Limpiggycase

Chairperson'sCorner—RiskManagementSection'sTerribleTwosbyDavidN.
Ingram2RiskManagementSectionGrowingbyMikeBoa_4DynamicRiskModelingbyJamesE.
Rech_5ChiefRiskOfficerForum—PrinciplesforRegulatoryAdmissibilityofInternalModelsbyJohnHeleandHenkvanBroekhoven_____7InternalControls—TheCOSOWaybyDorothyL.
Andrews10Standard&Poor'sEnterpriseRiskManagementEvaluationofInsurersbyDavidN.
Ingram_14ERM≠EC2bySimSegal18ThirteenWaystoKillaCompanybyJenniferBowen21RiskManagementInvestorSurveybyMaryEllenLuning28PolicyholderBehaviorintheTail:VariableAnnuityGuaranteedBenefitsSurveyResultsbyJamesReiskytl_32ContinuingEducation—2006ActivitiesbyHubertMueller_34March2006,IssueNo.
7PublishedinSchaumburg,Ill.
bytheSocietyofActuariesRiskTableofContentsManagementChairperson'sCornerTheRiskManagementSectionrecentlyturnedtwoyearsold.
Asmanyofyouwhoareparentsknow,atwo-year-oldcanbequiteahandful.
Two-year-oldsarejustlearningthattheycanbemuchmoreindepend-ent.
Theyhavemoreskillsthanone-year-oldsandmuchmorepotentialtocausetrouble.
Ourgrouphasbeengrowinglikeasmallchild.
Lastcount,ourmembershipwasaround3,000withmorethan500fromtheCAS(seethesec-tionmembershipstatisticsatendofthisarticle).
Ourparents,theSOAandtheCAShavebothrecentlyexpandedtheircommit-menttoriskmanagementandtotheimportanceofourmission.
Ourresponse,likeatwo-year-old,hasbeentorunoffinseveraldirectionsatonce.
However,unlikethetypicaltwo-year-old,wehavetheadvantageofhav-ingmanylegsandmany,manyhands.
Andhope-fully,manyheadsaswell.
Thesectioncouncilhasexpandedfromninetotwelvemembersandwehavealsoconsolidatedourcommitteestructure.
Wewillbeoperatingwithonlysixcommittees.
Eachcommitteewillbeempoweredandexpectedtoactwithinthebroaddirectionsthathavebeensetdownbythecouncil.
Wewillbemakingaspecialefforttobreakoutofthestandardforsectioncouncils,wherethecouncilmembersactassurrogatevol-unteersfortheentiresectionmembership.
Committeeswillbemadeupofsectioncouncilmembersworkingwithnon-councilmembers.
YOUcanhelp.
Youdonotneedtobeelectedtothecounciltomakeacontribution.
Thenewslettercommittee,headedbyKenSengTanandRonHarasym,willbecontinuingtoproducethisexcellentpublication.
Theyarelookingforhelpfrompeoplewhoarewillingtowritearticlesandfromfolkswhowouldbeabletohelpdeveloparticles.
HubertMueller,LarryRubinandTonyDardisareleadingtheeffortsfortheContinuingEducationCommittee.
TheywillbedevelopingsessionsforSOAmeetingsaswellasforstand-alonecontinuingeducationprogramslikelastyear'sverysuccessfulwebcast.
TheRiskManagementSectionwillbeagainco-sponsoringtheERMSymposiuminChicagothisApril.
ValentinaIsakinaisleadingtheOrganizingCommitteeandcouncilmembersHubertMueller,FrankSabatiniandKevinDicksonhavebeenworkingonthecommittee.
Youshouldhavealreadyheardaboutthepro-gramfor2006.
Wewillbestartingworkonthe2007ERMSymposiumthissummerandwillbelookingfornewcommitteemembers.
FredTavanandRonHarasymareheadingtheResearchCommittee.
Theyhaveaworkingcommitteethatisreviewingresearchproposalsandfindingstafffortheprojectoversightgroups.
TheyneedseveraladditionalpeopletoworkwiththeResearchCommitteeandonover-sightcommittees.
Therestofthesectionactivitiesfallintothecat-egoryofSpecialProjects.
Therearequitearangeoftheseincluding:RiskIndex,OperationalRisk,RiskManagementTerms,RiskPreference,ERM,ExtremeValueModeling,PolicyholderBehavior,ActuarialValueProposition,BestPracticesandLocalNetworking.
Inaddition,whilewegenerallyfindtwo-year-oldsdoing"side-by-side"play,IamhopefulthattheSOAandCASmembersofthesectionwillveryquicklymoveintoactualcollabora-tion.
Andalongthewaytocollaboration,wequicklyneedtolearnfromeachotherthetech-niques,skillsandexperiencesthatwehavede-velopedinourseparateactuarialpractices.
Todothat,wearetryingtoencouragefolkstosharetheirriskmeasurementandmanagementexperiencesinhopesthatsharingwillhelpRiskManagementQMarch2006RiskManagementSection'sTerribleTwosbyDavidN.
IngramQPage2DavidN.
Ingram,FSA,isdirectorofenterpriseriskmanagementwithStandard&Poor'sinNewYork,NY.
Hecanbereachedatdavid_ingram@standardandpoors.
com.
Chairperson'sCorneridentifytheseareasforlearningandcollabora-tion.
Ifyouhaveanythoughts,ideasorexperi-encestosharethatmightfurtherthislearningandcollaboration,pleasecontactmeoranymemberofthesectioncouncil.
Finally,beforeyoupassonthecontentofthispublication,Iwanttomentiononemorethingaboutvolunteering.
Fiveyearsago,whenthefirstRiskManagementTaskForceCommitteeswereformed,fewofthevolunteershadmuchriskman-agementexperience.
Today,ifyoudidarollcallofthoseinitialvolunteers,themajorityofthemnowhavefull-timeriskmanagementjobs.
Thisisstillanewandgrowingarea.
Theriskmanagementjobcandidateswithvolunteerexperiencewillhavetwoadvantagesoverthosewithout.
Theywillhavelearnedsomevaluablethingsaboutsomespecificareasofriskmanagementthroughtheirvolunteerworkandtheywillhavethatvol-unteerworkontheirresume.
Ifyoumightbein-terested,contactmeoranysectioncouncilperson.
Wecanfillyouinonthesesuccessstoriesandoncurrentvolunteeropportunities.
!
March2006QRiskManagementPage3Q".
.
.
Riskmanagementjobcandidateswithvolunteerexperiencewillhavetwoadvantagesoverthosewithout.
"2006SOASectionMembershipStatistics(asofJanuary12,2006)SectionTotalSectionMembersEducation693FinancialReporting3,758ActuaryofFuture899Futurism676Health3,486International1,747Investment4,778LTCI1,649ManagementandPersonalDevelopment1,302MarketingandDistribution2,108Pension3,950ProductDevelopment3,869Reinsurance2,392RiskManagement(excludeCASmembers)2,306SmallerConsultingFirm585SmallerInsuranceCompany685Taxation575Technology1,503GrandTotals36,961RiskManagementSectionGrowingbyMikeBoaYear2005wasayearofremarkablegrowthforthejointCAS/SOARiskManagementSection,ascasualtyac-tuariesjoinedinforce.
Morethan500membersoftheCAShavejoinedthesection,bringingthetotalsectionmembershiptoabout3,000.
Therapidgrowthwillhelpthesectiontoachieveitsprimary2006objectivesofexpandingERMeduca-tionalopportunities,fos-teringriskmanagementresearch,andsupportingtheinitiativesinpromot-ingactuariesasriskman-agers.
Theworkofthesectionisaccomplishedbyitsteams,andtheadditionalCASmemberswillhelpgettheworkdonewhileprovidingacasualtyperspective.
Theteamsinclude:MembershipValueTeamCommunicationsandPublicationsTeamNewsletterTeamContinuingEducationTeamBasicEducationTeamRiskManagementResearchTeamMarketplaceRelevanceTeamProfessionalCommunityTeamTheResearchTeamprovidessupportanddi-rectiontoseveralspecificinitiative-relatedresearchcommittees,twoofwhichareledbyCASmembers.
MarkVerheyenischairingtheOperationalRiskManagementCommittee,whileMichaelBelfattiischairingtheStandardRiskManagementTermsCommittee.
Inadditiontothechancetobeonthefrontlineinadvancingactuariesintheriskmanagementarena,membersofthesectionenjoyotherbene-fits,suchasreceivinginvitationstosectionnet-workingevents,liketheoneheldduringthe2005CASAnnualMeetinginBaltimore.
TheCASVicePresidentofRiskIntegrationandERM,JohnJ.
Kollar,ledadiscussionover(real)breakfastwithabout25otherearly-risersaboutthevaluethatactuariescanbringtotheriskmanagementprofession.
SectionmembersalsoreceiveRiskManagement,thesectionnewsletter,ande-mailcommunicationsannouncingupcomingre-search,projects,continuingeducationeventsandotheractivities.
Additionalinformationaboutthesection,in-cludinganapplicationtojoin,canbefoundontheSOAWebsiteathttp://www.
soa.
org/ccm/content/categoryID=342001.
!
RiskManagementSectionGrowsRiskManagementQMarch2006QPage4MikeBoaisdirectorofcommunicationsandresearchfortheCasualtyActuarialSocietyinArlington,Va.
Hecanbereachedatmboa@casact.
org.
Theyearwas1995andtheCasualtyActuarialSociety'sDynamicFinancialAnalysisCommittee(DFAC)hadpub-lisheditsfirstDFAHandbook.
Thehandbookprovidedacookbookapproachoutliningnu-merousconsiderationsinthedevelopmentofadynamicfinancialanalysismodel.
Themerefactthatthehandbookconsistedofnearly85pagesofconsiderationsandwasreflectiveoftheperceivednotionthatDFAmodelingwasa"dauntingtask;"fewfirmswerewillingtoex-pendthetimeandcostsindevelopingandmain-tainingsuchextensiveDFAmodels.
TheDFAworldcontinuedtoevolve.
TheDFACcreatednew"chapters"fortheDFAHandbookinordertoexpandthescopeoftheoriginalhandbook.
Morerecently,theDFACchangeditsnametotheDynamicRiskModelingCommittee(DRMC).
Thenamechangewasindicativeofthemorecurrentthoughtprocessthatadvocatestheusefulnessofdynamicmodelingtechniquesforprojectswithvariousscalesofscopeandpur-pose.
Firmswerewillingtofundthedevelop-mentofmorespecificdynamicriskmodelsaddressingreinsurancestrategies,stochasticreserving,aggregatelossdistributions,catas-trophemodeling,predictivemodeling,etc.
Thepowerofdynamicmodelingtechniquesinad-dressingspecificmanagementconcernsisnowevident.
TenyearshavepassedsincethatoriginalDFAHandbook.
Lastyear,theDRMCsetupaWorkingParty(WP)torewriteandupdatetherenamedDRMhandbook.
Attestingtoawidein-terestinthehandbook,themembershipoftheWPwasfairlyevenlydistributedamongcreden-tialedactuaries,studentsandnon-actuaries.
Geographically,whilethemajorityofmemberswereintheUnitedStates,wehaveseveralmem-berslocatedoutsidetheU.
S.
boarders.
Theconceptofthisnewhandbookistobuilduponthesubstantialcontributionsfromprevi-ousauthors.
Assuch,muchofthecurrentun-dertakingisbasedonreorganizing,updatingandeditingthecurrentDFAHandbookincon-junctionwithotherexistingDRMCpublishedarticles.
Therevised(andrenamed)"DynamicRiskModelingHandbook"isintendedtoprovideabasicunderstandingofandpracti-calguidelinesforthedevel-opmentandimplementationofdynamicriskmodelscom-montothepropertyandcasu-altyinsuranceindustry.
Inaddition,itishopedthattheDRMHandbookwillbecomeabasicreferencesourcefortheeducationalneedsoffu-turemodelersandthepracti-calday-to-dayapplicationneedsofexperiencedpractitioners.
Thescopeofthisrewriteisto:Restructureandedittheexistingcorechaptersofthecurrent"DynamicFinancialAnalysisHandbook"forconsistency.
Addnewchaptersincluding"Introduction,""AssetModeling,""CoherentRiskMeasures",and"PresentationsofDRMResults.
"Introducepracticalexampleswithinthesecore(andnew)chapterstoillustratetheconceptsusingthePublicAccessDFAModelwherethatwouldbehelpfultoillustrateaconcept.
AddabibliographyrelatinginvestmentconceptsonthesyllabustoDynamicRiskModeling(thesyllabusnowhassectionsonPart8thataddressinterestratemodelsforexample).
Addatleastfive(5)casestudieschoosingfromthesubjectslistedbelow(asAppendices)–InterestRateModelsInvestmentPolicyStatements(IPS)andtheeconomicdependenciesbetweenAssetClassesandLiabilitiesFinancialRiskManagement(hedgingandtheuseoffinancialderivatives,options,swapsandforwards)SurplusAllocationReinsuranceAlternativegrowthstrategiesbylineofbusinessDynamicRiskModelingPage5QDynamicRiskModelingbyJamesE.
Rechcontinuedonpage6QJamesRech,ASA,ACAS,MAAA,isaconsultingactuaryandvicepresidentwithGPW&Associates,Inc.
inPhoenix,Ariz.
Hecanbereachedatjrech@gpwa.
com.
March2006QRiskManagement""Itisanticipatedthattheexamplesandcasestudiesusedinthehandbookwillbebasedonthepublicaccessmodelandwillbecoordinatedtothegreatestextentpossible.
.
.
AlternativeinvestmentstrategiesIntegratesome(orall)ofthenewcasestudieswiththeWorkProductsoftheWorkingPartyonExecutiveLevelDecisionMakingusingDRMandtheWorkingPartyonthePublicAccessDFAModel.
Setupguidelinesforfutureenhancements,correctionsoradditionstothenewhandbook.
Itisanticipatedthattheexamplesandcasestud-iesusedinthehandbookwillbebasedonthepublicaccessmodelandwillbecoordinatedtothegreatestextentpossiblewiththedevelop-mentoftheCasualtyActuarialSociety'snewRiskModelingWorkshop.
Documentation,en-hancementsand"conversion"to"opensource"forthepublicaccessmodelisthesubjectofan-otherWPbeingsponsoredbytheDRMC,soco-ordinationofthehandbookwiththisWPisalsopartoftheprocess.
ThegeneralguidelinewasfortheDRMHandbookWorkingPartytocompleteitsassign-mentwithinoneyear.
