六種公司管理風(fēng)險(xiǎn)[文獻(xiàn)翻譯]

1、1外文題目：SixWaysCompaniesMismanageRisk出處：2009（3）harvardbusinessreview作者：RStulz原文：SixWaysCompaniesMismanageRiskAsinveststotuptheirlossesfromthefinancialcrisismanywillbeaskingthemselvesHowdidWallStreetmessupsobadlyWhatwentwro

2、ngwithallthosecomplicatedmodelsEvenbackinNovember2007befethecrisishadreallyhitthestockmarketsonecommentatintheFinancialTimeswrote“ItisobvioustherehasbeenamassivefailureofriskmanagementacrossmostofWallStreet.”O(jiān)fcoursefina

3、ncialinstitutionscansufferspectacularlossesevenwhentheirriskmanagementisfirstrate.Theyareafterallinthebusinessoftakingrisks.Whenriskmanagementdoesfailhoweveritisinoneofsixbasicwaysnearlyallofthemexemplifiedinthecurrentcr

4、isis.Sometimestheproblemlieswiththedatameasuresthatriskmanagersrelyon.Sometimesitrelatestohowtheyidentifycommunicatetherisksacompanyisexposedto.Financialriskmanagementishardtogetrightinthebestoftimes.InthefollowingpagesI

5、’llexplethesixpathstofailureindetail.distributionofhousepricechangeswouldgraphicallymapabellcurvearoundthemeanchange.Thehizontalaxisofthegraphwouldmeasurepricechangestheverticalaxiswouldmeasuretheprobabilitythateachchang

6、ewouldoccur.Theshapeofthebellwouldbedeterminedbyvolatility:Thegreaterthevolatilitytheflatterwiderthecurve.Yourmodelingexercisecouldhavegonewrongintwoways.Firstifhousepricevolatilitywashigherinthefuturethaninthepast(which

7、provedtobethecase)youwouldhavesubstantiallyunderestimatedtheprobabilityofaplungeinprices.Yourbellcurveshouldhavebeenflatterwiderbecausetheprobabilitiesoflargerpricemovements(updown)weregreaterthanyouthoughttheywouldbe.if

8、youhadoverestimatedthemeanchangeyourcurvewouldbetoofartotherightgeneratingfaultyestimatesoftheprobabilitiesofallpossiblechanges.3movements(updown)weregreaterthanyouthoughttheywouldbe.ifyouhadoverestimatedthemeanchangeyou

9、rcurvewouldbetoofartotherightgeneratingfaultyestimatesoftheprobabilitiesofallpossiblechanges.Secondyoumighthavewronglyassumedthatthedistributionoffuturehousepricechangeswasdescribedbythebellcurve.Butifchangesinpriceweren

10、otnmallydistributedthenagraphofthedistributionmightbeskewed—justastheprobabilitydistributionofacoin’sfallingheadsupwouldbeskewedifthecoinwereslightlybent.Bothkindsoferrillustratehowdifficultitistousepastdatatopredictthef

11、uture.Supposeyouhadlookedat30years’wthofhousepricechangestoestimateyourmeanstarddeviation.Perhapsthattimeframewasn’tlongenough—especiallyifthetwodecadesprecedingitweremevolatile.Abellcurveprojectedfrom50years’wthofdatawo

12、uldhavebeenflatterthantheoneyouplottederhapslessfartotheright.supposea50yearanalysisshowedthatlargechangesweremelikelythanwiththenmaldistributionbuta30yearanalysisdidnot.Inthatcasethe30yearanalysiswouldunderstatetherisko

13、faplungeinpricescomparedwiththe50yearanalysisbecausewiththelatterthedistributionwouldhavefattailsrelativetothenmaldistribution.Evenifyoucouldhavecapturedthefullimpactonyourbalancesheetofasignificantdownturninrealestateyo

14、uwouldhavebeenvulnerabletoanotherkindoferrinestimatingtheindirecteffects.Yourbanklikeanyotherfinancialinstitutionwouldhavehadpositionsinmanyassetclasseswhosepricechangescrelatedwithchangesinrealestate.Pricemovementsinall

15、assetclassescrelatewithoneanothermelesshecrelationshavetobeestimatedcrectlyifyouaretodetermineyourbank’sfullexposuretoagivenrisk.Thehigherthecrelationstheriskieryourtotalptfolioofassetstherefethemeequitycapitalyouwillnee

16、dtoprotectyourinstitutionagainstlosses.Obviouslyinestimatingcrelationsfromhisticaldatayouwouldhavebeenvulnerabletoalltheproblemswehavejustdescribed.Tocomplicatemattersfurthercrelationsarenotconstantinfacttheyareknowntoin