.1.1.AssessmentofEuropeaninsuranceregulation 66
4.2.1.2.AssessmentofU.S.insuranceregulation 67
4.2.2.Mis-managementofshort-termfundingraisedthrough
commercialpapersorsecuritieslending 67
4.2.2.1.AssessmentofEuropeaninsuranceregulation 67
4.2.2.2.AssessmentofU.S.insuranceregulation 68
4.3.Internationalregulatoryinitiatives 68
5. Mitigatingmeasures 71
5.1.Principlesofselectingamitigatingmeasure 71
5.2.Mitigatingmeasurestargetedtospecicidentiedissues 72
5.2.1.Measure1:Implementcomprehensive,integrated
andprinciple-basedsupervisionofinsurancegroups 72
5.2.2.Measure2:Strengthenliquidityriskmanagement 73
5.3.Additionalmeasurespromotingnancialstability 74
5.3.1.Measure3:Enhanceregulationofnancialguaranteeinsurance 74
5.3.2.Measure4:Establishmacro-prudentialmonitoring
withadequateinsurancerepresentation 74
5.3.3.Measure5:Strengthenriskmanagementpractices 75
7/30/2019 Geneva Association Systemic Risk in Insurance Report March2010
7/129
iii
AppendixA.Timelineofthecrisis 77
AppendixB.Casestudiesoftroubledinsurers 87
B.1.AIG 87
B.2.ING 90
B.3.TheHartford 91
AppendixC.Sizeanddiversicationofinsurers 93
C.1.Sizeoftop10Europeaninsurersandbanks 93
C.2.Sizeoftop10NorthAmericaninsurersandbanks 94
C.3.DiversicationofEuropeaninsurers 95
AppendixD.Insurerwind-ups:EquitableLifeandHIH 97
D.1.Casestudy:thefailureofEquitableLifeAssuranceSocietyintheU.K. 97
D.2.Casestudy:thefailureofHIHInsuranceinAustralia 99
AppendixE.EUinsuranceregulation:SolvencyII 103
AppendixF.U.S.insuranceregulationwithfocusonRBC 105AppendixG.Swissinsuranceregulation:theSwissSolvencyTest(SST) 107
AppendixH.SolvencyIIandBaselII:acomparisonofcriticalaspects 109
References 111
Glossary 113
Exhibits
Exhibit1:Breakdownofinsurancepremiumsbylineofbusinessandgeography 7
Exhibit2:Premiumspercapitainlifeandnon-lifeinsurance(inUSD,2008) 8
Exhibit3:Chronologyofthecrisismarketdevelopments2007-2009 11Exhibit4:U.S.nancialstressindex2007-2009 12
Exhibit5:Re-capitalisationandcreditlosses 13
Exhibit6:Riseandfallofwholesalefundingandinterbanklending 13
Exhibit7:Borrowingratesduringthecrisis 14
Exhibit8:Insurancepricingduringthecrisis 14
Exhibit9:Crisisimpactbyexposuretobankinginoperations 15
Exhibit10:Insurancecompaniesaffectedduringthecrisis 16
Exhibit11:AIGrevenuesbydivision 17
Exhibit12:Monolinersexposuresduringthecrisis 19
Exhibit13:Monolinerscorrelationtothehousingmarkets 19
Exhibit14:Insurerscash-owsbysource 20Exhibit15:Operatingnetcash-owsandinvestmentsbymajorEuropeaninsurers 21
Exhibit16:Estimatedcashcoverageofinsurancecompanies 21
Exhibit17:Insurancedebtissuanceduringthecrisis 22
Exhibit18:Denitionsofsystemicriskandsystemicrelevance 23
Exhibit19:Economiccapitalbyrisksource 24
Exhibit20:Interconnectedness:insurancevs.banking 25
Exhibit21:Worldwidereinsurancecapitalinows1990-2008 27
Exhibit22:ConcentrationofEuropeanprimaryinsurancemarkets 27
Exhibit23:ConcentrationoftheglobalandU.S.reinsurancemarkets 28
Exhibit24:TimingofWorldTradeCentreclaimspayments 28
Exhibit25:Approachforassessingsystemicrelevance 33
Exhibit26:Universeofactivitycategoriescarriedoutbyinsurancecompanies 34
7/30/2019 Geneva Association Systemic Risk in Insurance Report March2010
8/129
iv
Systemic Risk in InsuranceAn analysis of insurance and nancial stability
Exhibit27:Inter-connectionsofkeyriskactivitiesinwhichinsurerscanengage 35
Exhibit28:Insurersassetsaresignicantlylowerthanbanksassets
andcapitalmarketsassets 36
Exhibit29:Insurersexposuretoothernancialinstitutions 37
Exhibit30:Insurersandtheequitymarkets 37
Exhibit31:Illustrationofconvexitymismatch 38
Exhibit32:ExampleofimpactofprogrammetradingDutchpensionfundhedge
oneurolong-termswaprates(December2008) 39
Exhibit33:Catastrophelossesvs.bankfailures 41
Exhibit34:Catastrophelosscoverage 42
Exhibit35:Timingofinsuranceclaimsettlementexampleofmedicalliability
andWorldTradeCentre 43
Exhibit36:Illustrativeexample:riskofjumpinlongevity(U.K.) 45
Exhibit37:Casestudyonasbestosis 45
Exhibit38:EffectofnancialcrisisonGermanlifelapserates 47
Exhibit39:Allianzsurrendersandnancialresources(2008) 47
Exhibit40:Illustrationofreinsurancespiral 51Exhibit41:Comparisonofinsurance,reinsuranceandretrocessiongross
writtenpremiumvolumes 52
Exhibit42:Cessionandretrocessionratesformajorprimaryinsurersandreinsurers 52
Exhibit43:ILSmarketgrowthandinsurersexposuretoILS 54
Exhibit44:Evolutionofweatherderivatives 54
Exhibit45:Liquidityriskfromstocklendingandcommercialpaperfunding 56
Exhibit46:LossesavertedbybailoutofAIGssecuritieslendingoperations 56
Exhibit47:Coverageofshort-termliabilitiesbyoperationalcash-ows
andliquidassets 57
Exhibit48:Overviewofcreditinsurancemarket 59
Exhibit49:Size,substitutabilityandinterconnectednessofmonoliners 60Exhibit50:AIGinvolvementinCDSanditsinterconnectednesstomajorbanks 61
Exhibit51:ExampleofdisconnectionbetweenCDSnotionalsandunderlyingdebt 62
Exhibit52:CDSsellersandbuyersbysector(December2006) 62
Exhibit53:AIGrevenues2005bybusinesslines 87
Exhibit54:Sizeoftop10Europeaninsurerscomparedtotop10banks 93
Exhibit55:Sizeoftop10NorthAmericaninsurerscomparedtotop10banks 94
Exhibit56:RelativepremiumcontributionsbyEuropeancountriesforEuropean
topinsurers 95
Exhibit57:EquitableLifelapserates2000-2004 98
Exhibit58:U.K.andEuroareaGDPgrowthratesandU.K.pensionerincome 99
Exhibit59:NationalAustralianeconomyanddwellingmarket 100Exhibit60:PillarstructureofSolvencyII 103
Acknowledgements
ThisGenevaReportis theproductofa collectiveexercisecomprisinga steeringcommittee,several
working groups and many individual efforts that drafted and discussed all the texts plus a series of
commentsfromoutsideexpertsandexternalsupportfromOliverWymanallundertheauspicesofThe
GenevaAssociation.
ItispartofalargereffortbyTheGenevaAssociationtostimulateandtoconductworkonthecredit
crisis,newregulatoryinitiativesandtheirimpactoninsurance.
7/30/2019 Geneva Association Systemic Risk in Insurance Report March2010
9/129
v
Steering Committee Members
Aegon:PatriciaPlas
Allianz:VolkerDeville
Aviva:JimWebber
AXA:BenotClaveranne Met Life:StanTalbi
Munich Re:JoOechslin
Swiss Re:RajSingh
The Geneva Association:PatrickM.Liedtke
Zurich Financial Services: AxelP.Lehmann
Working Group Members
Aegon:ChristianPierotti,PatriciaPlas
Allianz:MichaelButtstedt,ArneHolzhausen
Aviva:HughFrancis,TimHall,NickKitching,
AXA:BenotClaveranne,FabriceLorillon
Generali:GianfrancoVecchiet
Lloyds:AlastairEvans,GavinSteele
MAPFRE:LuigiLubelli
Met Life:LoriEvangel,SusanGreenwell
Munich Re:MatthiasKubicek,MichaelMenhart
Partner Re:FranoisVilnet
Prudential plc: MilesCelic,SvenKasper, PauletteKing,SallyPadget
Prudential Financial:ElisaPuzzuoli
SCOR:PhilippeTrainar
Sul America Seguros:OswaldoMario
Sun Life Financial: ClaudeAccum Swiss Re:PhilippeBrahin,AstridFrey,ThomasHess, KarenHudson
The Geneva Association:PatrickM.Liedtke,JanMonkiewicz
Tokio Marine:MakotoHori,MasaakiNagamura
TrygVesta:MartinBgeMikkelsen
Vittoria Assicurazioni:AndreaAcutis
Zurich Financial Services:PeterBuomberger,DanielHofmann,RoySuter
TheGenevaAssociationthanksOliverWymanfortheexternalsupport.
External Support
Oliver Wyman:AlexandreCarrion,ClaudiaFell,BernhardKotanko,AndrewRear
7/30/2019 Geneva Association Systemic Risk in Insurance Report March2010
10/129
vi
Systemic Risk in InsuranceAn analysis of insurance and nancial stability
7/30/2019 Geneva Association Systemic Risk in Insurance Report March2010
11/129
1
Foreword
BasedonthetargetsputforwardbytheG20,preliminaryconsiderationsweregivenin2009
tointroducingmacro-prudentialsupervision.Thediscussionhingesonthe questionofwhether
and towhat extent systemic risksexistand how theseriskscan beaddressed bysupervisory
regulations.