Unfortunately,theoriginaltimeframehasturnedouttobeoptimistic.
Whilethemajorityofthehandbookisnowinitseditingstage,westillrequireagreateffortononeoftheprimarycasualtychapters,Price/ReservingModels.
Butfirst,thebasicsofthehandbook:ThestructureoftheDRMHandbookwillin-cludeninechaptersandthreeappendices.
Chapters1.
Introduction2.
OverviewoftheDRMProcess3.
Strategies4.
Scenarios5.
AssetModeling6.
Price/ReserveModeling7.
PerformanceandRiskMeasures8.
CoherentMeasuresofRisk9.
PresentationsofDRMResultsAppendices1.
Bibliography2.
DRMChecklistofConsiderations3.
GlossaryofTermsBecausethedistributionofthehandbookwillbedigital,itisintendedthatitwillberegularlyup-datedtohighlighttechnicaladvancesandinno-vationsinriskmodeling.
WhilethemajorityoftheDRMHandbookispro-gressingnicely,therearestillacoupleofareasinwhichthehandbookrequiresadditionalefforts.
ThefirstareaisinthedevelopmentofChapter6,Price/ReserveModeling.
Thesecondareaisinthedevelopmentofthefivecasestudiesforthehandbook.
WearestilllookingforwritersforthesecriticaladditionstotheDRMHandbook.
Ifyouareinterested,pleasecontactRunYanatrun.
yan@mercer.
comormyselfatjresch@gpwa.
com.
!
QPage6DynamicRiskModelingRiskManagementQMarch2006DynamicRiskModelingHandbookQcontinuedfrompage5Page7QCROForumTheChiefRiskOfficer(CRO)Forum,comprisingriskofficersofthemajorEuropeaninsurancecompaniesandfi-nancialconglomerates,wasformedtoaddresskeyriskissues.
Itisatechnicalgroupfocusedondevelopingandpromotingindustrybestprac-ticesinriskmanagement.
Themembershipismadeupofthefollowingcompanies:AEGONNV,AllianzAG,AvivaPLC,AXAGroup,Converium,Fortis,Generali,INGGroup,MunichRE,PrudentialPLC,SwissRe,WinterthurandZurichFinancialServices.
IntheNovember2005,IssueNo.
6ofRiskManagement,theJune2005CROForumstudy,"AFrameworkforIncorporatingDiversificationintheSolvencyAssessmentofInsurers"waspresented.
ThisarticlesummarizestheotherCROForumstudypublishedatthattime.
IntroductionTheCROForumhasundertakenastudytobenchmarkinternalmodels,soastodiscusstheadmissibilityofthesemodelsforregulatorypur-posesinthecontextofSolvencyII.
Ininternalmodelsthetrueriskprofileandsolvencyposi-tionofacompanyisreflectedandthereforetheuseofinternalmodelsprovidesarealincentiveforimprovedriskmeasurementandriskman-agement.
Thestudypresentstheresultsofthebenchmarkingofinternalmodelsandalsopre-sentsaproposedsetofprinciplesthatcouldbeusedbyregulatorsforvalidatingandadmittinginternalmodelsforregulatorycapitalpurposes.
InventoryofRiskMeasurementFrameworksUsedbyCROForumMembersTheCROForumestablishedabenchmarkingteamconsistingofDamirFilipovicandDanielRostoftheUniversityofMunich,withMercerOliverWymanforsupplementarysupport.
Ade-tailedquestionnaire,setupbythebenchmark-ingteam,wascompletedbyallparticipantsandthreeregulatoryagencies.
Thesurveyoutcomeswerediscussedwiththeriskmanagementde-partmentsofall13-membercompanies.
TherewerealsoresponsesfromBPV(Switzerland),DNB(Netherlands)andBaFin(Germany).
OverviewoftheResultsoftheBenchmarkingStudyThemostimportantandinterestingconclusionisthattheapproachesusedbytheparticipantsinthebenchmarkstudyarehighlysimilar.
Insomecasestherearedifferences,oftendrivenbydifferencesinthetypeofbusiness.
Withrespecttotheframeworkdefinition,mostoftheparticipants(69percent)useaVARap-proachwithaone-yeartimehorizon,andmorethan75percentfollowaneconomicapproach.
Thevastmajority(85percent)onlyuse,atmax-imum,oneyearofnewbusiness.
Onlyafew(15percent)usetheIAA(AGlobalFrameworkforInsurerSolvencyAssessment)advisedTailVaRastheriskmeasureduetothecomplexityofthismeasure.
Withrespecttotheconfidencelevels,allparticipantsuseaconfidencelevelabove99.
5percent(99.
6percentto99.
99percent).
Forsolvencypurposesaregulatoryconsensusappearstobeconvergingto99.
5percent.
Overall,themodelingmethodologiesusedareinlinewiththoserecommendedintheIAASolvencyFrameworkpaper.
TheimportantissueisthatALLriskshouldbemeasuredinaconsistentway.
Allparticipantsmodelandmeasuremarketriskandcreditrisk.
Mostpar-ticipants(morethan75percent)alsomodelcreditriskforreinsurers.
Asexpectedthequan-titativemeasurementofoperationalriskisstillinitsinfancy.
Theinterestingconclusionsontheframeworkimplementationarethatmost(69percent)haveadetaileddocumentationsystem,butthatonlyabouthalf(54percent)haveaformalizedsign-offprocedure.
Alsoitseemsthat,al-thoughcriticalforgainingseniormanagementcommitment,linkstomanagementcompensa-tionarestillintheirinfancy.
ChiefRiskOfficerForumPrinciplesforRegulatoryAdmissibilityofInternalModelsbyJohnC.
R.
HeleandHenkvanBroekhovencontinuedonpage8QJohnC.
R.
HeleisdeputychieffinancialofficerofINGGroupandgeneralmanagerofING'sCorporateInsuranceRiskManagementDepartmentinAmsterdam,TheNetherlands.
HecanbereachedatJohn.
Hele@ing.
com.
HenkvanBroekhovenisariskmanageratINGGroups'CorporateInsuranceRiskManagementDepartmentinAmsterdam,TheNetherlands.
HecanbereachedatHenk.
v.
Broekhoven@ing.
com.
March2006QRiskManagement""Peoplehavetroubleincorporatingaprioriprobabilities,whichcanbethemostimportantfactorwithqualitativeinformationinestimatingprobabilities.
OverviewoftheResultingPrinciplesDefinedbytheChiefRiskOfficerForumPrincipleRiskModelingFrameworkInternalmodelsneedtobebasedontheadversemovementintheEconomicValueofthedifferencebetweentheAssetsandLiabilities,calibratedtoanannualized99.
5percentprobabilityofsolvency.
Modelingapproachesbasedonlongertimehorizonsoralternativeriskmeasures(e.
g.
,TailVaR)arepermissible,providedthecalibrationapproachusedcanbeshowntobeconsistentwithanannualized0.
5percentprobabilityofeconomicinsolvencyOneyear'snewbusinessshouldbeexplicitlymodeling,basedonassumptionsthatareconsistentwithbusinessplans,wherethishasamaterialimpactontheriskprofileofthegroupAssetsthatarenotlikelytobeavailableintheeventofinsolvency(forexample,profitsfromfuturenewbusiness,thecomponentofdeferredtaxassetsarisingfromlossescarriedforward),shouldnotbeincludedasavailablecapitalintheinternalmodelBestestimateliabilitycashflowsshouldbediscountedatswaprates,astheyaretypicallythemostliquid,completeandreliablerisk-freeratesavailable.
Thisismoreconservativethanusingatrulyeconomicdiscountratethatwouldincludeanallowanceforthecreditspreadoftheinsureritself(orofthecounterpartytowhomtheliabilitieswouldbetransferredintheeventofinsolvency)PrincipleModelingMarketRiskAllsourcesofmarketriskneedtobemodelingprobabilisticallywithinter-factordependenciesexplicitlymodeling.
Choiceofmodelingapproach(simulation-basedoranalytical)andgranularityofmodelingneedstobeproportionatetotherisks/businessesbeingmodeling.
Forexample:Interestrates—CashflowmatchingtakenaccountofbymodelingofthewholeyieldcurveFXmismatchrisk—Currencymismatchesbetweenassetsandliabilities/supportingcapitalexplicitlymodelingEquityrisk—Equityriskmodelingbasedonanalysisoftherelevantmarketindexwhereconcentrationinindividualsectors/individualstocksdiffersfromtheindex,suchconcentrationsshouldbeexplicitlymodelingRealestaterisk—Realestateriskmodelingbasedonanalysisoftherelevantpropertymarketindex,orreasonableproxiesifsuchanindexisunavailable—whereconcentra-tioninindividualsectors/individualstocksdiffersfromtheindex,suchconcentrationsshouldbeexplicitlymodelingDerivatives/marketriskmitigation—Explicitmodelingthroughsimulation/scenarios,withcounter-partycreditriskalsobeingmeasured.
Embeddedoptionsandguaranteesexplicitlymodelingthroughsimulationmodeling:Managementactions(e.
g.
,bonusratesonparticipatingbusiness,dynamicassetallocationpolicies),wherematerial,shouldbeexplicitlyandreal-isticallymodeling,withmodelingmanagementactionscodifiedaspolicyanddisclosedtothesupervisorPolicyholderbehavior,wherematerial,shouldbeexplicitlyanddynamicallymodeling,withkeyassumptions(whichcouldbeeitherexpert-opinionbasedorempiricallybased)beingdisclosedtothesupervisorParameterizationofvolatilityanddepend-enciesbetweenmarketriskfactorsshouldbederivedfromanappropriateandreliablytimeseriesofmarketdata,andshouldbeestimatedaccountingfortaildependencies(e.
g.
,understressedconditions)PrincipleModelingCreditRiskAllsourcesofcreditriskneedtobemod-elled,ordemonstratedtobeinsignificant.
InvestmentsReinsurance/derivativecounterpartyfailureCreditinsuranceTradecreditors,debtorsAlldifferentmanifestationsofcreditriskshouldbemodelingDefaultriskMigrationriskSpreadriskPrinciplesforRegulatoryAdmissibility.
.
.
Qcontinuedfrompage7CROForumRiskManagementQMarch2006QPage8March2006QRiskManagementCROForumPage9QCreditinsuranceshouldbemodelingusingmethodologiesthatreflectthespecificexposurecharacteristicsandriskmitigationoptionsinherentinthebusinessIfcreditexposurescanbeaccuratelyrepre-sentedbyexternalcreditindices(e.
g.
,Euro'A'corporatebondindex)andcreditconcentrationsarenotmaterialrelativetotherelevantindex,thendefaultrisk,migrationriskandspreadriskcanbemodelingonintegratedbasisthroughdirectmodelingoftheindex(e.
g.
,throughanEconomicScenarioGenerator)Ifrepresentativecreditindicesarenotavail-able,orcreditconcentrationsarematerial,thendefaultandmigrationriskneedtobemodelingexplicitlyinamanneralignedwiththeprinci-plesofBaselII.
PrincipleModelingInsuranceRisksForlife/healthinsurancemortality,morbidityandpersistencyriskshouldallbemeasured,ensuringthatparameter,processandcalamityrisksareallcoveredbythemodelingFornon-lifeinsurancetheriskassociatedwithcurrentyearunderwriting(premiumrisk)andprioryears'underwriting(reserverisk)shouldbothbemeasured(eitherinanintegratedmodel,orseparately),againensuringthatparameter,processandcalamity/catastropherisksareallcoveredbythemodelingForbothlife/healthandnon-lifeinsuranceprocess,catastrophe/calamityandparam-eterriskshouldbemeasuredusingeitherscenarioorprobabilisticapproachesProcess(orvolatility)risk,theriskassociatedwiththeanticipatedyear-toyearvolatilityininsuranceresult,shouldbemeasuredprobabilistically,supportedbyscenarioanalysiswhereappropriateSeparateestimationofcalamity/catastropheriskshouldbecarriedoutusingscenarios/probabilitydistribu-tionsbasedonscientificanalysisandexpertopinionParameterrisk–ifsignificant,levelandtrendriskshouldbemeasuredseparatelybasedonacombinationofscientificanalysis,expertopinionandanalysisofhistoricalexperienceReinsurance/risktransferProportionalreinsurancecanbemodelingconsistentlywiththeapproachusedformodelingthegrosslossesForadditionalcredittobegivenfornon-proportionalreinsurancesce-narioorprobabilisticapproachesmustbeusedCapitalmustbeheldtocovertheriskofcounterpartyfailure,takingintoaccountpossibledependenciesbetweenthesizeofgrosslossesoccurringandcounterpartyfailurePrincipleModelingOperationalriskOperationalriskneedstobeexplicitlyaccountedforunderPillar1,inamanneralignedwiththeprinciplesoftheBaselIIapproachReactiontotheStudyRegulatorsandotherinterestedpartieswereap-preciativeandcomplimentarytowardthestudy.
SolvencyIIismovingaheadin2006withinter-nalmodelsasacorefoundationintheframe-work.
WorkisnowunderwaybytheCommitteeEuropeanAssurance(CEA),theEuropeanInsurersAssociation,andalsowiththehelpoftheCROForum,onthedevelopmentofarecom-mendedstandardmodelforinsurers.
SolvencyIIregulationsareexpectedtobedraftedby2007foranexpectedintroductionby2010,creatingamodernfinancialregulatoryinsurancesystemforEurope.
!
MoredetailedinformationcanbefoundintheJune10,2005CRO-forumpaper:"PrinciplesforRegulatoryAdmissibilityofInternalModels.
"Copiescanbeobtainedatthesecretariat,CROForum:GiselleLim.
gisellelim@kpmg.
comFurtherinformation:RiskManagement,IssueNo.
6,November2005:"ChiefRiskOfficerForum:AframeworkforincorporatingdiversificationsinsolvencyAssessmentofinsurers.
""SolvencyAssessmentModelsCompared,"CEAandMercerOliverWyman,March2005IAA:"AGlobalFrameworkforInsurerSolvencyAssessment"2004TheCommitteeofSponsoringOrganizations(COSO)oftheTreadwayCommissionwasstartedbyprofession-alsfromthefollowingfiveprofessionalorgani-zations:TheAmericanAccountingAssociation,TheAmericanInstituteofCertifiedPublicAccounts,TheFinancialExecutivesInstitute,TheInstituteofInternalAuditors,andTheInstituteofManagementAccountants.
ActuariesliketothinkofCOSOasaeuphemismforaccount-antstakingovertheworld,especiallyinviewofitssponsorship.
