... “Financiallinkagesamongfirms and marketscouldheightenthisrisk.Derivativesclearlyhaveexpanded the financiallinkagesamong the institutionsthatusethem and the marketsinwhichtheytrade.Variousstudiesof the October1987marketcrashshowedlinkagesbetweenmarketsforequities and theirderivatives.Accordingtothosestudies,pricesin the stock,options, and futuresmarketswererelated,sothatdisruptionsinonewereassociatedwithdisruptionsin the others.”6. ... The federalgovernmentwouldnotnecessarilyintervenejusttokeepamajorOTCderivativesdealerfromfailing,buttoavertacrisis, the FederalReservemayberequiredtoserveaslenderoflastresorttoanymajorU.S.OTCderivativesdealer,whetherregulatedorunregulated.”Inresponseto the GAO’s1994warningsabove, the financialindustry’sresponsewasbothaudible and caustic.MajorWallStreetfirmspushedbackwithconcertedlobbyingeffortstoblockanyregulatorychangesat the pass,while“ChickenLittle”accusationswereleveledat the authors,Congressionalrequestersof the study, and anyindependentfirm,suchasWeissResearch,thatmadeforecastsbasedonitsconclusions.7 The industry’sprimaryargumentindefenseofderivativeswasthattheyhelpedtoreduceriskthroughhedging, and thateachderivativespositionwasgenerallybalancedagainstoffsettingpositions.However,manylargefinancialinstitutions—suchasBearStearns,LehmanBrothers,MerrillLynch and the AmericanInsuranceGroup(AIG)—wentfarbeyondhedging,transformingtheirderivativesdivisionsintomajorprofitcentersbasedonspeculativetrading.Moreover,theydidnotadequatelyprotectthemselvesagainstdefaultsbytheirtradingpartnersoranticipate the severityof the systemriskstressedby the GAO.Subsequently,asdetailedin the GAO’sfollow‐upreport,FinancialDerivatives:ActionsTakenorProposedSinceMay1994,8some,mostlycosmetic,changesweremade.Buttheydidnothingtoslow the meteoricgrowthof the veryinstruments and practicesthat the GAOidentifiedasposing the greatestthreatstofinancialinstitutions and the financialsystem.Specifically, ... inaworst‐casescenario, the contagioncouldshutdown the entirebankingsystem.Thisfearisnottotallyunjustified.Banking and relatedpanicshaveoccurredbefore—notonlyinpriorerasbutalsoinmorerecenttimes.InJanuary1991,forexample,duetoafloodofwithdrawalsbypanickeddepositors,RhodeIslandGovernorBruceSundlundeclaredabankingemergency and shutdownall45 state charteredsavingsbanks and creditunionsinhis state. Shortlythereafter,wewitnessedasimilarsituationinMaryland. The lessonwasthatpanic and contagionwasnotstrictlyaphenomenonthatendedin the 1930s.AnInstructiveCaseStudy: The RoleofRegulatorsinaMajorDisintermediationCrisisAmongLifeInsurersEvenmorepertinentlessonscanbelearnedfrom the disintermediationthatstruckseverallargeU.S.life and healthinsurersin the early1990s.Theirproblemscanbetracedto the early1980swhenmanyinsurershadguaranteedtopayhigh,fixedyieldstoinvestors,butfoundthemselvesunabletomeetthosepromisesasinterestratesdeclined.Tobridge the gap,severalreachedouttolowerrated,higher‐yieldingsecurities,includingjunkbonds and unratedbonds.Untilthisjuncture,higherriskbondportfoliosin the industrycouldbeexplainedasastopgapsolutiontofallinginterestrates.Butafewinsurers—especiallyExecutiveLifeofCalifornia,FidelityBankersLife, and FirstCapitalLife—saw the potentialofhigh‐riskbondportfoliosasamajorbusinessopportunity.Thesecompaniesweren’tsimplyreluctantbuyersofjunk and unratedbondstofulfillpriorcommitments.Theirentirebusinessplanwaslargelypredicatedon the conceptofjunkbondsfrom the outset. The keywastokeep the junkbondaspectlargelyhiddenfrompublicview,whileexploiting the faith the publicstillhadin the inherentsafetyofinsurance.Tomake the modelasuccess,however,theyneededtwoadditionalelements: the cooperationof the WallStreetratingsagencies and the blessingof the state insurancecommissioners. The cooperationof the ratingagencieswasrelativelyeasy.Formanyyears, the standardoperatingprocedureof the leadinginsurancecompanyratingagency,A.M.Best&Co.,wastoworkcloselywith the insurers:If the companydidnotlike the rating and requesteditnotbepublished,Bestcomplied.If the companywassatisfiedwith the rating,Bestwouldcharge the companytoprintitsratingreports,whichcouldbeusedby the insurer’ssalesforcetomarketitsproducts.25Threenewerentrantsto the businessofratinginsurancecompanies—Moody’s,S&P, and Duff&Phelps(nowFitch)—offeredessentially the sameservice.Butinsteadofearningtheirmoneyfromreprintsofratingsreports,theycharged the insurancecompaniesaflatfeeforeachrating,rangingfrom$10,000to$50,000perinsurancecompanysubsidiary,peryear.Later,A.M.Bestdecidedtochangeitspricestructuretomatch the otherthree,charging the ratedcompaniessimilarup‐frontfees.Asawhole, the ratingsprocesswasstackedinfavorof the companiesfromstarttofinish. The companieswereempoweredtodecideif and when and