Followingthere-establishmentoftheFinancialStabilityForumastheFinancialStabilityBoard
inApril2009,thediscussiontookonanewintensity.InOctober2009,theFSBincooperation
withtheIMFandtheBISpublishedareportonsystemicrisksthatincludedconcretedenitions
andadescriptionofrstregulatoryapproaches.TheninJanuary2010,theJointForumpublished
areportaddressingkeyissuesandcontainingconcreterecommendationswithregardtonancial
marketregulation.Withrespectto(re)insurance,theIAIShasofcoursealsoplayedasignicant
roleforsometimeonsystemicrisks.
Inthepast,therewasnoperceivedneedtospecicallyaddresssystemicriskininsurancebut,
givenrecentdevelopments,itwashightimefortheinsuranceindustrytoengageinthedebateon
systemicrisksandthewaytheyarehandledintermsofregulationandsupervision.The present report is devoted tothis task. The GenevaAssociation, numbering almost 80
CEOsfromtheworldslargestinsuranceandreinsuranceundertakingsamongitsmembers,has
withthehelpofindividualmemberstakenupthechallengeofstructuringandmoreaccurately
presentingkeyaspectsofthediscussionsurroundingthetopicofsystemicrisksintheinsurance
industry.
This appears to be necessary as in the public debate the business model ofthe insurance
industryisunfortunatelynotalwayssufcientlydemarcatedfromthebusinessmodelofother
nancialservicesproviders,suchasthebanks.Thewaysystemicrisksaretreatedmust,however,
take account of precisely these specic characteristics of the business models and particular
activitiescarriedoutbyinstitutions.
Thisreportisdesignedtoenrichtheongoingdiscussionalthoughitisclearthatthedebateon
systemicriskswillnotendwiththepublicationofthisreport.
Itisimportanttometopayrecognitiontotheprofessionalworkofthosewhohavecontributed
tothisreport.Thecompletionofthisrstcomprehensivereportonsystemicrisksintheinsurance
industrywithinsuchashorttimeisaremarkableachievement,forwhichIwouldliketothankthe
authorsandcontributorsonbehalfofallGenevaAssociationmembers.
DrNikolausvonBomhard
President of The Geneva Association
Chairman of the Board of Management, Munich Re Group
7/30/2019 Geneva Association Systemic Risk in Insurance Report March2010
12/129
Systemic Risk in InsuranceAn analysis of insurance and nancial stability
2
7/30/2019 Geneva Association Systemic Risk in Insurance Report March2010
13/129
3
Thenancialcrisishasexposedawsinthesupervisorysystemandengenderedcallstofurther
regulatethenancialsector.Amongthemanyproposalsunderconsiderationorimplementation
is the idea of applyingmore stringent supervisionand, perhaps, more onerous regulations to
systemically relevant institutions. This proposal is usually conceived as applying to banks.However,someinstitutionsandgovernmentshaverecentlysuggestedthatasimilarapproachbe
takentoinsurers.Thisreportexaminestheperformanceoftheinsuranceindustryduringthecrisis,
assessestheapplicationoftheFSBsproposalonsystemicrisktoinsurance,anddevelopsinitial
recommendationsto addresscurrentregulatory gapsandstrengthenindustry riskmanagement
practices.
Banksandinsurersplayedmarkedlydifferentrolesinthenancialcrisis:notonlywere
banks,notinsurers, thesourceofthe crisis, bankswere also muchharderhitbyit. Excluding
thoseinsurerswithlargequasi-bankingoperations,insurersreceivedlessthanUSD10billionin
directStatesupportduringthecrisis,comparedwithoverUSD1trilliongiventobanks.
Theinsurancebusinessmodelencompassing bothinsurers and reinsurershasspecicfeaturesthatmakeitasourceofstabilityinthenancialsystem.Insuranceisfundedbyup-
frontpremiums,givinginsurersstrongoperatingcash-owwithoutrequiringwholesalefunding.
Insurancepoliciesaregenerallylong-term,withcontrolledoutows,enablinginsurerstoactas
stabiliserstothenancialsystem.Duringthecrisis,insurersmaintainedrelativelysteadycapacity,
businessvolumesandprices.
Thosefewinsurerswhoexperiencedseriousdifculties,mostnotablyAIG,werebrought
downnotbytheirinsurancebusinessbutbytheirquasi-bankingactivities.Similarly,the
troubled monoliners (FSA, AMBAC, MBIA et al.) concentrated exclusively on nancial
guaranteesandCDSwritingandtrading.Morethan90percentofStatesupporttoinsurerswent
tothosewithsignicant,failingnon-insurancebusinesses.
TheFSB,BISandtheIMFrecentlygavetheirdenitionofsystemicrisk,whichwassupported
bytheG20nanceministersandcentralbankgovernors.AlthoughtheFSBdenitionisonlyone
amongmanyputforwardinrecentmonths,itisthemostcommonlycitedandprovidesastarting
pointforthepurposesofthisreport.
TheFSBusesthreecriteriatoassessthesystemicriskpresentedbyaninstitution:size,
interconnectednessandsubstitutability.TheIAIShasaddedtimethatis,thespeedofloss
transmissiontothirdpartiesasafourthcriterion.Thisisofparticularrelevancetoinsurance,as
insuranceclaims,unlikebankingobligations,donotimmediatelygeneratecashoutows.
Wedonotdisputethesecriteriaforsystemicrisk. Evenmoreimportantlyfortheregulatory
purposes,theyshowhowsystemicriskaccrues,nottorms,buttospecic activities ofthose
rms.
Executive summary
7/30/2019 Geneva Association Systemic Risk in Insurance Report March2010
14/129
4
Systemic Risk in InsuranceAn analysis of insurance and nancial stability
ApplyingtheFSBcriteriatothemainactivitiesofinsurersandreinsurers,weconclude
that nonepose a systemic risk. These activities include investment management (investing
policyandshareholdersfunds),liabilityorigination(providingprotectionandguarantees),risk
transfer(throughreinsurance,securitisation,etc.),andcapitalmanagement.
Noneofthesepassthetestforsystemicrelevance,foratleastoneofthefollowingreasons: theirlimitedsizemeansthattherewouldnotbedisruptiveeffectsonnancialmarkets;
theslowspeedoftheirimpactallowsinsurerstoabsorbthem,suchascapitalraisingover
timeor,inaworstcase,engaginginanorderlywind-up;
featuresoftheirinterconnectednessmeanthatcontagionriskwouldbelimited.
Wendthatonlytwo,non-core activitiesofinsurerscouldhavethepotentialforsystemic
relevance, assuming that they are conducted on a huge scale and using poor risk control
frameworks:
derivativestradingonnon-insurancebalancesheets;
mis-managementofshort-termfundingfromcommercialpaperorsecuritieslending.
Currentandalreadyapprovedinsuranceregulatoryregimes ,suchasSolvencyIIinthe
EuropeanUnion,adequatelyaddressinsuranceactivities.Theremainingquestioniswhether
existingregulationadequatelymitigatespotentialsystemicriskfromthesenon-coreactivitiesor
whetheritneedssupplementingorreplacingwithnewmeasures.
We conclude that principle-based group supervision applied to all entities within
an insurance group (regulated and non-regulated), supported by sound industry risk-
managementpractices,willmitigatepotentialsystemicriskrelatedtotheseactivities.
SolvencyIIrepresentssuchacomprehensiveandeconomicbasedregulatoryframeworkthat
itshouldnotbeconfusedwithBaselII,despitenumericalequivalence.
Wealsobelievethatinsolvenciesneednotbeavoidedatanyprice. Facedwithaverylargeevent,aninsurercanfail;but,incontrasttowhatwehavewitnessedinthebankingsector,winding-
upaninsurerisanorderlyprocessthatdoesnotgeneratesystemicrisk. Webelievecross-
bordercrisismanagementremainsanarearequiringimprovedcoordinationamongsupervisors.
Inseekingtocloseremaininggapsinthesupervisoryframework,regulatorsshouldavoid
thetemptationtoplacespecialburdensonspecicinstitutions. Thisapproachcoulddistortthe
insurancemarketbyskewingpricing,reducingaggregatemarketrisk-bearingcapacity,drawing
supervisorsattentionawayfromriskyactivitiesgoingonelsewhere,andcreatingmoralhazard
inthesetoobigtofailinstitutions.Theconsequencesofgettingsystemicriskreformswrong
wouldnotonlybeseverelydamagingtotheinsuranceindustrybuttotheeconomyaswell.
Werecommendvemeasures.Thersttwoareputforwardtoaddressgapsinregulation
andindustrypracticeidentiedinthisreport;measuresthreetoveaim tostrengthennancialstability:
1. Implement comprehensive, integrated and principle-based supervision for insurance
groups.
2. Strengthen liquidity risk management.
3. Enhance regulation of nancial guarantee insurance.
4. Establish macro-prudential monitoring with appropriate insurance representation.
5. Strengthen risk management practices.
These measures demonstrate the industry engagement to contribute to the discussion on
systemicrisk.Theindustrystandsreadytotakeanyactionnecessarytomaintainstabilityinthe
insurancesystemitself,contributetothestabilityoftheoverallnancialsystem,andperformitsenablingroleintherealeconomy.
7/30/2019 Geneva Association Systemic Risk in Insurance Report March2010
15/129
5
0. Introduction
0.1. Purpose of this report
Thenancialcrisishasengenderedwidespreadcallsto furtherregulatethenancialsector.
Among the many proposals under consideration or implementation is the idea of applyingmore stringent supervision and, perhaps, more onerous regulations to systemically relevant
institutions.Thisproposalisusuallyconceivedasapplyingtobanks.However,someinstitutions
andgovernmentshaverecentlysuggestedthatasimilarapproachcouldbetakentoinsurers.
Thisreportexaminesthisproposal.
Unless stated otherwise, the term insurance is held to mean both primary and
reinsurance.Fortheavoidanceofdoubt,thenancialguarantorsknownasmonoliners
areexcludedfromthismeaning.
0.2. Report structure
Inthischapterwegiveabriefintroductiontotheinsuranceindustry,itseconomicroleandits
mainbusinessmodels.InChapter1weexaminethedifferentrolesplayedbybanksandinsurers
intherecentnancialcrisis,whichhasgenerallyhitinsurersfarlesshardthanbanks.