TheCOSOhasasitsprimarygoaltheimprovementofcorporatefinancialreporting,whichmakesitastrongholdintheemergingpracticeofEnterpriseRiskManagement.
TheCOSOpublishedInternalControl—IntegratedFramework,in1992inresponsetorecentcorporatescandalsandauditimpropri-eties.
Itshouldnotbeasurprisetoanyonethatbusinessscandalsleadtoincreasedregula-tions.
TheSecurityandExchangeCommission(SEC)andtheNationalAssociationofInsuranceCommissioners(NAIC)haveastheirmissiontoprotectconsumerinterestsfromtheeffectsofcorporatemisconduct.
Theironlyweaponsarelegislationandregulation,buttheyareaimedatthegood,thebad,andtheuglyalike.
Paradoxically,theSECandtheNAIC,ineffect,contributetotheerosionofconsumervaluebecausetheburdenofincreasedlegisla-tionandregulationchallengethebestandbiggestofcompaniestosurviveprofitablyundertougheconomicandregulatoryconditions.
TheCOSOprinciplesofinternalcontrolareintend-edtobeself-policing,byprovidingaframeworktoplaceundersurveillancetheactivitiesofkeyareasofacompany.
Asurveillancesystemshouldlinkkeyactivitiesacrossanorganizationandillustratetheimpactontheorganizationofafailureinakeyactivity.
Forexample,ifpolicieserrorfromareservevaluationrun,thenthesur-veillancesystemshouldcapturethemissingpoliciesandtriggeranalerttoindicate,attheveryminimum,thatthenumberofpoliciesval-ueddoesnotagreewiththepolicycountofthevaluationfile.
Whilemorecomplicatedalertsarepossibleandappropriate,itwasraretofindinsurerswiththissimplemodelinplacetovali-datereservesinmymanyyearsofperformingactuarialauditsoninsurancecompanies.
Thenewapproachtoriskmanagementasem-bodiedintheCOSOprincipleslooksatorgani-zationalriskfromabroaderperspectivethanwouldtraditionalriskmanagement.
Traditionalriskmanagementwaspurelyconcernedwiththefrequencyandseverityofexpectedlosses.
Thenewriskmanagementparadigmhasamuchwiderwingspanandcirclesoveramuchwiderlandscapeofanorganizationwithitsinternalcontroldoctrines.
TheCOSOdefinesinternalcontrolasaprocess,effectedbyanentity'sboardofdirectors,managementandotherpersonnel,designedtoprovidereasonableassurancere-gardingtheachievementofobjectivesinthefol-lowingcategories:1)Effectivenessandefficiencyofoperations,2)Reliabilityoffinan-cialreporting,and3)Compliancewithapplica-blelawsandregulations.
Itisimportanttounderstandthefundamentalconceptsuponwhichthisdefinitionrests.
First,internalcon-trolisaprocess,ameanstoanend,notanendinitself.
Second,internalcontroliseffectedbypeo-ple.
Itisnotmerelypolicymanualsandforms,butpeopleateveryleveloftheorganization.
Third,internalcontrolcanbeexpectedtoprovideonlyreasonableassurance,notabsoluteassur-ance,toanentity'smanagementandboard.
Fourth,internalcontrolisgearedtotheachieve-mentofobjectivesinoneormoreseparate,butoverlappingcategories.
Let'sexaminebrieflyeachofthesefundamentalconcepts.
InternalControls—TheCOSOWaybyDorothyL.
AndrewsTheCOSOWayRiskManagementQMarch2006QPage10DorothyL.
Andrews,ASA,MAAA,isvicepresidentoftheriskmanagementdepartmentofWachoviaCorporationinCharlotte,N.
C.
ShecanbereachedatDorothy.
Andrews@wachovia.
com.
ProcessThemostimportantthingtounderstandaboutinternalcontrolisthatitisamanagementtoolconsistingofanetworkofbusinessactivitiesthatarenotonlyinter-related,butalsoreactivetonegativestimuliwithinthenetwork.
Thisnet-workextendstoandisingrainedineverycorneroftheorganization,makingitasmuchoftheessenceoftheorganizationasthatexpressedbytheorganization'smissionstatement.
Inthisway,internalcontrolisnotintendedtorelievemanagementofanactiveandparticipatoryroleinrunningthebusinessortheresponsibilityofadverseconsequencesofbusinessactivities.
TheCOSOwaydescribesinternalcontrolsas"built-ins"ratherthan"built-ons"toanorgani-zation'sinfrastructure.
Thedifferenceisthatbuilt-incontrolsareinternaltoaprocess,whilebuilt-oncontrolsareexternaltoaprocess.
Forexample,enablingvaluationsystemstopro-grammaticallyverifypolicycountsandpremi-umsagainstfinancialledgeramountsisanexampleofabuilt-incontrol.
Inthisscenario,discrepanciesarehighlightedimmediatelyandappropriateactionscanbetaken.
Abuilt-oncontrolwouldinvolveamanualreconciliationofthetwofiles,which,dependingonresources,mayormaynotgetdone.
Built-incontrolsarethehandmaidenstoeffectivequalityinitiatives,aidinginthecontainmentofthecostofdoingbusinessanddecreasingreactiontimetoad-verseevents.
PeopleWeallknowthecliché,"Ourpeopleareourgreatestasset,"orsomethingsimilar.
Theseas-sets,however,canerodecompanyvalueifill-trainedtoperformasneeded.
Internalcontrolisimplementedbyeverymemberoftheorganiza-tion,fromtheboardmemberstothereception-istsandsecurityguards.
Theyallhavearoletoplayineffectingsoundinternalcontrolmanage-ment.
Mostpeopleinanorganizationdonotun-derstandtheimpacttheirjobshaveontheworkproductivityofothers.
Forthisreason,itisim-portanttotrainassociatesatalllevelsofanorganizationintheprinciplesofriskmanage-ment.
Theprinciplesemphasizetheimpactandinter-relationshipsamongfirmactivities.
Informationisamostvaluableassetinacompa-nyandseniormanagementdependsonhighqualityinformationtosteertheorganizationinaprofitabledirection.
However,theflowofinfor-mationinmanyorganizationsisalotlikeplay-ingthefamiliar,childhoodtelephonegame.
Inthetelephonegame,amessageiswhisperedfromonepersontothenextuntilitgetstothelastpersonintheline.
Thelastpersonstandsupandrecitesthemessageandacomparisonismadetothecontentofthemessagewhisperedbythefirstpersonintheline.
Withnearperfectprobabili-ty,therecitationmadebythelastpersonhasnorelationshiptothecontentoftheinitialmessagewhispered.
Thisgameepitomizestheflowofin-formationinmostinsurancecompanieswithseniormanagementasthefinalstop.
Theinstal-lationofasoundsetofinternalcontrolswillim-provethehandoffofinformationaroundtheorganization,andempowermanagementtobet-termanagethecompany.
Keytoinstallinginter-nalcontrolsisanassociateeducationprogram,whichfocusesontheinterplayandimpactofac-tivitiesconductedthroughouttheorganization.
Attheveryminimum,riskmanagementeduca-tionshouldbeginwithnewhiresandthenex-tendtootherswiththegoalofchangingthecurrentculturetoamorerisk-consciousone.
ReasonableAssuranceAnorganizationmaynotsucceedwithinternalcontrols,butitclearlycannotsurvivewithoutthem.
Theyarenotabsoluteinthepreventingmanagementfromnavigatingtheorganizationinthewrongdirection,however.
Bytheirverynature,internalcontrolshavelimitations,asitisnearlyimpossibletomanageforeveryopera-tionalandenterprisecontingency.
But,internalcontrolsystemsdoallowforretrofittingandup-gradingasanorganizationseesfittonarrowtherangeofeventsthatcannudgeitoffcourse.
ThisMarch2006QRiskManagementTheCOSOWayPage11Qcontinuedonpage12Q"Thedifferenceisthatbuilt-incontrolsareinternaltoaprocess,whilebuilt-oncontrolsareexternaltoaprocess.
"impliestheremustalwaysbesomeoneonwatchandreadytoreacttoadverseindicatorstriggeredbythesystem.
ObjectivesCompanyobjectivesgenerallyfallintooneofthreecategories:operations,financialreporting,andcompliance.
Operationalobjectivesincludeallthoseobjectivesrelatingtotheeffectiveandefficientuseoffirmresources.
Financialobjec-tivesrelatetothepreparationoffinancialstate-ments.
Andcomplianceobjectivesrelatetocompli-ancewithlawsandregula-tions.
Operationalobject-ivesdifferfromtheothertwointhattheachievementofthelattertwoobjectivescanbemeasuredbyexter-nalmeans.
Forexample,ei-theracompanyiscompliantwithalaworitisnot.
Operationalobjectivescomeintwoflavors:inter-nalandexternal.
Theachievementofinternalop-erationalobjectivesissub-jecttothepeopleandprocessesofanorganization.
Externaloperationalobjectivesarenotalwayswithincompleteandtotalcontroloftheorganization.
Forexample,theachievementofaspecifiedinvestmentreturnisnotinthesolecontrolofmanagement.
Theinternalcontrolin-frastructureshouldberesponsiveinmeasuringthefitorlackoffitbetweenexternalorganiza-tionalobjectivesandunfoldingexperience.
Itshouldberecognizedthatanorganization'sob-jectivesmayfallintomorethanonecategorytoaddressdifferentneedsandassignaccountabili-tyformeetingthoseobjectivestodifferentoffi-cersofthecompany.
Theoverlapshouldnotpreventareasonableassignmentofexpectationsinmeetingeachcategoryofobjectives.
TheFiveComponentsofInternalControlTheCOSOhasdefinedinternalcontrolascon-sistingofthefollowingfivecomponents:controlenvironment,riskassessment,controlactivities,informationandcommunication,andmonitor-ing.
Eachofthesecomponentsisworthyofmoreattentionthanthetreatmentgivenhere.
However,acoloringoftheroleofeachcompo-nentinbuildinganeffectiveinternalcontrolsys-temisimportanttocompletethisdiscussion.
UnderCOSO,acontrolenvironmentisthesumtotalofthepeoplemakinguptheorganization.
Theirintegrity,ethicalvalues,andcompetencearethemaindriversofacompany'ssuccessorfailure.
Educationbecomeskeyinmakingsureeachmemberofanorganizationunderstandstheriskculturemanagementvaluesandinmakingsureallmembersunderstandtherequiredcom-petenciesrequiredfortheirrole.
Theriskassessmentfunctiononabasiclevelidentifies,analyzesandmanagesrelatedrisks.
Onahigherlevel,riskassessmentinvolvestheintegrationofriskrecognitionwithobjectivesre-latedtosales,production,marketing,financialandotheractivities.
Thisintegrationshoulden-ablealltheseactivitiestoworkintandemtomax-imizecompanyvalue.
Controlactivitiesconsistofthepoliciesandpro-ceduresthatmonitortheexecutionofmanage-mentdirectives.
Theseactivitiescomeinmanydifferentformsdependingonthedirective.
Approvals,authorizations,verifications,recon-ciliations,reviewsofoperatingperformance,se-curityofassets,andsegregationofdutiesareamongthetypesofcontrolactivitiessupportedbyasystemofinternalcontrol.
Theyaredesignedtopreventintentionalandunintentionalbreach-esoftheriskpolicyofanorganization.
Itisuniversallyagreedthatthedeliveryofquali-tyinformationisthecentralingredienttogooddecisionmaking.
TheCOSOrecognizesallthesourcesofbothinternallyandexternallygenerat-eddataandsupportsacompleteinventoryofsuchtodefinetheinter-relatednessofallthepieces.
Theseinter-relationshipsformthebasisInternalControls—TheCOSOWayQcontinuedfrompage11TheCOSOWayRiskManagementQMarch2006QPage12ofariskmanagementsurveillancesystemandareintegraltoaninternalcontrolprocess.
Thecommunicationtoandeducationofassociatesfurthercementstheimportanceoftherolesper-formedbyothersandtheimpactofthesevariousrolesinconcertandinisolation.
Lastly,theeverimportantactivityofmonitoringisanecessaryeviltoensuretheprocessinwork-ingasdesired.
Periodicevaluationsareneces-sarytoflagirregularitiesinthesystem.
Thescopeandfrequencyoftheseactivitiesisafunctionofthedegreetowhichmanualprocess-esareinvolved.
Moremanualtasksnaturallybecomecandidatesformoremonitoringtomaintainequilibriuminthesystem.
Itisimpor-tanttoreportimbalancesupstreamforimmedi-ateresolutiontoempowermanagementtoadjustthecourseoftheorganizationtowardamoreprofitabledirection.
Insummary,installinginternalcontrolsisnosmalltask.
Manyorganizationshaveantiquatedsystemsanddependonmanualprocessescon-trolledbypeopletounderstandtheorganiza-tionalmechanicsthatdrivebottomlineresults.
Italsobecomesverychallengingtoassesshowandwhenpertinentdataadverselychangesformorifithaschangedatall.
Maintainingdataintegrityasdataflowsthroughouttheorganiza-tionmustbeatoppriorityandakeyobjectiveindesigninganinternalcontrolprocess.
Asecondpriorityanddesignincentivemustbethealign-mentofindividualgoalswithcompanyobjec-tives.
Historyhasshownusthatamisalignmentisoftentherootcauseforthedeteriorationofcompanyvalue.
Itismoretruethannotthatthelikelihoodofacatastropheeventbringingdownanorganizationismuch,muchsmallerthanthatofmismanagement.
Therefore,ifanorganiza-tionneedstworeasonsforinstallinginternalcontrolprocesses,thenmaintainingdatain-tegrityandpreventingmismanagementareverystrongones.
Requiredreadingforallriskofficers:InternalControl—IntegratedFramework,September,1992andEnterpriseRiskManagement—IntegratedFramework,September,2004,bytheCommitteeofSponsoringOrganizationsoftheTreadwayCommission.
BothareavailablefromtheAmericanInstituteofCertifiedPublicAccountants(www.
aicpa.
org)forlessthanonebusinessscandaloronefaultyaudit.
!
March2006QRiskManagementTheCOSOWayPage13Q"Historyhasshownusthatamisalignmentisoftentherootcauseforthedeteriorationofcompanyvalue.
"InOctober2005,Standard&Poor'sRatingsServicesannouncedanewaddi-tiontotheanalysisprocessthatleadstotheratingsofinsurers:EnterpriseRiskManagement(ERM)eval-uationcriteria.