bywhomtheyweretoberated.Theyweregivenapreviewoftheirratingbeforeitwasrevealedto the public.Theyhad the righttoappealanadverserating and delayitspublication. And, asmentionedabove,iftheywerestillunhappywith the rating,mostof the ratingagenciesallowedthemtosuppressitspublication.Notsurprisingly, the ratingagenciesgaveoutgoodgradeslikecandy.AtA.M.Best, the gradeinflationwassoseverethatindustryinsiderswidelyrecognizedthata“good”Bestratingwasactuallybad.Ithadtobe“excellent”toreallybegood.Thus,inthisfriendlyratingsenvironment,itwasnotdifficultfor the insurerswithlargejunkbondportfoliostogetexcellentgradesfrommostof the ratingagencies.Gettingsimilarcooperationfrom the insuranceregulatorswasnotquiteaseasy.Infact, state insurancecommissionersweregettingsoconcernedabout the industry’sbulginginvestmentsinjunk and unratedbonds,theydecidedtosetupaspecialofficeinNewYork— the SecuritiesValuationOffice—tomonitor the situation.Akeyquestionhotlydebatedbetween the industry and regulatorswas:What’sajunkbond? The standardWallStreetanswerwasundisputed:anybondwitharatingfromS&PorMoody’sofdouble‐Borlower.However,insurerswithsubstantialholdingsofjunkbondswerenotsatisfiedwiththatdefinition.Theyknewitwouldexpose the truesizeoftheirjunkbondholdingsto the public.Sotheylobbied the insurancecommissionerstoredefine the definitionofajunkbond. The commissionersinitiallystruggledwiththisrequest,buttheyultimatelyobliged.Insteadof the standardWallStreetdefinitionofjunk, the SecuritiesValuationOfficeestablished the followingfourbondclasses“yes,”“no*,”“no**,” and “no.” The firstcategorywasconsideredinvestmentgrade,while the three“no”categorieswereconsideredjunkbonds.However, the “yes”categoryactuallyincludedbillionsofdollarsofbondsratedBBorlower (the standarddefinitionofjunk)by the leadingratingagencies.Thiscontinuedforseveralyears.Finally,however,afterprotestsfromindustrywatchdogs, the insurancecommissionersrealizedthisamountedtoajunkbondcover‐up and made the decisiontoend the charade,adopting the standarddouble‐Bdefinition, and reclassifyingover$30billionin“secure”bondsasjunkbonds.Basedon the faultydefinitionofjunkbondsuseduntil1989, the insurancecommissionershadreportedthatFirstCapitalLifehad$842million,or20.2percentofitsinvestedassetsinjunkbondsatyear‐end1989.However,basedon the correct,standarddefinitionofjunkbonds,which the commissionersfinallybeganusingin1990,itturnedoutthatFirstCapitalactuallyhad$1.6billioninjunkbonds,or40.7percentofitsinvestedassets.FidelityBankersLife’sjunkbondholdings,previouslyreportedat$639million,or18.3percentofinvestedassets,jumpedto$1.5billion,or37.6percentofinvestedassets.Alltold, the industry’sjunkbond26holdingsreportedby the regulatorssurgedfrom$51billiononDecember31,1988to$84billiononDecember31,1990,withvirtually the entireincreaseattributableto the changeindefinition.20Large,institutionalinvestorsholdingguaranteedinvestmentcontracts(GICs)were the firsttorushfor the exits,creatingasilentrunon the companies’assets; and inresponse,company and state officialsdeclaredthattheywere“safe.”But the presspublicized the newofficialjunkbonddata,triggeringmasswithdrawalsby the public.Tostem the tideofdisintermediation,allfourweretakenoverby the state insurancecommissioners. And thisaction,inturn,was the preludetoanevenlargerfailure—MutualBenefitLifeofNewJersey,whichfellunder the weightoflossesinspeculativerealestate.Meanwhile, the state guaranteemechanismalsofailed.Mostinsurancepolicyholdershadbeengiven the impressionthat,in the eventofafailure,their state guaranteeassociationswouldpromptlyreimbursethem,muchlike the FDICdoesfordepositorsinfailedbanks.However,asarule, the insuranceguaranteefundshadlittleornofunds;theirstandardoperatingprocedurewastoraise the moneywithnewpremiumassessmentsafter the fact.Thatapproachtendstoworkefficientlywhenjustafewsmallcompaniesfail.Butwhen the failuresarelarge,thereisinsufficienttime and resourcestoraise the neededpremiumsfrom the survivinginsurers,mostofwhicharesmallerthan the largefailedcompanies.Asaresult,inadditiontotakingover the operationsof the failedinsurers, the state insurancecommissionershadnochoicebuttodeclarealong‐termblanketmoratoriumonallcashwithdrawalsbypolicyholders.Wereviewed the statutoryfilingsofeachof the failedinsurers and determinedthattheyhad5.95millionindividual and grouppolicyholders,amongwhich1.9millionheldfixedannuities and otherpolicieswithcashvalue.Consumersinthislattergroupweredeniedaccesstotheirfundsformanymonths.Moreover,asadevicetolegallyavoidinvoking the state guaranteemechanism,ratherthandeclaring the companiesbankrupt,theypronounced the firms“financiallyimpaired,”or“inrehabilitation.”Aftermanymonths, the authoritiesthencreatednewcompanieswithnew,reformedannuitiesyieldingfarlessthan the originalpolicies,whilegivingpolicyholderstwochoices.Either...