Chapter2considerstheFSBsproposedcriteriafordesignatinganinstitutionassystemically
relevant. We conclude that, in the rst instance, it is not insurance institutions but insurance
activitiesthatshouldbeconsideredinrelationtosystemicrisk.
Chapter3isdevotedtoexaminingthemainriskactivitiesofinsurersinlightoftheFSBcriteria.
Weconcludethatthesecriteriaexplainwhymostinsurersactivitiesdonotposeasystemicrisk
likesomebankingactivitiesdo.
Havingidentiedtwoactivitiesthatcanpotentiallyaddtoasystemicriskscenario,Chapter4considerstheextenttowhichtheyarealreadyadequatelydealtwithbyexistingregulations.
Finally,inChapter5,wesuggestmeansforclosingthegapsincurrentregulationsandpropose
waysofimprovingourownpractices.
0.3. The economic and social role of insurers
Insurers main functionsthe functions of insuranceare the provision of protection by
acceptingrisksfrompolicy-holders,poolingtheserisks,managingthemactivelyandpotentially
transferringtheminparttoreinsurers.
Duetotheirroleandthelong-termhorizonofmanyinsurancecontracts,insurershavelarge
amountsofinvestmentsundertheirmanagementtobackfutureclaimsandarethereforesignicant
players,withothernancialinstitutions,innancialintermediationandcapitalaccumulation.
7/30/2019 Geneva Association Systemic Risk in Insurance Report March2010
16/129
6
Systemic Risk in InsuranceAn analysis of insurance and nancial stability
0.3.1. Providing protection for individuals and their possessions
Insuranceprovidesamechanismforthepoolingandtransferofthenancialconsequences
ofrisk.Inexchangeforapremium,theinsurercommitstonanciallycompensatingthepolicy-
holderforexpectedlossesresultingfromthecontract,operatingexpensesandcapitalcosts.
Insuranceactivitiesarebroadlydividedintolifeandnon-lifeinsurance.Lifeinsuranceprotects
againstthelossoffamilyincomeduetodisabilityordeathoragainsttheriskofoutlivingyour
nancial resources. Non-life insurance contracts provide protection against damage, loss, or
injurytotheinsured,aswellaslegalliabilityfordamagescausedtootherpeopleortheirproperty.
Reinsurersprovideprotectiontoinsurersthemselves.
Therecoursetoinsuranceallowspeopleandbusinessestoengageinactivitiesthatotherwise
wouldbetoofraughtwiththeriskoflosses.Theabsenceofinsurancewouldcontributetorisk
aversion and dampen the entrepreneurial spirit and many business initiatives would not be
undertaken.Withoutinsurance,noonecouldaffordtotaketherisksnecessarytogrowahealthy
economy.
Insuranceisalsoamajorcontributortoeconomicwelfareandfairnessandanessentialproviderofsocialprotection:
Contractual savings, particularly life insurance and pension, play an essential role in
providinglong-termnancialsecurityandgrowth,inparticularincountrieswherethisrole
isnotcoveredbytheState.
Non-lifeinsurancehelpsindividuals andrm hedge against importantrisksaccidents,
property,re,interruptionofproductiveactivity,generalandprofessionalliability,motor
thusallowingthemtoengageinseveraleconomicactivitiesandexpandtheproduction
frontier.
Becausetheyassumerisks,insurerspayparticularattentiontomanagingtheserisks,including
transferringthemtoreinsurers.Insurersalsoinvestgreatlyinhelpingtheirpolicy-holdersmanageandmitigatetheirownrisks.Therisksinsurerstakeonarediverse,andinsurersconstantlylook
fornewareasinwhichtheymayacceptasinsurablesomeoftherisksfacedbyotheractorsinthe
economy.
Insurersarenottheonlyentitiestoacceptrisksbanksalsoacceptrisks.Howevertheirrisks
arefundamentallydifferentfrominsurancerisks.Insuranceriskisidiosyncraticand,forthemost
part,independentof theeconomiccycle. Byacceptingshort-term,liquiddemanddepositsand
grantinglong-termloans,banksassumetwomajorrisks:creditriskrelatedtolendingactivities
andliquidityriskduetothemismatcharisingfromborrowingshortandlendinglong.Thesebank-
specicriskstendtobecorrelatedwiththeeconomiccycle.
0.3.2. Financing the economy through the premiums raised
For life insurance and many lines of non-life insurance, claims payments occur many
years after premiums have been collected. Consequently insurers have stable asset
portfolios to manage with a long-term horizon. Their investment activities in equities
and bonds provide a key link in capital markets, supporting the market place for savers
andborrowers.
Insurance therefore plays a key role in nancial intermediation by re-investing long-term
savingsthroughdebtandequityholdings.Insurersfundingprolesanddiversecustomerbases
allowthemtotakealong-termassetallocation,whichiskeyfornancingenterpriseswithlong-
termcapital.
7/30/2019 Geneva Association Systemic Risk in Insurance Report March2010
17/129
7
Insurersseeinvestmentriskdifferentlyfromassetmanagersasaninsurancecompanycannot
ignoretheliabilitysideofitsbalancesheet.Atypicalassetmanagerinvestsonbehalfofclients
andisfocusedonmaximisingthevalueofinvestmentsrelativetoaprescribedbenchmark.But
ininsurance,theinvestmentfunctionmustnotonlyachieveadequatereturns,itmustmanagethe
potentialmismatchinassetsandliabilitiesasaresultofchangesincapitalmarketconditions.
0.4. The importance of insurance in numbers
The importance of insurance is reected in the volume of premiums generated and
assetsmanaged.
WorldwideinsurancepremiumsamounttoUSD4.4trillion,whichrepresents7.3percent
ofworldwideGDP.Theexhibitbelowshowsthesplitofinsurancepremiumsbygeographyand
betweenlifeandnon-life.
Exhibit 1: Breakdown of insurance premiums by line of business and geography
Premium volume by business segment
USD BN, 2008
Source: Swiss Re sigma, IAIS Global Reinsurance Market Report 2009
1. Gross reinsurance premiums shown by region of ceding insurer
159
Europe
1,703(40%)
North America
1,344(32%)
59 66 1618
739
996
707
625
719
273
76
83
Asia Pacific
1,012(23%)
ROW
159(4%)
1,782(42%)
2,43658%
Non-life
insurance
Lifeinsurance
Reinsurance1
InsurershaveUSD18.7trillionofassetsundermanagement,1whichisroughly11percentof
theworldstotalnancialassets.
Insurancepenetrationishighlycorrelatedwiththelevelofdevelopmentofcountries.OECD
countriesachieveveryhighpenetrationlevelswithpremiumspercapitalforbothlifeandnon-life
ofaroundUSD3,000percapita(2008),asopposedtolessthanUSD500indevelopingcountries.
1. International Financial Services London Research, October 2009.
Introduction
7/30/2019 Geneva Association Systemic Risk in Insurance Report March2010
18/129
8
Systemic Risk in InsuranceAn analysis of insurance and nancial stability
Exhibit 2: Premiums per capita in life and non-life insurance (in USD, 2008)
1,856
2,022
2,448
2,133
1,187725
220107811917
79692947380
500
1,000
1,500
2,000
2,500
3,000
3,500
4,000
4,500
North
America
Western
Europe
Japan &
industr.
As iancountries
Central and
Eastern
Europe
Latin
America and
Caribbean
Middle East
and Central
As ia
South and
East Asia
Af rica
Life Non-Life
Source: Swiss Re Facts and Figures 2008
0.5. The different types of insurers business models
Insuranceactivitiesarecarriedoutbydifferenttypesofplayers.Theseare:
purelifeinsurers;
purenon-lifeinsurers(Property&CasualtyInsurers);
compositeinsurers(mixoflifeandnon-lifeinsurers); reinsurers;
insurerswithlargenon-insuranceoperations;
monoliners/nancialguarantee.
Life insurance and non-life insurance must be provided in different legal entities; this is
understoodtohavebecomeageneralrequirementinallregulatoryjurisdictionsoutofadesireto
protectlifeinsuranceinvestments(heldonbehalfofsavers)fromadefaultinanon-lifeinsurance
companyofthesamegroup.Many,butbynomeansall,ofthelargeglobalinsurers,arecomposite,
offeringbothlifeandnon-lifepoliciestomakebestuseoftheirdistributionnetworkandcustomer
reach.Thelifeandnon-lifeinsurancebusinessescontinuetobeseparatelegalentities.Allinsurers
carryoutsomeformofinvestmentmanagementactivity.Variouspartsoftheinvestmentactivitymaybeoutsourced,butthecriticalresponsibilityforappropriatemanagementofasset-liability
matchingremainswiththeinsurerinanycase.
Reinsurers are insurers for insurers, allowing insurers to reduce their exposure to peak
lossesandconcentration.Overthelast150years,reinsurancehasevolvedasaneffectivemeans
ofcopingwiththegrowingnumberandincreasinglycomplexnatureofrisks,representingan
essentialproviderofriskdiversicationonaglobalbasisandstabilisingfactorfortheinsurance
system,andthusthewidereconomy.
Morerecently,someinsurershavemovedintoareasofothernancialservicestovarying
degrees:
Certaininsurancegroupscontainabankalongsidetheirinsurancesubsidiariesandutilisetheopportunitiestocross-sell.Thistakesthecommondistributionmodelofbancassurance
(retailinginsuranceproductsthroughanagreementwithabank)onestepfurther.
7/30/2019 Geneva Association Systemic Risk in Insurance Report March2010
19/129
9
Moreover,inlinewithnancialinnovation,insurershavebecomeinvolvedinderivative
activities.Certaininsurershavegonebeyondusingtheseinstrumentsforhedgingtheirown
investmentsandsetupdedicatedsubsidiariestransactingincomplexderivativeinstruments.
Finally, certain groups have developed into nancial conglomerates with substantial
bankingandderivativetradingactivities.We discuss the impact of the banking crisis on the banking subsidiaries of insurers in
Chapter1.