WithintheevaluationofERMcapabilities,Stand-ard&Poor'swillprimarilybelookingathowmanage-mentofaninsurerdefinesthelosstoleranceofthefirmandtheprocessesthatarebeingperformedtoas-surethatlossesexperi-encedbytheinsurerarekeptwithinthatlosstoler-ance.
Inaddition,theERMevaluationwillfocusonthedegreetowhichin-surermanagementviewsriskandreturnforrisk-takinginsettingcorporatedirectionandinstrategicdecision-making.
TheERMevaluationwillprimarilybeasubjec-tiveviewofqualityofmanagementpractices.
Thefocuswillbetolookforpracticesthatarebeingcarriedoutinasystematicandconsistentwaythatwillleadtothecontroloffuturelossesinapredictablemannerandthatwillleadtoanoptimalrisk/rewardstructurefortheinsurer'sbusinesses.
TheERMpracticeswillbeviewedincomparisontotherisksofthecompanyandtothepracticesofpeercompanieswithsimilarrisks.
Standard&Poor'swilllookforsophisti-catedrisk-managementpracticestodealwithsophisticatedrisks.
Insurerswillbeviewedashaving"excellent,""strong,""adequate"or"weak"ERM.
Toreachthoseviews,Standard&Poor'swillevaluateERMqualityinfiveareas:I.
RiskManagementCultureRiskmanagementcultureisthedegreetowhichriskandriskmanagementareimportantconsiderationsintheeverydayaspectsofcor-poratedecision-making.
Toevaluateriskmanagementculture,Standard&Poor'swilllookatthestaffingandorganizationalstruc-tureofthepeoplewhoarechargedwithexecut-ingtheriskmanagementfunctionintheinsurer.
Thegovernancestructureasitrelatestoriskmanagementisanotheraspectofrisk-managementculture.
Afavorableindicatorofrisk-managementcultureisastructurethatisindicativeofahighdegreeofinfluenceondecision-makingbyriskmanagementstaff.
Communicationofriskandriskmanagement—bothinsideandoutsideoftheinsurer—arealsoindicatorsofrisk-managementculture.
Aninsurerwithastrongriskmanagementculturewillhaveaverytransparentriskmanagementprocesswithinthecompanyandwithotherinterestedpartiesthroughtheirpubliccommunications.
II.
RiskControlsRiskcontrolisachievedthroughidentifying,measuringandmonitoringrisksaswellasbysettingandenforcingrisklimitsandmanagingriskstomeetthoselimitsthroughriskavoid-ance,risktransferandriskoffsetorotherrisk-managementprocesses.
Standard&Poor'swillevaluaterisk-controlprocessesforeachoftheimportantrisksofaninsurer.
Consistencybe-tweentheoverallcorporaterisktolerancesandthespecificrisklimitswillbeanimportantcon-sideration.
Summarydescriptionsofrisk-con-trolprogramsaswellasexamplesofactualexecutionwillbereviewed.
Standard&Poor'swillbelookingforinsurersthathaveprogramsthatarestructuredtoeffectivelydelivertheriskcontrolneededtomaintaintheexposuresandlosseswithintherisktolerancesaswellascon-sistentexecutionofthoseprogramsthatissuffi-cientlyembeddedineverydaypracticesthatfutureexecutioncanbereliablyinferred.
QPage14Standard&Poor'sEnterpriseRiskManagementEvaluationofInsurersbyDavidN.
IngramDavidN.
Ingram,FSA,isdirectorofenterpriseriskmanagementwithStandard&Poor'sinNewYork,NY.
Hecanbereachedatdavid_ingram@standardandpoors.
com.
EvaluatingInsurersRiskManagementQMarch2006III.
ExtremeRiskManagementExtreme-eventriskmanagementisconcernedwiththeimpactoflowfrequencyadverseeventsonthecompany.
Lowfrequencyeventscannoteasilybemanagedviaacontrolprocessbecausethemonitoringisnotexpectedtoshowanyre-sultsinmostperiods.
Commonextremeeventriskcontrolpracticesincludetrendanalysis,stresstesting,contingencyplanning,problempostmortemandrisktransfer.
Standard&Poor'swillbelookingforinsurerstoshowthattheyarepracticingextremeriskmanagementinadvanceofproblemeventsandwillalsobelook-ingfortheresultsofeffectiveextremeeventriskmanagementduringandafteradverseevents.
Thoseresultswillincludepromptinformationontheexposureoftheinsurertolossfromtheac-tualevent,promptandsurefootedinsurerre-sponsetotheevent,lossesthataremoderatedinsomefashion,andaclearsetoflessonslearnedandadjustmentstofutureprocedures.
IV.
RiskandEconomicCapitalModelsRiskandeconomiccapitalmodelsareanim-portantpartofERM.
Effectiveflowofinforma-tionaboutriskpositionsandtheirpossibleimpactontheinsurerarekeytoeffectiveriskmanagement.
Standard&Poor'sassessestheriskmodelsofaninsurerinrelationtotherisksoftheinsurerandtotheprocessesthattheinsurerhastousetheinformationfromtheriskmodels.
Aninsurerwitheffectiveriskmodelswillbeabletoshowthatthemodelsproducetheinformationneededtoperformthebasicrisk-controlfunctionsthatareneed-edtosustainlossestowithintheirrisktoler-ances.
Thismeansthattheriskmodelsneedtoproduceinformationthatissufficientlyaccu-rate,up-to-dateandtimelytodrivecorrectandwell-timedrisk-managementdecisionsandactions.
Thesemodelsneedtobeclearlyunderstoodbymanagement.
Theinsurerneedstodemonstratearegularprocessofmodelvalidationaswellasaprocessforup-datingdataaboutthebusinessactivitybeingmodeledandtheassumptionsthatareusedinthemodel.
Themodelneedstobesufficientlyrobusttoproduceinsightintoalloftherisksthatareretainedaswellastherisksthatarewrittenbutnotretained.
Themodelsneedtoprovideinformationthatisbothdescriptiveofthesizeoftheriskandinformationthatisac-tionableinmanagingtherisks.
Ifthosetwodifferentobjectivesaremetbydifferentmod-els,thenthetwomodelsneedtobereconciledregularly.
Toaccomplishstrategicriskmanagement,in-surersneedtodeterminetheriskcapitalthatisassociatedwiththeirproducts,investmentsandoperations.
Evaluationofaninsurer'sprocessesfordevelopingriskcapitalinvolveslookingattheunderlyingassumptions,dataflows,valida-tionandcalculationprocesses.
Insurersthatuseregulatoryorratingagencyrisk-basedcapi-talformulaswithoutmodificationwillbepressedtodemonstratethatthosemodelsap-propriatelycapturetheactualrisksoftheirspe-cificbusiness.
Insurersthatmodifythoseformulasinanappropriatemannertoreason-ablyapproximatethecapitalneededtosupporttheirriskpositionsareseentohaveadequatepracticeinthisarea.
Economiccapitalmodelsaresophisticatedanddetailedmodelsthatpro-ducespotvaluesforcapitalneeds,oftenlinkedcloselytospecificmarketvaluesontheexactdayofthecalculation.
Forverycomplexrisks,economiccapitalmodelsmightbetheonlyman-nerofreasonablyidentifyingcapitalneeds.
Atthispoint,Standard&Poor'swillbelookingforappropriateprocessestodevelopriskcapitalamountsthatareconsistentwiththeinsurer'srisksandrisk-managementprograms,thathaveanupdateandvalidationprocessthatproducesaresultthatisconsistentwiththeintentoftheinsurer,andthatareproducedonaschedulethatwillsupportusageintheinsurer'sstrategicrisk-managementprocesses.
Standard&Poor'swillbecontinuingtodeveloprobustprocessesofevaluatinginsurers'EconomicCapitalprocessessoastobetterin-formouroverallviewofthefinancialstrengthandcapitalizationinparticular.
Thisreviewwillonlybeperformedforcompaniesthatarefoundtohaveeffectiveandcoordinatedprocessesforriskcontrol,businesscontinuity,riskmanage-mentculture,andriskmodels.
March2006QRiskManagementcontinuedonpage16QPage15Q"AsERMbecomesthemantrafortoday'sinsuranceorganizations,companiescanfailtoassesstherealimpactofreinsuranceintheERMintegrationprocess.
"EvaluatingInsurersV.
StrategicRiskManagementStrategicriskmanagementistheprocessthataninsurerusestoincorporatetheideasofrisk,riskmanagementandreturnforriskintothecorpo-ratestrategicdecision-makingprocesses.
Riskcapitalisusuallyakeyconceptintheseprocesses.
Standard&Poor'sanalysisofstrate-gicriskmanagementwillstartwithunderstand-ingtheriskprofileoftheinsurerandgettingmanagementexplanationofthereasonsforrecentpastchangesintheriskprofileaswellasexpectedfuturechanges.
Riskpro-filecanbeexpressedintermsofriskcapitalforvariousrisksorforeachofthebusinessesofthein-surer.
Insurersmightalsobeabletoexpressanun-derstandingofthesensi-tivitiesofthatriskprofiletothetimeviewandthelosstoleranceofthemetricused.
Standard&Poor'slooksatthemethodusedfortheallocationofanydiversificationbenefitthatisincorporatedintotheriskprofileandtheimpactoftheallocationchoiceonthestrategicdecisionsmadeusingtheriskcapital.
Strategicprocessesthatcouldbeaffectedbyriskandriskmanagementthinkingincludecap-italbudgeting,strategicassetallocation,prod-uctrisk/rewardstandards,risk-adjustedfinancialtargets,andperformancemeasure-ment,dividendpracticesandincentivecom-pensation.
Thedegreetowhichriskcapitalisvitaltotheseprocessesandtowhichriskandriskmanagementareaconsiderationontheseprocessisindicativeofthequalityofstrategicriskmanagement.
ConcludingRemarksTheevaluationsofeachofthesefiveareaswillbecombinedintoasingleclassificationofqual-ityofERM.
Thedegreeofimportanceofeachfactorinthatjudgmentwillvaryonanindivid-ualbasisamonginsurersaccordingtothespe-cificsituationoftheinsurer.
(SeeTable1onpage17).
TheimportanceofERMinthatprocesswillde-pendontherisksoftheinsurerandthecapacityoftheinsurertoabsorblosses.
Foraninsurerwithahighcapitalpositionand/orexcellentac-cesstocapitalandabusinessplanthatconcen-tratesonretainingonlythoserisksthatarelesscomplexandwellunderstoodbythecompany,ERMwillbelessimportantinformingtheratingdecision.
Forinsurerswithtightcapitaland/orlimitedaccesstocapitalthatareexposedtoverycomplexrisks,ERMwillbeaveryimportantpartoftheratingdecision.
However,capitalisnotseenasasubstituteforERM.
Acompanywithahighcapitalpositionstillneedstobeabletodemonstratethatithastheabilitytomaintainthatpositionthroughlimitingfuturelosses.
Inaddition,Standard&Poor'swillcontinuetoviewaninsurerwithmorecapitaltobemorese-curethananinsurerwithlesscapital.
!
ERMEvaluationofInsurersQcontinuedfrompage15QPage16EvaluatingInsurersRiskManagementQMarch2006March2006QRiskManagementPage17QEvaluatingInsurersTable1:DefinitionsofERMClassificationsClassificationsDefinitionExcellentInsurerhasextremelystrongcapabilitiestoconsistentlyidentify,measureandmanageriskexposuresandlosseswithinthecompany'spredeterminedtoleranceguidelines.
Thereisconsistentevidenceoftheenterprise'spracticeofoptimizingrisk-adjustedreturns.
Riskandriskmanagementarealwaysimportantconsiderationsintheinsurer'scorporatedeci-sion-making.
StrongInsurerhasstrongcapabilitiestoconsistentlyidentify,measureandmanageriskexposuresandlosseswithintheenterprise'spredeterminedtoleranceguidelines.
AstrongERMinsurerissomewhatmorelikelytoexperienceunexpectedlossesthatareoutsideofitstolerancelevelthanisanexcellentERMinsurer.
Thereissomeevidenceoftheenterprise'spracticeofoptimizingrisk-adjustedreturns,thoughitisnotaswelldevelopedasthoseofleadingindustrypractitioners.
Riskandriskmanagementareusuallyimportantconsiderationsintheinsurer'scorporatedecision-making.
AdequateInsurerhascapabilitiestoidentify,measureandmanagemostmajorriskexposuresandlosses,buttheprocesshasnotbeencomprehensivelyextendedtoallsignificantrisksfacingtheenterprise.
Insurerloss/risktoleranceguidelinesarelessdeveloped.
Executionofitsexistingrisk-managementprogramsissufficient,albeitlesscomprehensive,thanarestrongandexcellentERMpractices.
Unexpectedlossesaremorelikelytooccur,especiallyinareasbeyondthescopeoftheexistingERMpractices.
Riskandriskmanagementareoftenimportantconsiderationsintheinsurer'scorporatedecision-making.
WeakInsurerhaslimitedcapabilitiestoconsistentlyidentify,measureandmanageriskexposuresacrossthecompanyand,thereby,limitlosses.
Executionofitsrisk-managementprogramissporadic,andlossescannotbeexpectedtobelimitedinaccordancewithasetofpredeterminedrisk/losstoleranceguidelines.
Riskandriskmanagementaresometimesconsideredintheinsurer'scorporatedecision-making.
Businessmanagershaveyettoadoptariskmanagementframework,aresatisfyingregulatoryminimumswithoutregularlyapplyingriskmanagementtotheirbusinessdecisions,orhaveveryrecentlyadoptedariskmanagementsystemthathasyettobetested.
Mostcompanieshavebeguntocon-siderimplementingenterpriseriskmanagement(ERM)insomeform.
ERMisaprocessthatincludesseveralsteps,including:1)EstablishinganERMframeworkandriskgovernance2)Riskidentification3)Riskassessment4)Riskresponse5)Incorporationintoperformancemeasurement/management6)ExternalriskreportingSomecompaniesaremoreadvancedalongthisprocessthanothers,thoughfewhavemasteredallofthestepsabove.
However,manyinsurancecompanieshavebecomeoverlyfocusedononeofthesestepsinparticular,riskassessment.
ManyinsurershearERMandimmediatelythinkEconomicCapital(EC)—theprocessofbuildingamodeltoquantifytheamountofre-quiredcapitalbasedonaninternalassessmentofcompany-specificrisksandcorrelations.
ThisispartlybecauseEChasthecompellingpotentialtoreducerequiredcapitalbyrecog-nizingriskdiversificationbenefits,aswellasmanyotherapplications.
Also,theactuariesin-volvedinERMareattractedbythechallengeofsuchacomplexmodelingexercise.
