1、外文翻译原文MONITORINGANDCONTROLLINGBANKRISKDOESRISKYDEBTHELPABSTRACTWEEXAMINEWHETHERMANDATINGBANKSTOISSUESUBORDINATEDDEBTWOULDENHANCEMARKETMONITORINGANDCONTROLRISKTAKINGTOEVALUATEWHETHERSUBORDINATEDDEBTENHANCESRISKMONITORING,WEEXTRACTTHECREDITSPREADCURVEFOREACHBANKINGFIRMINOURSAMPLEANDEXAMINEWHETHERCHANG
2、ESINCREDITSPREADSREFLECTCHANGESINBANKRISKVARIABLES,AFTERCONTROLLINGFORCHANGESINMARKETANDLIQUIDITYVARIABLESWEDONOTFINDSTRONGANDCONSISTENTEVIDENCETHATTHEYDOTOEVALUATEWHETHERSUBORDINATEDDEBTCONTROLSRISKTAKING,WEEXAMINEWHETHERTHEFIRSTISSUEOFSUBORDINATEDDEBTCHANGESTHERISKTAKINGBEHAVIOROFABANKWEFINDTHATIT
3、DOESNOTPOLICYMAKERSAREACTIVELYCONSIDERINGREQUIRINGBANKSTOISSUESUBORDINATEDNOTESANDDEBENTURESSNDAMANDATORYSNDREQUIREMENTAPPEARSTOBEANIMPORTANTPARTOFTHEMARKETORIENTEDREFORMS,CONTAINEDINTHECONSULTATIVEPAPERISSUEDBYTHEBASELCOMMITTEEONBANKINGSUPERVISION1999THEUSSHADOWREGULATORYCOMMITTEEHASCOMEOUTSTRONGLY
4、INFAVOROFMANDATORYSNDASAMECHANISMFORREALIZINGENHANCEDMARKETDISCIPLINEOFBANKSMOREOVER,THEGRAMMLEACHBLILEYACTOF1999MANDATEDAJOINTFEDERALRESERVEANDUSTREASURYSTUDYOFBANKSNDREQUIREMENTSTHISLEGISLATIONALSOREQUIRESALLLARGEBANKINGFIRMSTOHAVEATLEASTONEISSUEOFSNDOUTSTANDINGATALLTIMESPROPONENTSOFMANDATORYSNDRE
5、GULATIONSUGGESTTHATSNDWILLENHANCEMARKETDISCIPLINEANDCURBEXCESSIVERISKTAKINGINTWOWAYSTHROUGHRISKMONITORINGANDTHROUGHPREVENTATIVEINUENCETHEBENEFITSOFRISKMONITORINGAREREALIZEDIFINVESTORSACCURATELYUNDERSTANDCHANGESINAFIRMSRISKCONDITIONANDINCORPORATETHEIRASSESSMENTPROMPTLYINTOTHEPRICESOFRISKYDEBTISSUEDBY
6、THEFIRMIFTHEYDO,THENCHANGESINCREDITSPREADSWILLPROVIDEUSEFULINFORMATIONTOREGULATORSANDASSISTINSUPERVISIONMOREOVER,BANKSWOULDFACEINCREASEDCOSTSOFFUNDINGSHOULDTHEYADOPTRISKIERSTRATEGIESASARESULT,BANKSWITHSNDMAYBELESSLIKELYTOADOPTRISKYSTRATEGIESINTHEFIRSTPLACETHISISTHEPREVENTATIVEINFLUENCEROLEOFRISKYDEB
7、TTHEPURPOSEOFTHISPAPERISTOEXAMINEWHETHERSNDISSUEDBYBANKSANDBANKHOLDINGCOMPANIESBHCSTOGETHERREFERREDTOASBANKINGFIRMSENHANCESRISKMONITORINGANDPREVENTATIVEINFLUENCETOEVALUATERISKMONITORING,WEEXAMINEWHETHERCHANGESINFIRMSPECIFICRISKSGETREFLECTEDINCHANGESINCREDITSPREADSOFSNDISSUEDBYBANKINGFIRMS,AFTERCONTR
8、OLLINGFORECONOMYWIDEANDLIQUIDITYFACTORSTOEVALUATEPREVENTATIVEINFLUENCE,WEEXAMINECHANGESINTHERISKCHARACTERISTICSOFBANKINGFIRMSAROUNDTHETIMETHEYFIRSTISSUEDSNDALENGTHYLITERATUREEXISTSTHATADDRESSESTHEQUESTIONOFWHETHERTHEMARKETPRICESOFLIABILITIESRESPONDTORISKTAKINGBYINDIVIDUALBANKSTODATE,THERESULTSOFEMPI
9、RICALSTUDIESHAVEBEENMIXEDSTUDIESDONEPRIORTO1992FAILTOFINDASIGNIFICANTRELATIONSHIPBETWEENFIRMRISKANDYIELDSONSUBORDINATEDDEBT1MORERECENTSTUDIES,HOWEVER,DOINDICATETHATRISKISBEINGAPPROPRIATELYPRICEDFOREXAMPLE,FLANNERYANDSORESCU1996FINDTHATFORBANKSOVERTHE19831991PERIOD,YIELDSONSNDWEREAFFECTEDBYACCOUNTING
10、MEASURESOFRISKJAGTIANI,KAUFMAN,ANDLEMIEUX2002CONFIRMTHISRESULTFORTHEPOST1991PERIOD,APERIODTHATFOLLOWSTHEPASSAGEOFTHEFEDERALDEPOSITINSURANCECORPORATIONIMPROVEMENTACTFDICIA,WHICHSUPPOSEDLYREDUCEDTHEBREADTHOFTHESAFETYNETFORBANKS2ARELATEDLITERATURECONCERNSTHEEXTENTTOWHICHFINANCIALMARKETPRICESCONTAINTIME