Monolinershaveadifferentbusinessmodelfromotherinsurers.Monolinerssellonly,inone
formoranother,nancialguaranteesforinvestorsonanancialinstrument.Thesedifferencesin
businessmodelcametotheforeintherecentcrisisasmonolinersriskprolesweresimilartothat
ofabank,duetothepredominanceofcreditrisk.InChapter5weputforwardourviewthatthe
regulationofmonolinersshouldbeconsistentwiththatofbankscarryingoutthesameactivity.
Introduction
7/30/2019 Geneva Association Systemic Risk in Insurance Report March2010
20/129
10
Systemic Risk in InsuranceAn analysis of insurance and nancial stability
7/30/2019 Geneva Association Systemic Risk in Insurance Report March2010
21/129
11
Webegininthischapterbydescribingthedifferentrolesofbanksandinsurancermsinthe
nancialcrisis,anditsdifferenteffectsonthem.Thisisimportantwhenconsideringtheextension
toinsurersofregulatoryproposalsinitiallydesignedforbanks.
1.1. The nancial crisis: the different effects on banks and
insurers
Thestoryofthenancialcrisisisnowfamiliar,andadetailedtimelineofthecrisisisprovided
intheappendix.DecliningpropertyvaluesandclimbingdefaultratesintheU.S.ledtowrite
downsin mortgageassets. In 2007Countrywide,the largest subprimemortgageoriginatorin
theU.S.becameinsolventandBearStearnsbailedouttwoofitshedgefundsthathadinvested
heavilyinriskysecurities.InMarch2008,BearStearnswasrescuedfromfailurebyJPMorgan.
Exhibit 3: Chronology of the crisis market developments 2007-2009
800
1,000
Peak in equity
market prices
Br
Leh
(1
Bear Stearns
collapse
(March 2008)
bps)
800
0
200
400
600(October 2007)
CDSSpreads(
0
200
400
600
AllexceptS&P500
(indexed)
Jan-07 Jul-07 Jan-08
Source: Bloomberg, Datastream (US CDS spreads have been
US ins. sector CDS 5Y EU ins. sector CDS 5Y
eakdown of
man Brothers
5/09/2008)
AIG bailout
(16/09/2008)
Equity market
recovers
120
(March 2009)
S&
(in
0
20
40
60
80
P500
exed)
Jul-08 Jan-09 Jul-09 Jan-10
djusted to exclude AMBAC and MBIA)
S&P 500VIX US (USD ) 1M FWD 5Y swaption implied vol.
ByOctober2008AIG,FreddieMacandFannieMaehadbeenrescuedwithtaxpayersfunds
andLehmanBrothershadbeenallowedtofail.Uncertaintyaboutthesolvencyofmajorbanks
1. Setting the scene
7/30/2019 Geneva Association Systemic Risk in Insurance Report March2010
22/129
12
Systemic Risk in InsuranceAn analysis of insurance and nancial stability
withexposuretotoxicassetscausedtheinterbankmarketstoseizeup.Tostaveoffeconomic
calamity, governments propped up the nancial sector with extended deposit guarantees and
massiveinfusionsofdebtandequitycapital.
Not all kinds of nancial rm were equally responsible for the crisis, nor equally
affected by it. Specically, the crisis had a less dramatic affect on insurers than on banks.The crisis hit insurers later, required far less additional capital and had less affect on new
businessvolumes.
Exhibit 4: U.S. nancial stress index 2007-2009
5
6
Values higher
than 2 indicate
March 17:
Collapse of
Bear StearnsOct
0
1
2
3
severe stressof equity
-2
-1
Jan-07 Jul-07 Jan-08
Source: Zurich Financial Services
The index is a weighted average of the following seven weeklycorporate bond in the investment grade sector, Equity volatility
, - ,
transformation (for example log-differenced) and is standardise
The weights of each series is determined by a principal compon
November 19:
Fear of huge credit
card debts
ober 8:
March 6:
Renewed concerns
about banks' health
September:
Collapse of
Lehman and AIG
arkets
Jul-08 Jan-09 Jul-09 Jan-10
ata series: Spread on corporate bond in the high yield sector, Spread oneasured by CBOE VIX, Equity index measured by S&P 500, LIBOR-T-Bill
. -
, in other words each series has mean zero and standard deviation one.
ent analysis. Finally, the compound series is standardised.
1.1.1. Insurers vs. banks: capital and capacity
Lossesintheinsuranceindustryhavebeenonlyasixthofthoseatbanks,andthenewcapital
raisedonlyaninth.AIGaloneaccountsfor58percentofnewcapitalintheinsurancesector
and36percentofthecreditlosses,whichwereincurrednotinthecourseofcharacteristic
insuranceactivitiesbutbyAIGsmassivecreditdefaultswap(CDS)exposures .ExcludingAIGfromthegures,bankshadtoraise20timesmorecapitalthaninsurers.
Thesecomparativelysmalllossesunderscorethattheinsuranceindustryhaslostlittlebusiness
capacity,asmeasuredbyshareholdersequity.Infact,mostinsurerswereinapositiontoabsorb
comparativelylargecreditlossesontheirbalancesheets,whereasmanybankshadtotakerecourse
topublicfunding.
7/30/2019 Geneva Association Systemic Risk in Insurance Report March2010
23/129
13
Exhibit 5: Re-capitalisation and credit losses
Total capital raised1
Cumulative from 2007 (USD BN)
Total capital raised
Cumulative from 2007, as % of 2006 Shareholders Equity
Credit losses
Cumulative from 2007, as % of 2006 Shareholders Equity
Credit losses
Cumulative from 2007 (USD BN)
0
500
1,000
1,500
2,000
Insurers Banks
X 6
USD 271 BN
USD 1,715 BN
0
300
600
900
1,200
1,500
Insurers Banks
Europe AIG Other American Asia ING
USD 170 BN
USD 1,468 BN
X 958%
16%0%
20%
40%
60%
Insurers Banks
X 3.5
68%
26%
0%
20%
40%
60%
Insurers Banks
X 2.6
Banks suffered larger losses and required more
capital on an absolute basis
Source: Bloomberg (as at 10 Feb 2010), Oliver Wyman analysis, Datastream
1. This is equity and preferred share capital raised, from states and through the capital markets. It does NOT include asset relief or lending
provided by states that did not require a capital consideration
as well as on a relative basis
1.1.2. Insurers vs. banks: stable volume and pricing
Intherunuptothecrisis,leverageincreasedmaterially.Asset-backedcommercialpaperin
issuancepeakedattheendof2007,fallingawayduring2008asappetitesfortheunderlying
securitiesdeclinedandBearStearnswasrescuedbyJPMorgan.Asthecrisisdeepened,interbanklendingdroppedbymorethanhalfintheU.S.andliquiditywasrapidlywithdrawnfromnancial
markets.
Exhibit 6: Rise and fall of wholesale funding and interbank lending
US Asset Backed Commercial Papers (ABCP) in issuance & US Interbank lending volumeIndexed at 100 in December 2005
0
20
40
60
80
100
120
140
160
180
200
Dec-05 Jun-06 Dec-06 Jun-07 Dec-07 Jun-08 Dec-08 Jun-09
Asset Backed Commercial Paper Interbank Lending
Source: US Federal Reserve, Oliver Wyman analysis, Federal Reserve Board
Q3 2008: +72% since Jan 06
Q2 07: Withdrawal of
ABCP liquidity
Fallen 62%, from
USD 1.3TN to 0.4TN
Q3 08: Collapse of
Lehman shuts
interbank market
Index31/12/2005=100
Setting the scene
7/30/2019 Geneva Association Systemic Risk in Insurance Report March2010
24/129
14
Systemic Risk in InsuranceAn analysis of insurance and nancial stability
Duringthecrisis,centralbankspumpedliquidityintothebankingsectorinanattempttodrive
downinterestratesandencourageprivatesectorborrowing.However,therewasasignicant
increaseincomplaintsaboutcreditavailabilityandcreditcostbybusinessclientsandconsumers.
Exhibit 7: Borrowing rates during the crisis
5%
6%
EURIBOR and average interest rate on ne
December 2005-2009
1%
2%
3%
4%
InterestRates(%)
0%
Dec-
05
Mar-
06
Jun-
06
Sep-
06
Dec-
06
Mar-
07
Jun-
07
S
0
EURIBOR EU average interest on
Source: ECB, Oliver Wyman Analysis
loans
Q3 08: EURIBOR falls asBase Rate is slashed
p-
7
Dec-
07
Mar-
08
Jun-
08
Sep-
08
Dec-
08
Mar-
09
Jun-
09
Sep-
09
Dec-
09
new corporate loans Spread over EURIBOR
However,insurancepremiumshaveremainedrelativelystableduringthecrisis,reectingthe
industryssuperiorcapitalposition.Insurancemarketsremainedliquidandaclearingpricewas
alwaysavailable.