SuchcompaniesalsotendtobegintheECeffortveryearlyintheERMprocess,effectivelyjumpingaheadtotheriskassessmentstep(step3).
ECtakesalongtime,sothereisatendencytogetstartedinahurry.
Insurancecompaniestyp-icallyhaveahighlycomplexsetofrisksandsomeverylong-termcontracts.
Quantifyingtheserisksofteninvolvesadvancedtoolsandtechniques,whichcanpushtheenvelopeofmoderndata/projectionsystems.
ECcanbeavaluablecomponentoftheERMprocessforinsurancecompanies.
However,anover-emphasisonEC,tothepointofneglectofotherstepsintheERMprocess,canreducetheeffectivenessofanERMprogram.
Thisisanalo-goustobuildingacriticalmachinepartwithoutfirstconsideringhowitwillmeshwithitsneigh-boringpartsandgears.
Atbest,thiswillcausefriction;atworst,theprocesswillgrindtoahalt.
TheseERMprogramstypicallysufferfromanincompleteintegrationofECintodecision-makingprocessesandalackofbuy-infromin-ternalandexternalstakeholders.
Asaresult,theseERMprogramsareexperiencingdifficul-ties,regardlessofhowsophisticated,completeandaccuratetheirECmodelsmaybe.
TosuccessfullyimplementanERMprogramsupportedbyEC,insurancecompaniesmustbuildtheECmodelonlyaftercarefullyconsid-eringitsinteractionwitheachstepintheERMprocess.
ERMFrameworkThisstepinvolvesdefiningtheERMprocessstepsandhowtheywillinteract,developinganimplementationplan,anddefiningthemetricsandproceduralstructuresforkeystrategicERMdecisions—thosemadebytheERMcommittee.
BuildingtheECmodelwithoutanERMframe-workinplacerequiresassumptionsastotheex-tentandtimingofeachERMprocessstep.
ThiscaneasilyresultintheECmodelbeingunabletosupportotherERMstepsinatimelyfashion.
Onemid-sizeinsurerwasinthemidstofbuild-ingarobustECmodelwhentheERMframe-workwasrevealedrequiringthatECsupportproductpricingwithinaveryshorttimeperiod.
TheECmodelbeingdevelopedwastoorobusttocompletewithintherequiredtimeframe.
However,hadtheoverallframeworkandplanbeenknowninadvance,theECmodelcouldhavebeenbuiltinadvancingstagesofrobust-nesstoprovideatleastadequatepricingsupportinthenearterm.
AnotherimplicationofputtingECmodelingaheadofthisstepisthatECmaybeunabletosupportakeystrategicERMdecision—manag-ingenterpriseriskexposuretowithinriskap-petite.
Thecapital-onlybasisoftheECmeasuremaybeinconsistentwiththeERMframeworkdefinitionofriskappetite.
Forexample,riskap-petitemaybeexpressedasameasureofshare-holdervaluevolatility(basedonadiscountedprojectionofdistributableearnings)ratherthanQPage18"TosuccessfullyimplementanERMprogramsupportedbyEC,insurancecompaniesmustbuildtheECmodelonlyaftercarefullyconsideringitsinteractionwitheachstepintheERMprocess.
"ERM≠EC2bySimSegalSimSegal,FSA,MAAA,isaseniormanagerinDeloitteConsulting'sInsuranceandActuarialSolutionspracticeinNewYork,N.
Y.
Hecanbereachedatsimsegal@deloitte.
com.
ERM≠EC2RiskManagementQMarch2006ameasureofcapitalaloneasprovidedbytheECmodel.
ThiswouldcausedelayswhiletheECapproachisadjustedtosupportthis,thoughthelengthofthetimeneededwillvarydependingonthespecificECmethodologyemployed.
RiskGovernanceInthisstep,managementestablishestheorga-nizationalandfunctionalriskgovernancestructure,includingidentifyingtheexecutiveriskownersanddefiningtheirroles.
Notin-volvingtheexecutiveriskownersearlyonintheECprocesscanfosteroppositiontoEC.
Withoutinputfromexecutiveriskowners,themodelresultswillbesuspect.
However,thiscanbequicklyremediedoncetheyareen-gaged,simplybyrevisingmodelassumptionsandotherinputs.
Ofmoreconcernthoughisthelackofpoliticalbuy-infrominternalstake-holders.
Mostexecutiveriskownersarefromthebusinesssegments.
Excludingthesestake-holdersfromearlyinvolvementmaygivetheimpressionthatECisaneffortthatwillbecon-trolledandimposedbycorporate,withfewusefulapplicationsformanagement.
ThiswillcauseresistanceineveryarenaofERMinwhichECisintendedtooperate.
Thelongerthisnotionisallowedtotakehold,themorechallengingitistoovercome.
BecauseECisprimarilyintendedasatoolemployedbytherisktakersinthebusinesssegments,theearli-erthesestakeholdersareinvolvedandreceivethismessage,thebetter.
RiskIdentificationIftheECmodelprecedestheriskidentificationstep,theECmodelmaybeincomplete,havingignoredcertainrisks.
Forexample,keyrisks(toincludeinECquantification)mayhavebeendefinedinthisstepusingqualitativecriteria,whereastherisksincludedintheECmodelmayhavebeenbasedonquantitativethresholds.
ThiscanresultindelayswhilethemissingrisksareintroducedintotheECapproachandECre-sultsarerevisedbasedonnewriskcorrelationfactors.
Ifthisisnotcorrected,theECmodelwillbeunabletosupportdecisionsinvolvingtherisksexcludedandtheECamountforthere-mainingriskswillbebasedonanincompletecorrelationcovariancematrix.
RiskResponseThisstepincludesthefullrangeofdecisionsthatwillbesupportedbyriskinformationintheERMprocess.
PriortobuildingtheECmodel,itisimportanttounderstandthescopeofdeci-sionsthatthemodelmustsupport.
Withoutthis,theintegrationofECintokeydecision-makingprocessesmaybeincomplete.
Thereareanum-berofissuesthatmustbeaddressedinadvance,includingthefollowing:AtwhatleveloftheorganizationwillECbeex-pectedtosupportdecisions—enterprise,busi-nesssegment,productline,etc.
ThisimpactsECmodelstructureandrequireddataandas-sumptions.
Forexample,assumethattheECmodelwasconstructedtosupportonlybusinesssegment-leveldecisions—thelevelforwhichthiscom-panyhasexistingfinancialdataandsupportingalloca-tions(e.
g.
,investmentin-come,expenses,etc).
However,oncetheriskre-sponsestepisdefined,thereisarequirementthatECsup-portproduct-leveldecisions.
Thiswillcausesignificantdelaystoproducethere-quireddatainputsandmodelenhancementsandtosatisfyotherrequirements,suchastraininganadditionallayerofmanagementintheuseofEC.
Whattypesofdecisionswillbesupported—strategic(e.
g.
,strategicplanning,capitalman-agement,etc.
),tactical(e.
g.
,retentionefforts,hedgingprograms,etc.
),pricing,etc.
Thisim-pactstheprocesseswithwhichtheECeffortmustbecoordinated.
Thisinvolvescoordina-tionofpeopleandprocesses,integrationofsys-temsandbuildingapplicationsthatsupportthespecificdecisions.
Onelargemulti-lineinsurerdevelopeditsECmodelinisolation,withoutthecoordinationneededtointegratethemodelintodecision-makingprocessesthroughthecompa-ny.
Asaresult,afteralengthyandcostlyECmodeldevelopmentexercise,themodelwasonlyusedbythecorporateareaandremaineddisconnectedfromdecision-makingprocessesinthebusinesssegments.
Whatrisksmustbereflectedinthedecisionssupported—justfinancialrisksoralsoopera-tionalrisksThismayimpacttheECmodelingapproach.
Atmanycompanies,theECap-proachusesashortcutmethod(e.
g.
,afixedper-centageofcapital)forassessingoperationalMarch2006QRiskManagementPage19QERM≠EC2continuedonpage20Qrisks.
Someofthesecompanieslaterrealize,intheriskresponsestep,thatthereisaneedforamorerobustapproachtooperationalriskcon-sistentwiththatusedforfinancialrisk.
Thisre-sultsindelayswhiletheECmodelisenhancedtoaddressoperationalrisksinthesamewayitaddressesfinancialrisks.
Atcompanieswherethisissueisnotaddressed,theECmodelisun-abletosupportdecisionsinvolvingoperationalrisks,e.
g.
,evaluatingalternateriskmitigationtechniques.
PerformanceMeasurement/ManagementECmeasuresshouldnotbeintegratedintoper-formancemeasuresandcertainlynotintoincen-tivecompensationuntiltheECmodelisfullydevelopedandstabilized.
However,tosecureinternalstakeholderbuy-inandsupportfortheECeffort,itisimportanttoclearlycommunicateearlyintheprocessthatECmeasureswillulti-matelybeincorporatedintoperformancemeas-urement/management.
ThisdemonstratesseniormanagementcommitmentandwillaligninternalstakeholderinterestswiththeECeffort.
Inaddition,credibilitywithexternalstakehold-erssuchasratingagencieswill,inpart,dependonwhetherthisisbeingdone.
Alackofinternalstakeholderbuy-intotheECeffortisanindica-tionthatthecompanywillnothaveastrongERMprogram.
AlthoughECmeasureswillnotbeincorporatedintoincentivecompensationforsometime,theECapproachshouldconsideritsimplications.
OneimportantconsiderationisthatECishigh-lysensitivetoassumptions.
Tomaintainacred-ibleECmeasure,adisciplinedprocessshouldbeestablishedforthesettingandchangingofas-sumptions.
Thismayincludeacombinationofprovidingincentives(disincentives)foraccura-cy(inaccuracy)andestablishingcorporateguidanceandreviewprotocolsforanymaterialchanges.
ExternalRiskReportingSimilartotheperformancemeasurement/man-agementstep,ECmeasuresshouldnotbeusedinexternalreportinguntiltheECmodeliscred-ible.
However,internallycommunicatingthein-tenttoeventuallyincorporateECintoexternalreportingconveysmanagementcommitmenttotheECapproachandcanbeanadditionaltacticforsecuringinternalstakeholdersupport.
InsuccessfulECprograms,ECmeasuresarelikely,atsomepoint,tobeincludedinexternalreporting—whetherimplicitlyasapartofbusi-nesssegmentearnings(i.
e.
,interestonallocat-edEC)orinasegment-levelReturn-on-EC(ROEC)measureorinsomeothermanner.
Asaresult,itisusefultothinkthroughhowandwhentheECmeasuresshouldbesoemployed,andthelikelyimplicationsofdoingso,duringtheECdevelopmentprocess.
Thiscanassistindiscus-sionswithstakeholdersandinvariouschoicesmadeintheECdevelopmentprocess.
Ifthisisnotdone,thereisachancethatriskdisclosureswillnotbeinsynchwithEC,whichmaybeinter-pretedbyexternalstakeholdersasasignalthattheERMprogramisnotbeingimplementedaswellasitcouldbe.
AsinsurancecompaniesbeginimplementingERM,therearemanystepsintheprocessthatmustbeconsidered.
Theriskassessmentstep,oftenrepresentedbyEC,isacriticalstepinthisprocess,andwhendonecorrectlycanbethecatalystforapowerfulERMprogram.
However,companiesbelievingthatECcanop-erateinavacuumwilllikelyfindtheirERMprogramsoonrunningoutofair.
Incontrast,companiesrealizingandproactivelyaddress-ingtheinter-dependenciesbetweentheriskassessmentstepandotherERMprocessstepswillmorequicklyreapthebenefitsofasuc-cessfulERMprogram.
!
ERM≠EC2RiskManagementQMarch2006ERM≠EC2Qcontinuedfrompage19"CompaniesbelievingthatECcanoperateinavacuumwilllikelyfindtheirERMprogramsoonrunningoutofair.
"QPage20A2003studyofthe30largestcorporatefailures,fraudsandaccountingfias-coesyielded13attributessharedbyvariousgroupsofcompaniesthathadlandedintrouble:70%Unusuallyhighdependenceondebt,ormarginalabilitytomeetdebtrepay-mentrequirements;acquisitionssaddledthecompanywithhugedebts;oroverpaidforacquisitions.
57%Falsifiedfinancialstatementsand/ornonfinancialoperatingmetricstobooststockpriceand/orkeepfinanc-ingcostslow.
53%Unusuallyrapidgrowthand/orunder-pricedproductforrapidgrowth,and/orknowinglyacceptedmorehigh-riskbusinessthanotherfirms.
47%Failedtostress-testassetsandliabili-tiesunderavarietyofassumptionsaboutfutureeconomicandmarketconditions,toapplysophisticatedvaluationmethodstoembeddedderivatives,tocarefullystudycashflowimplicationsofproposedtransaction(s),ortoactonresultsofsuchanalysis.
47%Lackofintegrityinthecompany'sinternalprocesses,systemsandcontrols.
43%Managementfailedtosetappropriatestandardsofethics,integrity,accounting,orcorporategovernance;inadequateoversightbytheBoardofDirectors.
23%Topexecutives,andsometimesdirectors,usedshareholders'fundsaspersonalpiggybank,oftenwithoutinformingallappropriateboardmembers;insidertrading.
17%Management'sreluctancetoadmitproblemsledtohigher-riskinvestmentstrategiesorfinancialengineering.
17%Strategywasnotfocused,clearorconsistent;ormisunderstoodmarket.
17%Company'snaturewasfragile,basedonnontransparentleverage.
10%Significantfinancingarrangementsweretiedtothecompany'screditratingand,insomecases,stockprice.
Withoutthecreditratingorstockpricestrength,allthestructuresimploded.
10%Notabletoadaptandgrowasdeftlyascompetitors;notabletomatchcompetitors'priceprowess;inferiorproduct.
10%Roguetraderconcealedmountinglosses,orprincipalmisrepresentedproduct.
Exhibit1onpages22-24liststhecompaniesin-cludedinthestudyandprovidesabriefsumma-ryofthereasonsfortheirinclusion.
Exhibit2onpages25-26providesanexampleofeachofthe13attributesasmanifestedinonecompany,respectively,fromthestudy.
In2003,ItookonanewroleatJeffersonPilotFinancial(JP)asvicepresident,internalauditplanning&development.
Myprimarygoalwastodevelopandimplementrisk-basedauditplanning.