Exhibit 8: Insurance pricing during the crisis
Non-Life Insurance price changes i
year)
Casualty
Q1 & Q2 2007 Q1 &
France 10-20 1
Germany 10-20 1
UK 10-20 1
Italy 0
Spain 10-20 1
Property
Q1 & Q2 2007 Q1 &
France 10-20 1
Germany 10-20 1
UK 10-20
--
Spain 10-20 2
Motor
Q1 & Q2 2007 Q1 &
France 10-20
Germany 10-20
UK 10-20
Italy 0
Spain 10-20 2X% Decrease Flat
Source: Marsh EMEA Insurance Reports 2007,
n Europe (% from previous
Q2 2008 Q1 & Q2 2009
0-20 0
0-20 0-10
0-20 0-10
0-10 0
0-20 20-30
Q2 2008 Q1 & Q2 2009
0-20 0-10
0-20 0-10
0-10 0-10
- -- -
0-30 20-30
Q2 2008 Q1 & Q2 2009
0 0
0-10 0
0-10 0
0-10 0-10
0-30 0X% Increase
2008 and 2009
7/30/2019 Geneva Association Systemic Risk in Insurance Report March2010
25/129
15
1.2. The effects of the crisis on the insurance industry
Althoughthecrisishasgenerallyhadfarlesseffectoninsurersthanbanks,someinsurersdid
runintoserioustrouble.Forexample,intheU.S.asofmid-2009onlythreeinsurancecompanies2
had taken TARP3 funds.At the time 592 bankshad accessed the programme. Ingeneral, the
severityof insurers lossesdependedon theirbusinessmodel. Specically,thoseinsurerswithlargebankingoperationsorotherexposuretocreditrisksufferedthegreatestlosses:morethan90
percentoftheStatesupporttotheentireinsuranceindustrywasgiventothosefewinsurerswith
signicantbankingactivities.4
Exhibit 9: Crisis impact by exposure to banking in operations
Type
Size and risk of banking operations within insurance companies triggered insurers performance
None/limited
banking activities
Insurance companies
with none or
limited banking-
type operations
Bank-insurance
conglomerates
Insurance
conglomerates with
significant banking
operations in multiple
countries
Wholesale banking
operations
Insurance companies
engaged in
highly risky
wholesale banking
activities using
non-insurance entities
Monoliners
Insurance
companies selling
only financial
guarantee insurance
highly leveraged
and concentrated on
US public and
structured finance
Some examples
of exposure to US
housing market
leading to
State intervention
No evidence tosuggest systemic risk
Problems in banking
operations easily
sufficient to
overwhelm total
conglomerate
Insurance balancesheet ring-fenced
Severe problems
in wholesale
credit operations
unconnected
to insurance
balance sheet
Systemic threat
Severe losses led to
questioning of
business model
in general
(AMBAC, MBIA)
Systemic threat
Charac-
teristics
USD 8 BN USD 40 BN USD 180 BN Questioning of
business model
Perfor-
mance
State
support1
1. State support reflects capital injections and asset support provided by states. Exchange rates as of 31.12.2008
Source: Bloomberg, Company Reports, Oliver Wyman press search, Oliver Wyman analysis,
1.2.1. Insurers with limited banking activities
Insurerswithlimitedbankingoperationsincurredlimitedlosses.Despitetheturmoilinthe
interbanklendingmarkets,theinsurancemarketsremainedopenandinsurersoperatedasnormal.
Insurerswerecontaminatedbythebankingsectorinseveralways:
mark-to-marketdecreasesinvaluationof(subprime)mortgage-backedsecurities,corporatebondsandequitiesheldbyinsurers;
reducedliquidityofcertainassets;
increasedvolatilityofnancialmarkets(e.g.forvariableannuities);
exposurestodefaultedbanks(e.g.Lehman);
liquiditycrunchinthebankingoperationsofinsurers;
generaleconomicslowdown.
Lifeinsurerssufferedgreaterlossesthannon-lifeandcompositeinsurersbecausetheyhad
relativelygreaterexposurestonancialinstruments,throughembeddedderivativesinliabilities
2. U.S. Department of Treasury Ofce of Financial Stability Transactions Report (for the period ending July 16th 2009).
3. Troubled Asset Relief Program.
4. State support reects capital injections and asset support provided by States.
Setting the scene
7/30/2019 Geneva Association Systemic Risk in Insurance Report March2010
26/129
16
Systemic Risk in InsuranceAn analysis of insurance and nancial stability
andlargerinvestmentportfolios.TheU.S.lifeinsurersTheHartfordandLincolnreceivedState
aid to bolster their capital, having suffered losses from their exposure to devalued nancial
instruments.InJapan,YamatoLifeledfor bankruptcy.U.S. creditmarketexposureandhigh
levelsof anxiety onDutchnancialmarkets, contributing to widespread fears that any large
Dutchprovidercouldbeatrisk,ledAEGONtoaccepttemporaryStateaid.
Nevertheless,lifeinsurersexposuretoCDSandotherstructuredproductswassmallcompared
withbanks.Thisisaresultoftheirstrictunderwritingpoliciesandmodestlevelsofactivities
undertakenoffbalancesheetorthroughsubsidiaries.
Exhibit 10: Insurance companies affected during the crisis
Lincoln received USD 950 MM of funding through the US
Troubled Asset Relief Programme
Having suffered 5 consecutive loss making quarters, their
capital position had been eroded Losses were chiefly generated through their exposure
to the US housing market, through securitizations and
direct lending
Market commentators did not consider Lincoln to be in
immediate solvency danger, and that the recourse to TARP
funds may have been opportunistic
The funds have yet to be repaid
Lincoln National Corporation
US Life Insurer
33rd largest Japanese life insurer
Filed for bankruptcy on 10 October 2008, nearly a month
after the rescue of AIG
Debts exceeded assets by 11.8 BN (USD 117 MM) Operational inefficiency resulted in a high cost base
In an attempt to cover these costs, Yamato chased higher
yields, investing in the US housing markets
Bankruptcy was a result of poor investment and operational
management, rather than liquidity or exposure to other
financial institutions
Yamato Life
Japanese Life Insurer
USD 1.7 BN realized investment losses in 2008
(40% on CMBS portfolio)
Losses on assets trigger USD 1.0 BN of charges forguarantees in VA business from Q3 2008 to Q3 2009
VAs accounted for over 20% of The Hartfords USD 3
BN profit in 2007
VA guarantees include annual withdrawals of up to 7%
USD 5.9 BN raised capital
USD 3.4 BN capital injection by TARP
USD 2.5 BN invested by Allianz
The Hartford
US Composite InsurerAEGON
International Life Insurer(HQ in Netherlands)
Historic exposure to US credit markets from large US
operations (~ 2/3 of its business)
Total gross credit exposure of 390 MM to Lehman and
Washington Mutual
2.5 BN unrealised losses on credit portfolio
2008 strategy shift - already re-balancing
geographic profile
Dutch financial sector events created contagion fears
3 BN injection from Dutch government to bolster capital
and allow AEGON to retain AA rating
33% repaid in December 2009
Source: Oliver Wyman analysis, annual reports, press research
1.2.2. Bank-insurance conglomerates
Unsurprisingly,bank-insuranceconglomerateswerehitharderbythecrisisthanpureinsurers.
Thereislittledifferencebetweentheimpactofthecrisisonabank-insuranceconglomerate(e.g.
Fortis)andonaninsurance-bankconglomeratewithasimilarbankoperation(e.g.ING).5
ING is the largest insurance/bank conglomerate. Including asset relief ING received
more than USD 40 billion in State support. Its difculties stemmed from its banking
operations and especially its acquisition of a U.S. thrift when expanding its online savings
division.ThisacquisitionmadeINGsubjecttoregulationbytheOfceofThriftSupervision
(OTS). To comply with local thrift regulation, which requires more than 55 per cent of
assets to be allocated to mortgages, ING acquired a large portfolio of MBS, backed by
Alt-Amortgages.
When default rates on Alt-A mortgages began to climb, the market value of the
MBS portfolio plummeted, requiring ING to increase the quantity of capital it held
against them. In 2008, ING reported fair value losses of 2.6 billion on credit assets and5. By bank-insurance conglomerate we mean a combined bank/insurance group listed as a bank; by insurance/bank
conglomerate we mean a combined bank/insurance group l isted as an insurer.
7/30/2019 Geneva Association Systemic Risk in Insurance Report March2010
27/129
17
1.7 billion on equity securities. In October 2008 the Dutch government invested
10billioncapitalinINGandinFebruary2009ittookover80percentofINGsAlt-Aportfolio
at90percentofitsfacevaluefor20billion.6
Thebankinsurancemodelhashowevernotbeeninvalidatedasaresultofthecrisissincenot
allbank-insuranceconglomerateswerehoweverhitbythecrisis.Conglomerateswithbanksthathadstrongliquiditypositionsandlowinvolvementinstructuredproductscameoutofthecrisis
relativelyunharmed(BNPParibas,HSBC,CrditAgricole).
1.2.3. Insurers with wholesale banking operations: AIG
AIGwastheheadlinenear-failureinthe insuranceindustry.However,thesourceoftrouble
wasnotitsinsuranceactivitiesbutitsFinancialProductsdivision(AIGFP).AIGFPwasfounded
in1987asAIGscapitalmarketsdivision,domiciledinLondon.SincetheAIGholdingcompany
wasregisteredwithanequivalentregulator,theUSOfceofThriftSupervision(OTS),AIGFP
wasabletoevaderegulationbytheU.K.FinancialServicesAuthority(FSA).
AlthoughAIGFPhadalwayscontributedonlyasmallportionofAIGsrevenues,itmadehighlyleveragedtransactions.AsofSeptember2008,thenotionalvalueofAIGFPsderivatives
portfolio,concentrated in U.S. housing market and corporate CDOs and CLOs, was USD 2.7
trillion,ofwhichUSD440billioninwrittenCDSwasguaranteedbyAIGHolding.
Exhibit 11: AIG revenues by division
Asset Mgmt.
Foreign Life &
Retirement
28% US non-life Ins.32%
AIG 2005 revenues by business lines
5%
AIG FP
3%
US Life &
Retirement
15%
Foreign
General Ins.
11%
Financial Services
w /o AIGFP
6%
Sources: AIG Annual Reports
1. 2005 revenues shown as they represent the peak of
AIG FP vs. AIG revenues 2003-2008
80
120
160
inUSDB
N)
-80
-40
0
2003 2004 2005 2006 2007 2008
Revenues(
AIG FP Other AIG entities
IG FPs contribution to AIG revenues
AIGsproblemsbeganin2007.FollowingdowngradesofU.S.subprimesecurities,AIGsCDS
counterpartiesdemandedcashcollateral.InSeptember2008,AIGscreditratingwasdowngraded
from AA- to A-, triggering further cash collateral calls on its CDS contracts and securities
lendingprogramme.Unabletomeetthesecalls,AIGwasbailedoutbytheU.S.Treasuryon18th
September.Bytheendof2009,AIGhadreceivedatotalofUSD182billionoftaxpayersfunds. 7
TherearethreeimportantfeaturesofAIGFPsbusinessthatshouldbenoted.First,itwas
highlyleveraged:thetotalexposureofAIGFPwasmassivelygreaterthanitsassetbase.Such
leverageisnotpossibleinaregulatedinsurancebusiness.Second,thecollapsewastriggered
bya collateralcall, notan actualcredit event.Again,this isgenerally notpossible ininsurance
6. ING annual reports.
7. USD 70 billion government investment, USD 60 billion credit line and USD 52.5 billion to buy mortgage-linked assets
owned or backed by AIG (Bloomberg).
Setting the scene
7/30/2019 Geneva Association Systemic Risk in Insurance Report March2010
28/129
18
Systemic Risk in InsuranceAn analysis of insurance and nancial stability
contracts,whichpayoutonlywhenaninsuredevent(are,theft,lossoflife,etc.)hashappened.
Third,usingregulatoryarbitrage,AIGFPoperatedeffectivelyunregulated.Responsibilityfor
supervisionhasbeenacknowledgedbytheOTS.EventhoughtheOTSwasawareofAIGFPs
CDSexposureandtheguaranteesissuesbyAIG,theOTSdidnotattributeanyliquidityriskto
thissituation.Haditdoneso,theOTSwouldhavehadthepowertorequestreductionofCDS
exposure.
ThatAIGwasengagedintheseactivitiesatthisscalewasclearlyafactorwhichworsenedthe
globalcrisissignicantly.Interestingly,however,itcausedlimiteddamagetothecoreinsurance
businessesofAIG.
1.2.4. Monoliners
The group of nancial guarantors known as the monoliners (FSA, AMBAC, MBIA
etal.)wereamongthecasualtiesandarguablyamongthecausesofthecrisis.However,they
haveaverydifferentbusinessmodelfromtraditionalinsurers,withtheirexclusiveconcentration
onnancialguarantees.Ithasbeenarguedthatthe monolinersplayeda roleingeneratingandamplifyingthecrisis,andquestionshavebeenraisedabouttheirbusinessmodel.
Monolinershavehistoricallyprovidednancialguaranteestoreducetheborrowingcostof
U.S.municipalities(creditenhancement).Theirsolebusinesswastotakecreditrisk.Thecredit
riskhistoricallytakenwasdiversiedmunicipalriskofhighquality.Municipalshadverylow
losseshistorically.Asaresult,monolinershadacapitalpositionbasedonthatlowriskcredit
prole.Theybranchedout intostructured nancein the1990s rstinsuringtimely paymentof
interestandprincipaloffundedcashbonds;theylatermovedintonancialguaranteeassociated
withCDSofinvestmentgradecorporatecredit.Lastlytheytookonnancialguaranteesassociated
withCDSofCDOandMBSthatweresuperAAArated.
Theseactivitieshavethepotentialtotransmitlosseswhenthecreditratingoftheproviderof
thenancialguaranteeisdowngraded.Thisleadstoadevaluationofthewrappedsecuritiesand
mark-to-marketlossesforwhoeverholdsthem.
Themonolinersbuiltupundiversied,highlyleveragedportfolios,takingadvantageofthe
fact that a CDS backed by MBS requires less capital coverage than a municipal bond.This
businessmodelreliedheavilyonstrongcreditratings(asdidAIGFP)and,atthesametime,was
vulnerabletoerrorsofriskestimation.
Monolinersgenerallydidnotagreetopostcollateralundertheircollateralservicingagreements
because they lacked sufcient liquidity to support collateral calls. This heightened anxiety at
theircounterparties,whobecameconcernedthattheseinstitutionswouldcollapseleavingthe
counterpartiesunprotected.Thissituationeffectivelycreatedacondencecrisisinthemonoliners
evenathighratings.
Monolinersbusinessisdifferenttotraditionalinsurance,andwewouldsuggestthatregulators
considerthesystemicriskfromsuchinstitutionsexactlyastheywouldforanyhighly-concentrated
creditinstitutioninthebankingsectororelsewhere.
7/30/2019 Geneva Association Systemic Risk in Insurance Report March2010
29/129
19
Exhibit 12: Monoliners exposures during the crisis
MBIA
Net par amount written by bond type
0%
25%
50%
75%
100%
2006 2007 2008
US Public f inance US Structured f inance
Non-US Publ ic f inance Non-US Structured f inance
1.1% 1.1% 1.2%
0.4%
-0.9%
-2.0%
-1.0%
0.0%
1.0%
2.0%
2004 2005 2006 2007 2008
Source: MBIA and AMBAC annual reports
Monoliners leverage Monoliners exposures
1.1% 1.1% 1.2%
0.5%
0.1%
0.0%
0.5%
1.0%
1.5%
2.0%
2004 2005 2006 2007 2008
MBIA
Equity/net par amount outstanding
AMBAC
Equity/net par amount outstanding
AMBAC
Gross par amount written by bond type
0%
25%
50%
75%
100%
2006 2007 2008
Public finance Structured finance
International finance
Exhibit 13: Monoliners correlation to the housing markets
Indexed US Mortgage Default Rates and C
160
180
200
defaultrates
xed)
80
100
120
Dec-05 Jun-06 Dec-06 Jun-07
USMortgag
e
(Ind
US housing default rates
Source: Datastream, Oliver Wyman Analysis
AA
AAAMBAC
MBIA
Evolution of S&P ratings
DS spread time series
6,000
8,000
10,000
CDSspr
Dec-07 Jun-08 Dec-08 Jun-09 Dec-09
0
2,000
4,000
e
ad(bps)
MBAC CDS spread MBIA CDS spread
BB-BBBB+A-AA-
BBBBBA CC
Setting the scene
7/30/2019 Geneva Association Systemic Risk in Insurance Report March2010
30/129
20
Systemic Risk in InsuranceAn analysis of insurance and nancial stability
1.2.5. Conclusion
The insurance sector as a whole did not suffer the severe and widespread losses that the
bankingsector did.Where insurersencountereddifcultiesitwas predominantlythrough their
non-insuranceactivities.
1.3. Why did insurers fare better?
Insurersdonotrelyonwholesalemarketfundingforliquidity.Theyfundthemselvesthrough
premiums,withlong-termcapital tosupportrisk-taking positions.Their assetbases,whichare
substantialcomparedtotheircash-ows,mostlycomprisehighlymarketablesecurities.Whilst
insurersdoinvestinsomeilliquidorhigher-risksecurities,theirstrongtraditionofenterprise
risk management(which strengthened after the previous 2001-2003 equity crisis) and highly
regulatedbalancesheetsbothservetolimittheproportionofassetsatrisk.
Thus, during the crisis this natural long-liquidity position allowed insurers to be largely
unaffectedbytheliquiditycrunch,andinsurersinvestmentactivities,whichareguidedbytheneedtomatchliabilities,wouldhaveenabledthemtosurviveevenaprolongedandturbulent
marketdownturn.
Exhibit 14: Insurers cash-ows by source
5.1%
(3.7%)8.7%
12.0%
16.0%
hflo
w
Composition of Free Cash Flow yield for
generated by the life Value of In Force (VI
5.7%
0.7%
0.0%
4.0%
.
Freeca
nsurance
TotalLife
entreturn
essstrain
eneration
Generali
Investm
New
busin
Cashg
Source: Fat Tail Friday, Morgan Stanley Research esti
(5.7%)
(2.6%)(0.9%)2.3%
0.8%
uropean insurers the majority of cash is
) business unwind, FY09E
4.7%
10.4%13%
plusafter
dividends
ividends
low
Yield
inCapital
irements
Earnings
Other
Banking
agement
E
stimatedSur
FreeCash
Growth
req
TotalCash
Fundma
ates July 18 2008
Thedependabilityofinsurerscash-owsfromlargebackbooksofreliableinstitutionaland
retailcustomersallowedthemtocontinuetomakenetinvestmentsinsecuritiesatatimewhen
bankswereforcedtosellsecurities.
7/30/2019 Geneva Association Systemic Risk in Insurance Report March2010
31/129
21
Exhibit 15: Operating net cash-ows and investments by major European insurers
Net cash flows from operating activities
Top 5 European insurers
-50
0
50
100
150
2005 2006 2007 2008 H1 2009Cashflowfromoperations
(USDBN)
Allianz Aviva AXA ING Generali
Net investmentsTop 5 Insurers
-50
0
50
100
150
200
2005 2006 2007 2008 H1 2009Netinvestments(USDBN)
Source: Annual Reports, Oliver Wyman Analysis
1. Top 5 insurers by assets at 31.12.2006 in USD terms
Thestrengthofinsurerscash-ows,andtherelativestabilityoftheirbalancesheets,meant
thatinsurerswereabletomeettheirobligationsand,wherenecessary,toraisenewdebtcapital.
Exhibit 16: Estimated cash coverage of insurance companies
Source: Company data. European Life Goldman Sachs 2009
1: Normalised cash flow, pre-interest and tax, (excluding impairments) cover of interest expense
Estimated cash interest coverage of European life insurers1
0
1
2
3
4
5
6
7
L&G ING Prudential Generali Aegon AXA Std Life
Minimum to
meet obligations
Surplus
Setting the scene
7/30/2019 Geneva Association Systemic Risk in Insurance Report March2010
32/129
22
Systemic Risk in InsuranceAn analysis of insurance and nancial stability
Exhibit 17: Insurance debt issuance during the crisis
Source: Company data. European Life Goldman Sachs 2009
Debt issuance by European Life Insurers
2007-2009 (BN)
0
2
4
6
8
10
Q3 07 Q4 07 Q1 08 Q2 08 Q3 08 Q4 08 Q1 09 Q2 09
1.4. Conclusion
With a few notable exceptions, the insurance industry passed through the crisisrelatively
unscathed.Thoseinsurersthatencounteredthemostsignicantdifcultiessufferedthroughover-
exposuretonon-coreactivities.AIGisaclearexampleofthis.Itwasbroughtdownnotbyits
insurancebusinessesbutbyitscapitalmarketssubsidiary,AIGFP.Consequently,anyadditional
regulationforsystemicriskneedstotakeintoaccounttheactualactivitiesbywhichaninsurer
mightposesystemicrisk:thisisthefocusofourworkinthisreport.
7/30/2019 Geneva Association Systemic Risk in Insurance Report March2010
33/129
23
TheFSBandtheIMFrecentlydenedwhatsystemicriskmeanstothem.
G20nanceministersandcentralbankgovernorsendorsedthisdenition.