ItwasagreatopportunitytoapplytheknowledgeIhadgainedfromstudyingaboutEnterpriseRiskManagement(ERM),bypartic-ipatingontheSocietyofActuaries'RiskManagementTaskForce,aswellasmyunder-standingofJPgainedthroughmyworkinitscorporateactuarialdepartment.
Atthetime,JPdidnothaveacomprehensiveERMframeworkthatcouldbeusedasthebasisforsuchauditplanning.
In2004IcreatedaJP-specificframeworkforrisk-basedaudit-planning,butin2003IwasaskedtoprovideanauditprioritizationinashortertimethanIwouldbeabletocompleteonebasedonastudyofJP'sownrisks.
ThemethodologyIchosefortheinitialprioriti-zationwastostudythelargestcorporatefailuresthathadoccurred,determinetheattributestheyshared,andthenidentifytheareasoractivitieswithinJPthatcouldatleasttheoreticallybeex-posedtoanalogousrisks.
ThirteenWaystoKillaCompanyThirteenWaystoKillaCompanybyJenniferBowencontinuedonpage24QMarch2006QRiskManagementPage21QJenniferBowen,ASA,MAAA,issecondvicepresident,accountmanagerandvaluationactuary,SwissReLife&HealthAmericainArmonk,N.
Y.
ShecanbereachedatJennifer_Bowen@swissre.
com.
ThirteenWaystoKillaCompanyRiskManagementQMarch2006Exhibit1CompanyCountryBusinessTypeLoss(billions)WhenCauseWorldComUSAtele-communications$1042002Inappropriatelyaccountedfor$3.
8billioninexpenses;inflatedprofits.
EnronUSAenergy$682001D&O'screatedcomplexoutsidepartnershipsthatkeptbillionsofdollarsinlossesoffEnron'sbalancesheet.
Recordedequitywithoutreceivingthecash.
Underestimatedcosts;bookedallpro-jectedprofitsonfuturesales.
ArthurAndersonactedasEnron'soutsideauditorandalsoperformedinternalauditingservices.
AdelphiaCommunicationsUSAcable$602002$4.
6billionofundisclosedloanstofoundingRigasfamily.
Unconventionaltransactions,questionableaccounting.
GlobalCrossingBermudatele-communications$312002Boguscapacityswapsinflatedrevenue;insidertradingCaseStudiesoffailedEuropeanUnioninsurers15EUcountriesinsurers$301996-2001Fromapopulationof270actualcasesofactualfailuresandnear-misses,21casestudieswereformulated.
Eachcasestudyisanamalgamofmorethanonecase,topreserveanonymity.
ConferenceofSupervisoryServicesoftheEUcountriesperformedthestudytoidentifyrisksthatcanleadtofailure.
Managementproblemsappeartobetherootcauseofeveryfailureornear-failure.
PennCentralUSArailroad$301970Diversification;problemsfrommergerofPennsylvaniaRailroadandNewYorkCentralRailroadin1968;incompatiblecomputersystemsandsignalingsystems.
MirantCorporationUSAenergy$192003Liquiditystrain;lowpowerprices;sloweconomy.
Baldwin-UnitedUSApianomaker/insurer$171983Acquisitionsfinancedbydebt,butthecompanyportrayedthemascashdeals.
KmartUSAdiscountretailer$152002Cutbackonpromotionsduringeconomicdownturn;triedtocompetewithWal-Mart&Targetonsimilarbrandnames.
FINOVAGroupUSAfinancialservices$142001Cashflowtimingmismatch;imprudentlendingpractices;crisisofconfidenceonthepartofitsinvestorsandlenders.
NTL,Inc.
USAcableoperator$132002Debtsspiraledduetotech-boomspendingspree.
RelianceGroupHoldings,Inc.
USAinsurer$132001Duringanill-fatedaggressiveexpansioninthe1990s,thecompanywrotebillionsofdollarsinhigh-riskpoliciesatbargainprices,thenfounditselfresponsibleformassiveunexpectedlosses;divesteditselfofkeybusinesswhileretainingrun-offexposure.
NRGUSAenergy$112003Powerindustry'spost-Enroncreditcrunch.
QPage22ThirteenWaystoKillaCompanyMarch2006QRiskManagementExhibit1(continued)Page23QCompanyCountryBusinessTypeLoss(billions)WhenCauseContinentalIllinoisNationalBankUSAbank$101984Faultsinmanagement,internalcontrols,loanpricing;overlyaggressive;lendinginvolvementwiththreeofthelargestcorporatebankruptciesin1982;turnedincreasinglytoforeignmarketstofunddomesticoperations;littleretailbankingbusinessandthereforerelativelysmallamountsofcoredeposits.
FirstCapitalHoldingsUSAlifeinsurance$101991Irregularinvestmentpracticesandmanipulationoflifeinsurancestatements(FirstCapitalLifeIns.
Co.
andFidelityBankersLifeIns.
Co.
).
FederatedDepartmentStoresUSAretailer$81990SaddledbydebtfromthehighlyleveragedCampeauCorporatetakeoverofFederated.
ConsecoUSAfinancialservices$72002Subprimelending;$120millionD&Osettlement.
TycoUSAconglomerate$62002Taxevasion;CEOandCFOissuedbonusestothemselvesandotheremployeeswithouttheapprovaloftheboard;CEO,CFO&generalcounselgavethemselvesinterest-freeloansforpersonalpurchasesofproperty,jewelry,andotherfrivolity.
Theloanswereneverapprovedorrepaid.
WasteManagementUSAtrashhauler$61998,2001SEClitigation;inflatedcompany'searnings;restated1992-1997earningsby$1.
7billion.
LongTermCapitalManagementUSAhedgefund$51998In1997,concludingthatthecapitalbasewastoohightoearntherateofreturnoncapitalforwhichtheywereaiming,LTCMreturned$2.
7billionofcapitaltoshareholders,increasingitsleveragetoabout25to1.
Madethefirmriskierinthehopeofbolsteringreturnstoshareholders.
Marketconditionsdeterioratedsharply,leadingtomajorlosses.
Russiadevaluedtherubleanddeclaredamoratoriumonfuturedebtrepayments;resultinglossesonrelatedbondsandotherspeculativepositionscausedLTCM'sleverageratiotoclimbto45to1.
TheFederalReserveworkedoutarescuefinancedbyprivatebanksandbrokeragehouses.
MontgomeryWardHoldingCorp.
USAretailer$51997Inadequatebusinessstrategy.
FirstExecutiveCorp.
/ExecutiveLifeUSAinsurer$41991Investedheavilyinjunkbonds;falselyadvertisedproducts,speculatedwiththepremiums;adversepublicityfueledabankrun,forcinga$4billionportfolioliquidationbeforethemarketrose50-60%in1991-2.
CendantCorporationUSAtravel,realestate,financialservices$31998$500millionofrevenuereportedbyCUCfrom1995to1997wassimplyinvented.
Sixty-onepercentofCUC's1997netincomewasfake.
continuedonpage24QMysearchwasalmostentirelyGoogle-based,althoughIdidhavesomehelpfuldocumentsasaresultofmyparticipationontheRiskManagementTaskForce.
Thefirststagewastodeterminewhichcompanieswereworthyofinclusioninthiselitegroup.
IwasnotsureatfirsthowmanycompaniesIwouldin-cludeorwhattheminimumlossshouldbe.
IdecidedthatIwaslookingforfailures,fraudsandaccountingfiascoes.
Somethingverybadhadtohavehappened,althoughthecompanymighttechnicallyhavesurvivedit.
Iwasalsoflexiblewithrespecttoquantifyingthelossinvolved,becauseIwasgatheringinformationfrommanysourcesandthedatawereveryhet-erogeneous.
IdecidedtousewhateverIcouldfindinthewayofpre-eventassets(iftheresultwasabankrupt-cy,forexample),dollarsofincome-statementloss,dropinmarketcapitalization,etc.
BecauseIwastryingtoidentifyagroupofcompaniesforwhomthefinancialrepercussionsweregeneral-lyacceptedtobeverygreat,Iconsideredthisanacceptablemethodology.
Iendedupwith30companiesandaloss-amountthresholdofabout$1billion.
Exhibit1(continued)ThirteenWaystoKillaCompanyRiskManagementQMarch2006CompanyCountryBusinessTypeLoss(billions)WhenCauseHIHAustraliainsurer$32001Egregiousunder-reserving;inabilitytopriceriskproperly;inadequateconsiderationoftimingofcashflows.
Ratherthanrespondingtotheunderly-ingcausesofpoorperformance,HIHmanagementusedandreliedonquestionabletransactionsgiv-ingrisetodoubtfulaccountingentries.
Poorlycon-ceived&executedbusinessdecisions.
Riskswerenotproperlyidentifiedandmanaged.
Boardhardlyanalyzedcompany'sfuturestrategy.
Inadequatepoliciesandguidelinesinessentialareas.
HealthSouthCorporationUSAhealthcareservices$22003OverstatingearningstomakeitappearthatthecompanywasmeetingWallStreetexpectations.
SpiegelUSAretailer$22003Creditcardsforhigher-riskcandidates;merchandisingmissteps;failuretopubliclyreportimprobabilityofcontinuingasagrowingconcern.
AllfirstBankUSAbank$12002Complexandverydetermined,hiddentradinglosses;internalandexternalcollusion;controlsdidnotwork;foreignexchangetradingoperations.
BaringsBankUKbank$11996RoguetraderNicholasLeesonhidmassivelosses;internalstructureofBaringsFuturesSingaporewasseriouslyflawedbythefactthatLeesonhadcontrolofbothfrontandbackoffices.
Drexel,BurnhamLambertUSAinvestmentbank$11998180differentlawsuits;widevarietyofwrongdoing.
GeneralAmericanUSAinsurer$11999Liquiditystrainfrom7-dayputsonitsGICs.
QPage24ThirteenWaystoKillaCompanyMarch2006QRiskManagementExhibit2Page25QAttributeCompanyExampleUnusuallyhighdependenceondebtormarginalabilitytomeetdebtrepaymentrequirements;acquisitionssaddledthecompanywithhugedebts;oroverpaidforacquisitions.
WorldComAmassedabout$32billioninbothbondandbank-loandebtduringatwo-decadespreeofmorethan70acquisitions.
Falsifiedfinancialstatementsand/ornonfinancialoperatingmetricstobooststockpriceand/orkeepfinancingcostslow.
EnronUsedpartnershipstocreatetheillusionthatassetshadbeensold,funnelingcashintoEnronatcriticaltimes,whenthecompanywasstrugglingtomeetWallStreet'sexpectations.
Used"parking"transactions–wheretrueownershipofanassetishiddenthroughsecretguaranteesagainstloss.
Unusuallyrapidgrowthand/orunderpricedproductforrapidgrowth;and/orknowinglyacceptedmorehigh-riskbusinessthanotherfirms.
ConsecoFailedtotakeadvantageofopportunitiestoraisecasheitherbysellinginsurancecompaniesorissuingnewstock.
Triedtogrowitswayoutofitsproblems.
Madeloansformobilehomesandotherpurposesthatturnedouttoberiskierthanthoseitalreadyhad.
Theaggressivelendingwasacceleratedratherthanbeingreinedin.
Failedtostress-testassetsandliabilitiesunderavarietyofassumptionsaboutfutureeconomicandmarketconditions,toapplysophisticatedvaluationmethodstoembeddedderivatives,tocarefullystudycashflowimplicationsofproposedtransac-tion(s),ortoactonresultsofsuchanalysis.
LTCMFailedtoaccountforthefactthatasubstantialportionofitsbalancesheetwasexposedtoageneralchangeinthe"price"ofliquidity.
Ifliquiditybecamemorevaluable(asitdidfollowingthecrisis)itsshortpositionswouldincreaseinpricerelativetoitslongpositions.
Thiswasessentiallyamassive,unhedgedexposuretoasingleriskfactor.
AccordingtothecomplexmathematicalmodelsusedbyLTCM,thepositionswerelowrisk.
Stress-testingagainstthislowercorrelationmighthaveledLTCMtoassumelessleverageintakingthisbet.
Lackofintegrityinthecompany'sinternalprocess-es,systems,andcontrols.
AllfirstForfivefullfinancialyears,Allfirstcontrolsandtreasurymanagementapparentlyfailedtospotanyirregularorquestionabletrading.
Managementfailedtosetappropriatestandardsofethics,integrity,accounting,orcorporategover-nance;inadequateoversightbytheBoardofDirectors.
DrexelBurnhamLambertBrokerstradedonandexchangedinsideinformationobtainedwhileassemblingfinancialbackingforcorporateraiders.
MilkenwasengagedinstockparkingandcolludedwithBoeskyandotherstomanipulatethestocksoftakeovertargets.
Heactivelymisledregulators.
Topexecutives,andsometimesdirectors,usedshareholders'fundsaspersonalpiggybank,oftenwithoutinformingallappropriateBoardmembers;insidertrading.
Adelphia$4.
6billionofundisclosedloanstoRigasfamily.
continuedonpage26QIexcludedcompaniesformorereasonsthanjustsize.
Otherreasonsforexclusion:a)ongoinginvestigation—causesnotyetclear;binsufficientinformation;c)Asiancompanies,whosesituationswereoftennotclearenoughtome;d)toocomplicated;e)victimoflitigation;f)problemsweretooindustry-specific.
Iexcludedsavings&loancompaniesbecausethereweresomanyofthemandtheirproblemsweregenerallysimilarandspecifictothein-dustry,andnotrelevanttomycompany,whichwasinthelifeinsurancebusinessaswellascommunications(radio&TVstationsandsportsprogramming).
Ialsoexcludedbankslendingtotheenergyindustry,formuchthesamereasons.
Irecordedattributesforeachcompany,basedontheassessmentsthatIfoundinpublishedar-ticles.
Ionlyusedconclusionsthatauthorsofthearticleshaddrawn;noneofthecompany-specif-icanalysesweremyown.
Herearesomeofmyownobservations,afterstudyingthestoriesofsomanycompanies:IndustrySpecificRisksTherearedifferenttypesofindustry-specificrisks.
Someofthesehavetodowithregulation,somewithenvironmentalissuesandsomewitheconomicaspectsoftheindustry.
Lookingbackatsomeofthebigscandals,though,IseethatThirteenWaystoKillaCompanyRiskManagementQMarch2006Exhibit2(continued)AttributeCompanyExampleManagement'sreluctancetoadmitproblemsledtohigher-riskinvestmentstrategiesorfinancialengi-neering.
HIHExpansionintolinesofbusinessbeyondtheexpertiseoftheunderwriters.
Strategicdecisionsbasedonlimitedinformation.
Ratherthanrespondingtotheunderlyingcausesofpoorperformance,HIHreliedonquestionableaccountingtransactionswhichdisguisedtheseriousnessofthesituation.
Strategywasnotfocused,clear,orconsistent;ormisunderstoodmarket.
KmartKmart'sfailurewasafailureofmarketing.
Completelymisunderstoodmarket,guessedintheabsolutelywrongdirection,andwascompletelyoutoftouch.
Strategywasallovertheplace.
Company'snaturewasfragile,basedonnon-transparentleverage.
BaldwinUnitedAcquisitionsfinancedbydebt,butthecompanyportrayedthemascashdeals.
Significantfinancingarrangementsweretiedtothecompany'screditratingand,insomecases,stockprice.
Withoutthecreditratingorstockpricestrength,allthestructuresimploded.
GeneralAmericanDowngradetriggeredinvestorscallinginnearly$6.
5billioninGICs.
Notabletoadaptandgrowasdeftlyascompetitors;notabletomatchcompetitors'priceprowess;inferiorproduct.
PennCentralPennandNewYorkCentralculturesclashedbadly.
TherewasconfusionamongthecrewsandPennCentralhadproblemswiththeunionseventhoughitwasforcedtoguaranteeemploymenttoallexistingworkersasaconditionforthemerger.
Sometrainsweremisplacedfordays.
PiggybackvansusedbycorporationslikeEastmanKodakmissedconnections.
Freightbusinessbegantogoelsewhere.
MajorindustrialcustomersabandonedPennCentral.
Roguetraderconcealedmountinglosses,orprincipalmisrepresentedproduct.
BaringsBankRoguetraderhidmassivelosses.
QPage26someofthecompanykillersassociatedwithen-tireindustriesarereallyindustry-concentratedbadhabitsorsociallyacceptabledeviantbehavior.
Itwasnotnecessaryfortheseindus-triestohavecrashedandburned,eitherfinan-ciallyorreputationally.
SomeexamplesofthesebehaviorsarefraudandaggressivelendingbyS&Ls,insidertradingandstockmanipulationbyinvestmentbankers,andconflictsofinterestbyauditors.
Becausethesephenomenadonotnecessarilyarisefromtheinstitutionsthem-selves,theymustbechoicesmadebyindividu-alswhohappentoworkinthoseindustries.
Myobservationisthatitshouldbeacompetitivead-vantageinthelongruntonotengageinsuchbe-haviors.
ItisconvenienttocategorizetheS&Lfailuresashavingcommoncharacteristics,butitwasnotnecessaryforthosebehaviorstooccur.
Perhapsacertaintypeofpersonwasattractedbyanenvironmentthatallowedenoughfreedomforthosebehaviors,butitwasstillamatterofpersonalchoice.
ThePathtoRuinForpurposesofrisk-basedauditingitwasrele-vanttolookattheattributesofthesecompaniesandnotjustthecausesoftheirfailure.
Foronething,thecauseoffailureisusuallynotthatsim-ple.
ButIammoreawarenowofthepathtoruinandthedifferentstagesitmightgothrough.
Besidestwistsandturns,itmightmakeafewcirclesorbecomeasortofspiral.
Isawduringthisresearchthattherewerediffer-enttypesoffatalerrorsthatstartedcompaniesdownthatpath.
Buttherewerealsodifferentpointsatwhichcorrectionscouldhavebeenmade.
Therearedifferentdegreesofserious-nessofthetroublethatacompanyhasgottenit-selfinto,anddifferentdegreesofdesperationinitsresponsetothattrouble.
Thepointatwhichtheriskmanagerorauditorisgoingtomakeanobservationmightbeinanyofthosestages.
Thisisonereasonforaholisticap-proachtoriskmanagement.
Becauseyoudon'tknowwhetherthecompanymightbeinthebadstrategystage,theaggressivebehaviorstage,thelosscontrolstageorthedesperationstage,youhavenoideawhichstageyoumightneedtolookforwhenmakingplansforwhattoobserve.
HumanFactorsIstillhavenotseenanynew-fangledbusinessmodelthathasconvincedmethatgoodmanage-mentisanythingotherthanmaturity.
TheLTCMcasewasaboutjudgmentandmatu-rityintwoways:1)thefundmanagersreturnedcapitaltoinvestorsandincreasedleverage,chasinghighreturns;and2)theydidnotdoenoughstresstestingofkeyassumptionsintheircomplexmathematicalmodel.
EvenGeneralAmerican'ssituation,whichsomecouldsaywasacomplexALMmatter,couldbeviewedasacaseofrelyingtoomuchonoutsideconsultantsinmakingdecisionswithgreatriskpotential.
Also,with20/20hindsightwecanseethatreading,andgivingthoughtfulconsiderationto,akeyprovisioninacontractwasallthatwasneededtoseewhatahugeriskwasinvolved.
Thatdidnotrequireacomplexmathematicalmodel.
Inmostofthesecases,basichumannaturewasakeydriver,andbasicbusinessprinciplesplayedakeyrole.
Iagreethatitisveryimportanttohavegoodtools,andtoapplycontrolsatalllevels.
Butthisresearchshowedmethatlackofdiscipline,judgment,integrity,andasenseofresponsibilitybypeopleinpowerfulpositionswastheundoingofmanyofthesecompanies.
Itissadtoseethatithastakenastringofcorpo-ratedisasterstoraiseawarenessoftheneedformoreaccountabilityonthepartofcorporateex-ecutivesandboardmembers.
Itishardtobeatthesoberinginfluenceofsignificantjailsen-tencesasthebestdeterrentforembezzlementorfraud.
Tocounterthenextlevelofinadequacy,though,belowintentionalcrime,theRiskManagement(RM)culturehasemergedasthebestwaytoachievetheeffectivenessandaccountabilityofcorporateexecutivesandmanagement.
Riskmanagementwillnotbeeffectiveifitisviewedmerelyasanexerciseinfillingoutforms,report-ingmetrics,andestablishingcovariancematri-ces.
Itwillnotmeanachangeincorporatelifeuntilitisrepresentedbyrespectedexecutiveswhohaveaplaceatthetableandwhosevoicesareexpectedtobeheardwheneverimportantdecisionsarebeingmade.
TheBoardofDirectorsisassuredthatthevoicerepresentingRMisbringingupimportantconsiderations,askingappropriatequestions,levelingtheplay-ingfieldwithintheenterpriseonarisk-adjustedbasis,providinganaggregateriskprofilefortheoverallenterprise,andhelpingtoensurethattheriskstheenterprisetakesonare"calculat-ed"risks.
!
ThirteenWaystoKillaCompanyMarch2006QRiskManagementPage27QAuthor'sNote:MuchofthisarticleisexcerpteddirectlyfromtheinvestorsurveyperformedbyanE&YteamheadedbyThomasMcGrath,globalmanagingpartner,ClientServicesandAccounts.
AsmembersoftheRiskManagementSection,thereisnodoubtinyourmindsthatEnterpriseRiskManagement(ERM)isahottopic.
Morethanthat,itisatthetopofthemindsofseniorman-agement,boardsofdirectors,andinvestorsinallindustries.
Drivenbyfearofmajorbusinessfailuresandunderincreasingcompliance-re-latedpressures,companiesarefocusingontheirexposures,andriskmanagementhasbe-comepartoftheessentialfabricofthecorporategovernancestructure.
Ernst&YoungLLPisconductingaseriesofsur-veysaroundrisk,exploringattitudestowardriskanditsmanagement,comparingviewpointsacrosskeystakeholdergroupsincludingin-vestors,seniorexecutivesandauditcommit-tees.
Whilethesesurveysarenotspecifictotheinsuranceindustry,ortotheU.
S.
market,theyarequiterelevanttoourenvironment,especial-lythemostrecentsurveyof137majorinvestors.
Thesurveyparticipantsaretheinstitutionalin-vestorsandassetmanagerswhorunsomeoftheworld'slargestinvestmentportfolios,whorep-resentsignificantinvestorsinmostofourorgan-izations.
Theresultsofthesurveymaynotbethatsur-prisingtotheRiskManagementSection,butthesurveyconfirmsthedirectionwearetakingasanindustrywithincreasedfocus,changesinourcorporatecultureandintegratedapproachesaroundriskmanagement.
Forone,thesurveyshowedthatinvestorsdonotwanttoeliminaterisk(Exhibit1),butexpectrisktobemanagedeffectively.
Therefore,theymustunderstandtheriskstheyaretakingandthus,whenmakinginvestmentdecisions,transparencyisatthetopoftheirlistofpriorities(Exhibit2).
Investorswanttoknowmoreabouthowriskmanagementconceptsareappliedinyourorganization.
Thestudyclearlyshowsthatwhatyoucommunicateaboutriskmanagementisabsolutelyaffectingthosedecisions(Exhibit3),with61percentsay-ingthattheyhavedecidednottoinvestwhereriskmanagementwasdeemedinsufficient.
Basedonthesurveyresults,itisclearthatriskisplayinganincreasingroleandreceivinggreateremphasisininvestmentdecisions.
SowhatareinvestorslookingforTheinvestorssurveyedidentified"clearownershipofrisk"asaprimarysuccessfactor.
Inadditiontoaninte-gratedapproach,theyareclearlyfocusedonthecorporateriskculture,e.
g.
,anunderstandingofriskthroughouttheorganization,activeboardinvolvement,adedicatedriskfunction,etc.
(Exhibit4).
Havingaseparateriskmanagementfunctionwashighonthelistofsuccessfactors,buttheresultsindicatethatmostinvestorsfeelthatownershipstillliesultimatelywiththeCEOandtheBoard(Exhibit5).
Itisclearthatinvestorsarehungryformorein-formationonriskmanagementinmakingtheirdecisions,andinmanycasestheyarecurrentlybasingdecisionsonincompleteinformation.
Thegoodnewsisthatthereisapremiumavail-ableforthosecompaniesthatcandemonstratesuccessfulriskmanagementthroughfrequent,effectiveandtransparentinvestorcommunica-tions.
Insummary,ourabilitytomeetinvestorneedsandmaximizeshareholdervaluethroughcommunicatingourriskprofileanddemon-stratingtheeffectivenessofriskmanagementactivitiesisnotjustimportanttooperatingourbusinesses,butisalsoincreasinglydrivingthevalueandvaluationsinstitutionalinvestorsplaceonourorganizations.
!
QPage28RiskManagementInvestorSurveybyMaryEllenLuningInvestorSurveyRiskManagementQMarch2006"Ernst&YoungLLPisconductingaseriesofsurveysaroundrisk,exploringattitudestowardriskanditsmanagement,comparingviewpointsacrosskeystakeholdergroupsincludinginvestors,seniorexecutivesandauditcommittees.
"MaryEllenLuning,FSA,MAAA,isaseniormanagerinErnst&Young'sInsuranceandActuarialAdvisoryServicesprac-tice.
ThisarticleisbasedonasurveyofinvestorspreparedbyThomasMcGrath,globalmanagingpartner,clientservicesandaccounts.
Shecanbereachedatmaryellen.
luning@ey.
com.
Page29QMarch2006QRiskManagementInvestorSurveyExhibit1:InvestorRiskAppetiteExhibit2:TransparencyIsaPrioritycontinuedonpage30QInvestorSurveyRiskManagementQMarch2006Exhibit3:HowInvestorsUsetheInformationFutureChallengesforCompanies…accordingtoinvestors:"Decidingwhatriskstomitigateandwhatriskstoleaveexposed.
"FundDirector,UnitedKingdom"AsaninvestorIwantthecompaniestotrulycommitthemselvestochangewhatiswrongwhenitcomestoriskmanagement.
"SeniorInvestor,Brazil"Tobemoreintegratedandflexibleandhavetheabilitytochangeasthecompany'sgrowthandprofilechange.
"ChiefInvestmentOfficer,UnitedStates"Simplymakingsurethateverythingtheysayisthetruth.
"SeniorInvestor,UnitedStatesQPage30RiskManagementInvestorSurveyQcontinuedfrompage29March2006QRiskManagementInvestorSurveyExhibit4:SuccessFactorsExhibit5:OwnershipPage31QTheSocietyofActuaries'RiskManagementTaskForceistryingtode-velopbetterestimatesofpolicyholderbehaviorinthetail(PBITT).
Ourmissionistoexamineandultimatelygiveguidancetoactuar-iesonhowtosetpolicyholderassumptionsinextremescenarios.
Wearemostinterestedintheassumptionsusedbycompaniesorconsultantsforthescenariosinthe90CTEcalculationsifstochasticallymodeled,ortheassumptionsforeventsthatoccurabovetwostandarddeviationsofexpectedexperience.
OurfirsteffortwasanSOAquestionnairethatconfidentiallygatheredtherangeofassumptionsactuariesuseinpric-ing,reserving,andriskmanagementofmini-mumguaranteesonVariableAnnuity(VA)products:GuaranteedMinimumDeathBenefit(GMDB):guaranteesminimumaccountvalueatdeath.
GuaranteedMinimumIncomeBenefit(GMIB):guaranteesminimummonthlyincomeatannuitization.
GuaranteedMinimumWithdrawalBenefit(GMWB):guaranteesaminimumstreamofincome,provideditiswithdrawnwithinspecifiedlimitsovertime.
GuaranteedMinimumAccumulationBenefit(GMAB):guaranteesminimumaccountvalueataspecifiedfuturedate.
ThesurveyisavailableontheSOAriskman-agementWebsitehttp://www.
soa.
org/ccm/content/areas-of-practice/finance/research/policyholder-behavior-in-the-tail-survey-re-sults/.
Thequestionsthatwereaskedinthequestionnaireinclude:1.
Theprofileofthecompanies.
2.
Whatequitytailscenariosareassumed3.
Howthecompaniesmodelthelapseandtheirutilizationfunctions4.
Thelapseratesinthetail.
Inthisarticle,weprovideanexcerptofoursur-veyresultsonGMIBandhopefullytherebyen-couragereaderstoreviewourfullsummaryreportforgreaterdetails.
1.
TheProfileoftheCompaniesThefollowingtablegivestheprofileofthepar-ticipatingcompanies(inmillions)thatissueGMIB:2.
TheAssumedEquityTailScenariosDuetotheproliferationofguaranteedminimumdeathbenefitsandguaranteedlivingbenefit,atailscenarioismostlikelyonewithpoorequitymarkets.
However,dependingonthetypeofguaranteessold,atailscenarioforcompanyAmaynotnecessarilybeatailscenarioforcompa-nyB.