Exhibit 18: Denitions of systemic risk and systemic relevance
The risk of disruption to the flow of financial
services that is (i) caused by an impairment of
all or parts of the financial system; and (ii) has
the potential to have serious negative
consequences for the real economy
Fundamental to this definition is the notion that
systemic risk is associated with negative
externalities and/or market failure and that a
financial institutions failure or malfunction mayimpair the operation of the financial system
and/or the real economy
Source: FSB, IAIS
relevant institutions
Size: The volume of the volume of financial
services provided by the individual component
of the financial system
Interconnectedness: Linkages with other
components of the system
Substitutability: The extent to which other
componen s o e sys em can prov e e
same services in the event of a failure
Timing:Allow for the fact that systemic
insurance risk does not typically generate
immediate shock effects, but plays out over a
longer time horizon
The FSB denition is the most commonly referenced when systemic risk is discussed in
supervisoryandregulatoryforums.Wewillusethisdenitionasastartingpointforthediscussions
inthisdocumentalthoughthereisanactiveandongoingdebateaboutthedenitionofsystemic
risk.8
TheFSBhassetoutsomecriteriasize,interconnectednessandsubstitutabilitybywhichtherelevanceofparticularinstitutionstosystemicriskmaybeassessed.TheFSBhasalsospecied
secondarycriteria,qualiedascontributingfactors,thatpotentiallyincreasethevulnerabilityof
someinstitutions:namelycomplexity,9leverageandliquidityriskandlargemismatches.
8. Note: One company represented by The Geneva Association in this report disagrees with the FSB denition.
9. An institution can be qualied as complex if it
(a) operates diverse types of activities through numerous legal entities (e.g., simultaneously operating banking,
insurance and securities subsidiaries)
(b) operates across borders with centrally managed capital and liquidity (as opposed to simpler networks of national
subsidiaries)
(c) has exposures to new and complex products and markets that have not been sufciently tested.
Complexity per se would not be enough to guarantee a large systemic impact. However, countries may see complexity
as a source of vulnerabilityin particular if complexity is also associated with lack of transparency, difculties inunderstanding the exposures taken by the institution, and the potential magnication of information asymmetries in
the case of a systemic event (Source: FSB Guidance to Assess the Systemic Importance of Financial Institutions,
Markets and Instruments: Initial Considerations).
2. FSB and IAIS denition
of systemic risk
7/30/2019 Geneva Association Systemic Risk in Insurance Report March2010
34/129
24
Systemic Risk in InsuranceAn analysis of insurance and nancial stability
The IAIS have proposed adding time to these criteria, reecting the critical role that
timing (speed) plays in whether an event transmitted into the nancial system can be
absorbed by the system. We agree with this addition, and note its particular importance
toinsurance.
However,evenwhilstacceptingthebasiccriteriathattheFSBhasputforward,itisimportanttonotethattheimpactofthesecriteriaonnancialsystemriskcanbeverydifferentfordifferent
activities.Inthischapterweillustratethispointwithreferencetodifferentactivitiesthatnancial
institutions,betheybanksorinsurersmightcarryout.Ouraimisnottoarguethatthecriteria
arewrong,orthattheydonotapplytoinsurers,onlythattheyneedtobeappliedtoactivitiesand
nottoinstitutions.Applyingthecriteriablindlytowholeinstitutionsrisksaddinganadditional
regulatoryburdenonsystem-stabilisinginstitutionswhilstpotentiallymissingsomeinstitutions
carryingoutactivitieswhichdoposeasystemicrisk.Focusingonalistofinstitutionsisunlikelyto
detectormanagesystemicrisksmoreeffectively;instead,itislikelytoencourageriskmigration,
leadtounderestimationofsystemicriskandcreatedistortingmoralhazard.
2.1. SizeSizeisacrudemeasureofrisk.IfnoaccountistakenofEconomicCapital,thenassetsand
marketcapitalisationdonotmeasurerisk.Forexample,largeinsurersaretypicallywelldiversied
bothgeographicallyandacrosslinesofbusiness.Thisreectstheirbusinessmodel,whichcauses
themtobeexposedtoawiderangeofinsurance,market,businessandotherrisks.Becausethese
Exhibit 19: Economic capital by risk source
79%72%
40%
10%
23%
7%13%
4%
8%
36%
30%
41%
43%25%
17%
19%
45%
37%
7%9%
6%
28%
5%
1%
10% 11%
18%
32%
11%
8%7%
3% 2%
0%
20%
40%
60%
80%
100%
Commercial
Banking
Retail
Banking
Sales &
Trading
Asset
Management
Life
Insurance
P&C
Insurance
Reinsurance
Credit Market
Life Insurance P&C Insurance
Business Operational
Other
Source: 2006 ECAP Survey, IFRI CRO Summary, prepared by Oliver Wyman Companies Annual Reports
Breakdown of Economic Capital for European banks and insurers
Financial risks
Non-financial
risks
7/30/2019 Geneva Association Systemic Risk in Insurance Report March2010
35/129
25
risksareforthemostpartidiosyncraticandlargelyuncorrelated,thetotalrisktotheinstitutionis
lessthanthesumoftheindividualrisks.Thisdiversicationhelpsexplaintheresilienceofmost
largeinsurersduringthecrisis,anditisnotablethatatypicallargeinsurerismorediversied
thanatypicallargebank.Thussizeneedstodistinguishbetweenconcentrationofexposuresand
dilutionofriskthroughdiversication.
Size is of courseimportant, and it need not alwaysbe the case that a large bank is less
diversiedthanalargeinsurer.Buttherefore,thesignicanceofsizeforsystemicriskdepends
onthecompositionofaninstitutionsactivities,itsrespectivesizeanditsinterplaywithother
systemicriskfactors,suchasinterconnectedness.Itisnotsizeas suchthatpresentsdangerbut
undiversiedsize.Thisisthelessonofthemonoliners.Takingsizeasaninstitutionalvariable
ignoresthisproblem,andrisksthewrongjudgementsbeingmadeaboutsystemicrelevance.
2.2 Interconnectedness
Interconnectedness is a necessary condition for systemic relevance. Only if risk can be
transmittedcananinstitutionoritsactivitiespresentariskforthesystem.
Interactionsbetweeninstitutionswithinthenancialsectorareofdifferentkinds,including,
forexample,cross-ownerships,paymentsystemsinteractionsandexplicitrisktransferoperations
(reinsurance,derivatives).
But apparently similar types of interconnectednesscan have quite different effects onthe
nancialsystem.Bywayofexample,weconsidertwotypesofrisktransferactivity:reinsurance
transactionsbetweeninsurersandreinsurers,andCDStransactionsbetweenbanks.Bothinvolve
severalparties.However,whereasreinsurancetransactionsmitigatesystemicrisk(bysharingthe
existingrisksamongmanyplayersandallowingdiversicationofexposures),CDStransactions
canexacerbateit.And,whereasreinsurancetransactionsareasmallproportionoftheaggregate
insurancebalancesheet,pre-crisisCDStradingwassignicantinrelationtothetotalbanking
balancesheet.
Exhibit 20: Interconnectedness: insurance vs. banking
Comparison of inter-connectedness betw
Insurance
Low level of premiums ceded by insurers
USD BN
4,218
1,782
2,436
2.4% 5.6%
55% 28%
Ins uranc e Reins uranc e R et ro ces sio ns
15960
Life insurance Non-life insurance
Source: Swiss Re sigma, IAIS Global Reinsurance Market RepoNote: It should be noted that the insurance values are a flow wh
.
en insurance and banking operations
Banking
CDS notional exposures represent ~41 % of
worldwide banking assets, USD BN
95% of single name CDS
outstandings transacted
between banks
19,184
15,347
36,046
88,400
515
rt 2009, BIS, Oliver Wyman analysesreas the CDS is a stock Figure. However, the comparison of the
1,000
largest
banks
assets
TotalCDS
outstandings
Reporting
dealers
Banks
and
securities
firm
s
Others
1,000largest
banksassets
TotalCDS
outstandings
Reporting
dealers
Banksand
securitiesfirms
Others
FSB and IAIS denition of systemic risk
7/30/2019 Geneva Association Systemic Risk in Insurance Report March2010
36/129
26
Systemic Risk in InsuranceAn analysis of insurance and nancial stability
Reinsuranceeffectivelysharesariskwhosesizeislimitedtotheunderlyinginsurableinterest,
betweenseveralparties,andallowsdiversicationofexposures.Moralhazardismitigatedbecause
theprimaryinsurerremainsexposedtoaportionofanylossesinareinsurancetransaction.This
linksthedestinyofthecedinginsurerandthereinsurerinanytransaction.
Inthecontextofnaturalcatastrophesthestructureoftheinsuranceindustrygivesittwolinesofdefence.Reinsurerstakeonpeakrisksandactasarstlineofdefenceforextremeevents
(e.g.catastrophes),providinga rewallthat protectsprimaryinsurersfrommassivelossesand
potentialfailure.Intheeventofpeaklossescausingthefailureofreinsurers,thecapitalofprimary
insurersprovidesasecondlineofdefence.Itisintheorypossibleforaneventtobesohugethatit
overwhelmstheentireinsuranceindustry,butthiswouldthenbeanationalorglobalcatastrophe
ofunimaginablescale(forexample,HurricaneKatrinawaswellwithintheindustryscapacity).In
thiscircumstancetheeventwouldhavecausedtheloss,nottheactivitiesofinsurancecompanies.
InterbankCDStransactions,bycontrast,candisconnecttheriskfromholdingtheunderlying
exposure. When Lehman Brothers defaulted in September 2008, it had USD 155 billion in
outstandingdebt,yetUSD400billionofCDSreferencingthisdebthadbeensold.Itistherefore
possiblefortheCDSlossesarisingfromaneventtoimpactmanymoremarketparticipantsinasubstantialwaythantheeventitself;themarketinterconnectednesscanleadtoacontagionofthe
wholesystem.