Forexample,acompanywithsubstantialratchetguaranteesmaybemosthurtbyarapid-lyrisingscenariofollowedbyacrash,butacom-panywithmostlyreturnofpremiumguaranteeswillnotbebadlyhurtbysuchascenario.
Thewidevariationinstyleofin-forcebusinessmayexplainthewidearrayinresponsestothisques-tion,asdemonstratedinExhibit1onpage33.
3.
DescriptionofLapsesandUtilizationFunctionsItwasrespondedthat69percent(11outof16)usedynamicutilizationforGMIBs:Ofthe10thatdescribedtheirfunction,40percent(4outof10)explicitlystatedthatdynamicutilizationisafunctionofin-the-moneynessandattainedage.
Theremainderonlyreferstoin-the-moneynessasafactorfordeterminingdynamicutilization.
OnecarrierconsiderstheoptionvalueofexercisingtheGMIBversustheoptionQPage32VariableAnnuityGuaranteedBenefitsSurveyJamesReiskytl,FSA,MAAA,isretiredandlivesinMequon,Wis.
Hecanbereachedatjimreiskytl@wi.
rr.
com.
PolicyholderBehaviorintheTail:VariableAnnuityGuaranteedBenefitsSurveyResultsbyJamesReiskytlNetPremiumsAccountValueGuaranteedValueAverage1,3732,4393,44725thPercentile30630052175thPercentile1,3393,2525,406RiskManagementQMarch2006valueofholdingontothevariableannuityinadditiontoconsideringin-the-moneynessandattainedage.
4.
LapseRatesintheTailCarrierswerethenaskedtolisttheirlapseratesinthetailscenariotheydescribedatthebegin-ningofthesurveyunderfourdifferentbenefits(GMDB,GMAB,GMIB,GMWB).
Thecarrierswiththehighestandlowestoveralllapserates,alongwiththeaverageacrossthecarriers,arehighlightedasshowninExhibit2forGMIB.
Itisourhopethattheresultsofthissurveywillenhancetheactuary'sabilitytosetassumptionsfortheseproductsinextremescenarios.
Theymayalsoprovideabasisforfurtherdiscussionofwhatmaybecomecurrentpractices.
Forourfutureactivities,weplantoreportthedif-ferenceinRBCresultsusingtheseassumptionsforamodeledblockofbusinessatanupcomingSOAInvestmentSymposiumtotrytoprovideapossiblemeasureofthesereportedresults.
Wealsoplantodothissurveyagainthisyear.
HopefullythenextreportwillincludecompanyresponsesforthenewVARBCComponentRequirementsforDecember31,2005.
OurnextsurveywilladdresslapsesassumedonUniversalLifeproductswithsecondaryguaran-teesinthetails.
Wegreatlyappreciatethetimeandeffortsofthosewhoresponded.
Weencourageandwel-comecomments,questionsandsuggestionsfromallofyou.
PleasesendthemtoeitherJamesReiskytlatjimreiskytl@wi.
rr.
com,chairofthePolicyholderBehaviorintheTailWorkingGrouporStevenSiegelatssiegel@soa.
org.
!
Page33QMarch2006QRiskManagementVariableAnnuityGuaranteedBenefitsSurveyExhibit1:TailScenarioGivenbyRespondentsExhibit2:LapseRatesforGMIBsTheRiskManagementSectionisplan-ninganumberofContinuingEducation(ContEd)activitiesfor2006.
Thesewillinclude,butarenotlimitedto:I)SpringMeetingSessionsOnceagain,wehaveputtogetheraseriesofses-sionsfortheupcomingspringmeetings.
a)LifeSpringMeeting(May23-24,Hollywood,Florida)Sessionstobeoffered(withthesessioncoordi-natornotedinparentheses),include:ERMandCorporateGovernance(DavidIngram)Inatalkshowformat,ourpanelistswillbeaskedaseriesofquestionsoncorporategovernanceissuesrelatingtoriskmanage-mentandactuarialroles.
Attendeeswillre-ceiveapracticalunderstandingofhowERMfitsintocorporategovernance,andtheeffectivenessofcompanies'ERMprogressinthisarea.
EconomicCapital(EC)–RecentTrendsinImplementation(HubertMueller)ThissessionwillallowattendeestolearnhowothercompaniesareusingECinmak-ingbusinessdecisions,viewstowardsECbyratingagenciesandrecenttrendsinthemarketplace.
HedgingVariableAnnuities(VA)–APracticalDiscussion(DanGuilbert)Panelistswilldiscussthebenefitsofhedg-ingVAguarantees(includingdeathbene-fitsandlivingbenefits),thechallengestheyfaceandhowtheykeepscore.
Attendeeswillgetinsightintotheap-proachesusedbycompaniesforhedgingtheriskfromVAguaranteesandchallengesfacedinexecutionandmeasuringeffec-tiveness.
ImplementingStochasticMethodol-ogiesforReserves&Capital(HubertMueller)Thistwo-partseminarwilltakealookatre-centdevelopmentsregardingtheintroduc-tionofprinciples-basedrulesfordeterminingreservesandcapitalforlifeandannuityproducts,anddiscussinitialexperiencewiththeyear-end2005filingsforVAs(C-3PhaseII).
MeasuringandPricingforTailRisk(LarryRubin)Panelistswillreviewvariousapproachesformeasuringandmanagingtailriskandhowtopricefortailriskusingcapital-mar-ketconsistenttechniques.
Attendeeswillgainapracticalunderstandingofcurrentpricingtechniquesandriskmanagementapproachesformeasuringandmanagingtailrisk.
AvianInfluenza:IsYourCompanyPrepared(MaxRudolph)Thisfacilitatedworkshopisdesignedtobeafollow-uptothesessionontailrisk,focus-ingonavianinfluenza.
Attendeeswillshareinformationandtechniquesforstresstestingtheircompany'spreparednesstopandemicsthroughscenarioplanning.
ManagingMarketConductRisk(FrankSabatini)Thissessionwillexaminethedrivingforcesbehindmarketconducteventsfordistributionsystemsandinsuranceprod-uctsandthepracticesbeingemployedtominimizetheexposuretotheseevents.
Attendeeswillgainapracticalunder-standingofcurrentviewsonmarketexpo-suresandtheapproachesusedtomonitor,measureandmanagetheseexposures.
RiskManagementSectionBreakfast(DavidIngram)Inaddition,theRiskManagementSectionwillhaveabreakfastsessionopentoallmeetingattendeeswherewewilldiscusscurrentsectiontopicsandactivities.
Attendeeswillneedtoregisterinadvance.
b)HealthSpringMeeting(June23-24,Hollywood,Florida)OneriskmanagementsessionwillbeofferedattheHealthSpringMeeting,co-sponsoredwiththeHealthSection:RiskManagementforIndividualHealthProducts(TomCorcoran)Thissessionwillexaminecurrentriskmanagementissuesforindividualhealthproductsandrecenttrendsinthemarket-place.
II)AnnualMeetingSessionsWewillhaveasimilarslateofsessionsfortheAnnualMeetingtobeheldinOctober2006inChicago.
PleasecontactTonyDardis(Tony.
Dardis@TowersPerrin.
com)ifyouareQPage34HubertB.
Mueller,FSA,MAAA,isaprincipalwithTowersPerrininWeatogue,Conn.
HecanbereachedatHubert.
Mueller@towersperrin.
com.
ContinuingEducationContinuingEducation—2006ActivitiesbyHubertMuellerRiskManagementQMarch2006interestedinpresentingonriskmanagementtopics.
III)ERMSymposiumThe4thAnnualERMSymposiumwillbeheldApril23-25inChicagoinconjunctionwiththeBowlesSymposium;SOASeminaronCapitalEfficiencywillbeheldSept.
19-20,2006GARPWSonERMwasheldMarch2.
TheRiskManagementSectionisco-sponsoringthissym-posium,andhasbeenveryactiveinhelpingputtogetheranoutstandingscheduleofsessions.
Also,severalworkshopswillbeheldinadvanceofthesymposium.
Forfurtherdetails,pleasevisitthesymposiumWebsite(www.
ermsympo-sium.
org).
IV)SeminarsTheRiskManagementSectionCouncilhasalsobeenactiveindevelopingseveralseminarsthatareco-sponsoredwith,orledby,otherSOAsec-tionsandotherorganizations.
Theseinclude:Aone-dayworkshopattheGARPConventioninNewYorkAworkshoponERMheldattheGARPConventioninNewYork(March1-2,2006)SOASeminaronCapitalEfficiencyAseminarwillbeheldinSeptember2006(inconjunctionwiththeValuationActuarySymposium),focusedontheuseof(stochastic)embeddedvalue,ERMandALMformeasuring,optimizingandcommuni-catingcapitalefficiency.
Thisseminarisbeingco-sponsoredbytheRiskManagement,TaxationandFinancialReportingSections.
SOASeminaronManagementandPresentationSkillsforRiskManagementProfessionalsWearealsodevelopingaseminarthatwillfocusondevelopingmanagementandpresentationskillsforriskmanagementprofessionals.
Theprogramwouldencompasstheuseofanoutsideconsultanttoteachtech-niquesandanotherportionapplyingthesetechniquestoariskmanage-mentcasestudy.
Thisseminarwillbeco-sponsoredwiththeManagementandPersonalDevelopmentSection.
V)WebcastsWeareplanningonofferingatleastoneortwowebcastsonALM,ERMand/orrelatedtopicsthisyear.
VI)CASERMCourseTheCASisofferinganERMcourseauthoredbyStephenD'Arcy.
Forfurtherdetails,pleasevisittheCASWebsiteatwww.
CASACT.
org.
VII)On-LineERMCourseWeareintheprocessofdevelopinganon-lineERMtrainingcourse,intend-edtotrainnewentrantstotheriskmanagementfieldandrefreshskillsforthosework-inginthefieldalready.
Thiscoursewillbemadeupofex-istingERMtrainingmateri-als,supplementedbythepresentationsandpaperstobediscussedattheupcomingERMSymposium.
VIII)LocalRiskManagementMeetingsFinally,wearelookingtosetupregularlocalmeetingsofriskmanagementsectionmembers,allowingforanexchangeofideasandnetworkingamongriskmanagementprofes-sionals.
Thesemeetingswilleitherbesetupinselectedmetropolitanareas(likeBoston,NewYorkorChicago),orheldinconjunctionwiththemeetingsoflocalactuarialclubs.
CallforAdditionalVolunteersWehaverecentlyexpandedthelistofpeoplein-volvedinplanningContEdactivitiesoftheRiskManagementSection.
Currentvolunteersin-clude:EdBetzTonyDardisToddHendersonDavidIngramHubertMueller(Lead)LarryRubinFrankSabatiniRobertWolfSOALiaison:MichelRochetteWithlotsofactivitiesplannedthisyear,thereisalwaysroomforadditionalvolunteers.
Ifyouwouldliketohelpplanandcoordinatesomeoftheseactivities,orhaveotherideasorsugges-tions,pleasecontactHuberMuelleratHubert.
Mueller@TowersPerrin.
com.
Welookforwardtohearingfromyou!
Page35QMarch2006QRiskManagementContinuingEducationArticlesNeededforRiskManagementYourhelpandparticipationisneededandwelcomed.
Allarticleswillincludeabylinetogiveyoufullcreditforyoureffort.
Ifyouwouldliketosubmitanarticle,pleasecontactKenSengTan,editor,atkstan@uwaterloo.
caorRonHarasym,co-editor,atron.
harasym@aegoncanada.
com.
ThenextissueofRiskManagementwillbepublished:PublicationDateSubmissionDeadlineJuly2006April28,2006PreferredFormatInordertoefficientlyhandlearticles,pleaseusethefollowingformatwhensubmittingarticles:Pleasee-mailyourarticlesasattachmentsineitherMSWord(.
doc)orSimpleText(.
txt)files.
WeareabletoconvertmostPC-compatiblesoftwarepackages.
Headlinesaretypedupperandlowercase.
Pleaseusea10-pointTimesNewRomanfontforthebodytext.
Carriagereturnsareputinonlyattheendofparagraphs.
Theright-handmarginisnotjustified.
Ifyoumustsubmitarticlesinanothermanner,pleasecallJoeAdduci,(847)706-3548,attheSocietyofActuariesforhelp.
Pleasesendanelectroniccopyofthearticleto:Dr.
KenSengTan,ASA,Ph.
D.
UniversityofWaterlooWaterloo,OntarioCanadaN2L3G1phone:(519)888-4567ext.
6688fax:(519)746-1875e-mail:kstan@uwaterloo.
caorRonHarasym,FSAAEGONCanada,Inc.
7thFloor,5000YongeStreetToronto,OntarioM2N7J8phone:(416)883-5051fax:(416)883-5030e-mail:ron.
harasym@aegoncanada.
caThankyouforyourhelp.
March2006QRiskManagementRiskManagementIssueNumber7March2006PublishedbytheSocietyofActuaries475N.
MartingaleRoad,Suite600Schaumburg,IL60173-2226phone:(847)706-3500fax:(847)706-3599www.
soa.
orgThisnewsletterisfreetosectionmembers.
Asubscriptionis$15.
00fornonmembers.
Current-yearissuesareavailablefromtheCommunicationsDepartment.
BackissuesofsectionnewslettershavebeenplacedintheSOAlibraryandontheSOAWebsite:(www.
soa.
org).
Photocopiesofbackissuesmayberequestedforanominalfee.
2005-2006SECTIONLEADERSHIPEditorKenSengTan,ASAe-mail:kstan@uwaterloo.
caCo-EditorRonHarasym,FSAe-mail:ron.
harasym@aegoncanada.
caCouncilMembersDouglasW.
Brooks,FSAAnthonyDardis,FIA,FSAKevinDickson,FCASDanGuilbert,FSARonHarasym,FSADavidIngram,FSAValentinaIsakina,FSAHubertMueller,FSALarryRubin,FSAFrankSabatini,FSAKenSengTan,ASAFredTavan,FSASocietyStaffContactsClayBaznik,PublicationsDirectorcbaznik@soa.
orgNewsletterDesignJoeAdduci,DTPCoordinatorjadduci@soa.
orgFactsandopinionscontainedhereinarethesoleresponsibilityofthepersonsexpressingthemandshouldnotbeattributedtotheSocietyofActuaries,itscommittees,theRiskManagementSectionortheemployersoftheauthors.
Wewillpromptlycorrecterrorsbroughttoourattention.
Thisnewsletterwasprintedonrecycledpaper.
Copyright2006SocietyofActuaries.
Allrightsreserved.
PrintedintheUnitedStatesofAmerica.