Thiscontagionisaggravatedbytheproblemofopacity.Thecomplexityofmanyderivatives
andthefactthatmanyaretradedover-the-counterratherthanthroughexchangesmakeitdifcult
toassesstheriskpositionofinterbankcounterparties.Thisopacitywasanimportantfactorforthe
lossofcondenceatthepeakofthecrisisinlate2008.Inter-insurancermtransactionsenjoya
muchhigherdegreeoftransparency,beingdominatedbyasmallnumberofstandardmechanisms
thatarewellunderstoodbytheplayersinvolvedandthewidermarket.
Again,wedonotarguethatinterconnectednessisnotarelevantcriterionforassessingthe
systemic risk ofan insurer. Rather, we pointout that the interconnectednesscan bea highlysignicantfactorforsomeactivities,whereasforothersinterconnectednessisoflittlerelevance.
2.3. Substitutability
Thesubstitutabilityofan institutionforanancialservicemustbeassessedbyconsidering
twoquestions:
Doestheinstitutionhaveanytechnicalspecicitiesorplaysuchauniqueroleinamarket
thatitwouldbedifculttosubstituteanequivalentactorintheshort-termiftheinstitution
weretodisappear?
Isthecapacitythattheinstitutiondeploystoitsmarketsolargeoruniquethatotherscould
notstepinwithcapacitysufcienttoenablethemarkettoclear?
By these tests, insurance activity is substitutable, and therefore not systemically relevant
by this criterion. Firstly, no insurer has a monopoly in any material line of insurance,
nor does one institution play a central market role such as clearing or acting as a
securitiesexchange.
Secondly,insurancecapacityissubstitutable.Insurersderivetheircapacitytowritebusiness
fromanumberofsources,includingexternalcapital.However,wecanseereinsurersastheultimate
providersofcapacityinthesystem:solongasreinsurersareofferingcapacity,insurerswillwrite
business.Butcapacityinthereinsurancemarketiseasilysubstituted.Afteranaturalcatastrophe,
reinsurancecapacitydeclinesandreinsurance prices typicallyincreaseforseveral years. This
cycleattractsnewcapitalowingtobothexistingreinsurersandnewentrants(includingcapitalmarketsthroughside-carsandcatbondsecuritisations)while(expected)higherprotstopup
7/30/2019 Geneva Association Systemic Risk in Insurance Report March2010
37/129
27
capital reserves. Because these prots are not risk-free there is noarbitrage opportunity, and
thereforethecycleislong-lastingenoughtorestoresufcientcapacitytothesystem.
Exhibit 21: Worldwide reinsurance capital inows 1990-2008
0
2
4
6
8
10
12
14
16
18
20
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008
Y ear
0
50
100
150
200
250
300
350
400
450
Non-Bermuda (Equity+IPO) Bermuda (Equity+IPO) Cat Bonds Side cars Guy Carpenter RoL Index (RHS)
Source: Thomson, Guy Carpenter, AON Benfield, Dealogic, Oliver Wyman analysis
No data on inflows1990-1993
Hurricane Andrew 9/11 Hurricane KatrinaNew capital flows into nat cat reinsurance industry and nat cat reinsurance rates
USD
BN
RateonLineIndex(1990=100)
Of course, the recapitalisation of the industry after a major catastrophe does not happen
overnight.Howeverthereisalsonoovernightsubstitutabilityproblem.First,thetimingcriterion
raisedbytheIAISisrelevanthere.Whileanimmediatesubstitutewouldberequiredforseveral
bankingactivities,thisisnotthecaseforallinsuranceservices(andaswenoteinsection2.6,the
processforwinding-upaninsureralsomeansthatthislossofcoveragewouldnotbeimmediate).Second,theinsuranceandreinsurancemarketsarehighlydiversiedbytheusualmeasures
withineachcountryandworldwideforreinsurance,soonlyaneventwhichdestroyedtheentire
industrycouldmakeitimpossibletopurchasenecessaryinsurancecover.
Exhibit 22: Concentration of European primary insurance markets
Herfindahl-Index of primary insurance (lif
insurance markets1, 2007 (2008 for Spain,
0.10
0.12
0.14
0.16
0.18
hl-Index
0.00
0.02
0.04
0.06
0.08
Sp ai n UK Ger many Net her la
Herfind
Source: Oliver Wyman European Insurance Database 2008
1 Index based on gross premiums written by European insurers0.1Index
7/30/2019 Geneva Association Systemic Risk in Insurance Report March2010
38/129
28
Systemic Risk in InsuranceAn analysis of insurance and nancial stability
Exhibit 23: Concentration of the global and U.S. reinsurance markets
0%
20%
40%
60%
80%
100%
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
Number of considered reinsurers
Global US
Cumulative premiums by the largest reinsurers
Ranked in descending order by gross written premium, 2008
Source: AM Best, IAIS Global Reinsurance Market Report 2009, OW Press Research
1. Based on the Top 35 global reinsurers (gross written premium 2008: USD 156 BN) and the total gross written reinsurance premium according to the IAIS Global
Reinsurance
Market Report 2009 (2008: USD 159 BN) which is only considering reinsurers writing reinsurance in excess of USD 800 MM
2. Based on the Top 25 U.S. reinsurers (gross written premium 2008: USD 37.2 BN of total U.S. gross written premium of USD 37.7 BN)
Note: The Herfindahl index is a measure of the size of firms in relation to the industry and an indicator of the amount of competition among them; defined as the
sum of the squares of the market shares of the 50 largest firms
Cumulativegrosswritten
reinsurancepremiums
Duetotheorganisationandstructureoftheinsuranceandreinsurancemarkets,substitutability
is therefore not as relevant for insurance companies as it is for banks where exchanges and
paymentsystemsarecritical.
2.4. Timing
Insuranceclaims operate muchmoreslowly thanthe margin call,collateraland depositor
claimsonbanks.Forexample,lessthanhalfoftheclaimsontheWorldTradeCentreweresettled
twoyearsaftertheevent.Thisisadirectconsequenceofthenatureoflargeinsuranceclaims:they
relyinmanycasesonmultiplepolicies,courtjudgements,andindividualclaimants.
Exhibit 24: Timing of World Trade Centre claims payments
0%
20%
40%
60%
80%
100%
Q4 2001 Q4 2002 Q4 2003 Q4 2004 Q4 2005 Q4 2006 Q4 2007
Quarters after 11 September 2001
%o
ftotalnetclaimpaid
Cumulative proportion of claims made over time
Timing of World Trade Centre Insurance Claims
Source: Reinsurance Association of America, Catastrophe Loss Development Study
7/30/2019 Geneva Association Systemic Risk in Insurance Report March2010
39/129
29
The failure ofa bank and the consequent closure ofthe wholesale funding markets could
triggerthecollapseofthebankingsystemveryquickly.Bycontrast,thewind-upofaninsureris
likelytobeamoreorderlyprocess,asdiscussedinChapter2.6.Theslowpaceoffailureincreases
substitutabilitybyprovidingthetimerequiredtorebuildindustrycapitalandcapacity.Wecan
thusarguethatinsurancermfailurespresentlesssystemicriskthanbankfailures.Howeverin
doingso,wenotethatthedifcultiescausedbyAIGsFinancialProductsdivisionpresenteda
hugeandimmediatesystemicthreatinthejudgementofU.S.policy-makers:thisagainpointsto
theneedtoconsideractivities,ratherthaninstitutions,inconsideringsystemicrelevance.
2.5. Contributing factors to the assessment
of systemic importance
Each of the three further contributing factors (leverage, liquidity/mismatches and
complexity),independently,wouldnotcreatesystemicrisk,butshouldbeconsideredalongside
themaincriteria.
Wecantakeleverageandliquiditytogether,asinbankingactivitiesleverage(forexampletheuseofwholesalefundingtoincreasetheinstitutionslendingcapacityrelativetoitsdepositbase)
andliquiditymismatchesbothservetoexposetheinstitutiontotheriskofthewholesalemarket
drying-upforaperiodoftime.
But for typical insurancebusiness, the concept ofleverage is different from banking,and
therefore the nature of liquidity risk is different. Insurance activity is self-funding through
premiuminow,with long-termsourcesofcapitalusedtosupporttherisks accepted.Inother
words,leverageisnotpartofthebusinessmodelofinsuranceandinsurancecompaniesdonot
requireleveragetofunction.
Insurers do have to maintain appropriate liquidity, as they do have payments to make to
policy-holders.Many ofthese paymentsare plannedclaims andbenet payments,but insomeinstancespolicy-holdershavetheabilitytoacceleratepayments,throughpolicysurrenders.These
paymentsare funded,rst, throughpremiuminows,andnext throughsales ofsecuritiesheld.
Insurersarenotreliantonwholesalefundingtomeetpolicy-holderredemptionsorforanyother
coreinsuranceactivity.
Consequentlywhilstliquidityisarelevantissueforallnancialinstitutions,thenatureofthe
riskisverydifferentfordifferentactivities,asarethemetricsneededtomeasurethisriskandthe
actionsrequiredtomitigateit.
Finally,theFSBraisescomplexityasacontributingfactor.Complexitybyitselfisnotarelevant
issue,exceptinsofarasintra-grouptransactions,bypreventinganorderlywind-up,exacerbate
thesystemicriskfromthegroupsactivities.Forbanks,thisisparticularlythecasewherethe
intra-grouptransactionsarealsointer-country.Butforaninsurancegroupnotcarryingoutbankingactivities,intra-grouptransactionsareusedforcapitalmanagementefciency,shouldnotinterfere
withorderlywind-up(thisisdiscussedinmoredetailbelow),andshouldnotmaketheregulationof
theoverallentitymorecomplexaseachinsurancebalancesheetisaregulatedentity.Consequently,
forcertainactivities,complexityisasignicantcontributingfactor;forotheractivitiesitisof
norelevance.
2.6. Wind-up and run-off: insurance industry experience
Whileregulationattemptstolimitthepossibilityofinsolvencies,andinparticularinsolvencies
causedbyimprudentmanagementdecisions,companyfailuresassuchshouldnotbeconsidered
assomethingtobepreventedatanycost.Thedisappearanceofoldandappearanceofnewmarketparticipantsisanessentialelementofmarketeconomies.
FSB and IAIS denition of systemic risk
7/30/2019 Geneva Association Systemic Risk i