11、LYANDACCURATEINFORMATIONONTHEFINANCIALCONDITIONOFBANKSTHATISOFUSETOBANKSUPERVISORSEMPIRICALSTUDIESBYBERGER,DAVIES,ANDFLANNERY2000,DEYOUNGETAL2001,ANDEVANOFFANDWALL2000INDICATETHATNEITHERTHEMARKETNORSUPERVISORSPOSSESSALLINFORMATIONONFIRMSPECIFICRISKINALMOSTALLOFTHESNDSTUDIES,THECREDITSPREADOFSNDISDEF
12、INEDASTHEDIFFERENCEINBASISPOINTSBETWEENTHEYIELDTOMATURITYOFTHEISSUEANDTHEYIELDOFANEQUIVALENTTREASURYSECURITYFOREXAMPLE,FLANNERYANDSORESCU1996CALCULATETHEDEFAULTRISKPREMIUMASTHESNDYIELDMINUSTHEYIELDOFATREASURYBONDWITHAPPROXIMATELYTHESAMEMATURITYDATE3ONCEOBTAINED,THESESPREADMEASURESAREOFTENUSEDASDEPEN
13、DENTVARIABLESINAREGRESSIONEQUATIONAGAINSTRISKVARIABLES4NOSTUDIESINTHISLITERATURETHATWEAREAWAREOFATTEMPTTOEXTRACTTHETERMSTRUCTUREOFCREDITSPREADSFOREACHBANKTHEIMPORTANCEOFTHISISWELLRECOGNIZEDTHEREISNOWMUCHEMPIRICALEVIDENCETOSUGGESTTHATCREDITSPREADSOFDIFFERENTMATURITIESFORTHESAMEFIRMMAYMOVEINDIFFERENTD
14、IRECTIONSINPARTICULAR,THECREDITSPREADCURVECANMOVEUPWARD,DOWNWARD,ORREFLECTHUMPEDSHAPEDSHOCKSWECAREFULLYEXTRACTANENTIRECREDITSPREADCURVEFOREACHOFOURFIRMSFOREACHQUARTERTHENWEEXAMINEWHETHERFIRMSPECIFICRISKVARIABLESINFLUENCECREDITSPREADLEVELSANDCONFIRMTHATTHEYDO,EVENAFTERCONTROLLINGFORMARKETWIDEANDLIQUI
15、DITYVARIABLESHOWEVER,RELATINGLEVELSOFCREDITSPREADSTOLEVELSOFFIRMRISKVARIABLESISANECESSARYBUTNOTSUFFICIENTCONDITIONFORCREDITSPREADSTOSERVEASANINFORMATIONSIGNALONCHANGINGBANKRISKWENEEDCHANGESINBANKRISKTOBEREFLECTEDINCREDITSPREADCHANGESHENCE,WEEXAMINEWHETHERCHANGESINCREDITSPREADSREFLECTCHANGESINFIRMSPE
16、CIFICRISKSAFTERCONTROLLINGFORCHANGESINMARKETWIDEANDLIQUIDITYFACTORS,ANDWEDONOTFINDSTRONGANDCONSISTENTEVIDENCETHATTHEYDOOURRESULTCOULDBEDUETOTHEFACTTHATBANKINGFIRMSAREHIGHLYREGULATEDTHEREFORE,WEUSEASAMPLEOFNONBANKINGFIRMSASACONTROLGROUP,ANDFINDSIMILARRESULTS5NEXT,WEEXAMINEHOWBANKRISKCHANGESAROUNDTHET
17、IMEABANKINGFIRMFIRSTISSUESSND,USINGBOTHTHERAWRISKCHANGESOFEACHSNDISSUINGBANKINGFIRMANDTHEMATCHEDADJUSTEDRISKCHANGESCOMPUTEDASTHERISKCHANGESOFEACHSNDISSUINGBANKINGFIRMOVERANDABOVETHATOFAPORTFOLIOOFSIZE,LEVERAGE,ANDPROFITABILITYMATCHEDNONSNDISSUINGBANKINGFIRMSWEDONOTFINDANYSIGNIFICANTCHANGEINTHEFIRMSP
18、ECIFICRISKCHARACTERISTICSTHUS,WEFAILTOFINDEVIDENCECONSISTENTWITHTHEPREVENTATIVEINFLUENCEEFFECTFORSNDWECONCLUDETHATMANDATINGBANKINGFIRMSTOISSUESNDMAYNOTPROVIDETHEPURPORTEDBENEFITSOFENHANCEDRISKMONITORINGORPREVENTATIVEINFLUENCE,ASENVISIONEDBYITSPROPONENTSTHEREMAINDEROFTHEPAPERISORGANIZEDASFOLLOWSSECTI
19、ONIDESCRIBESHOWWECONSTRUCTOURRISKYANDRISKLESSBONDDATABASESSECTIONIIDESCRIBESTHEMODELWEUSETOCONSTRUCTTHECREDITSPREADCURVESFOREACHFIRMANDDISCUSSESTHEFITTOBONDPRICESSECTIONIIIDESCRIBESOURSETSOFFIRMSPECIFIC,MARKET,ANDLIQUIDITYVARIABLESSECTIONIVEXAMINESWHETHERRISKYDEBTFACILITATESMARKETMONITORINGOFBANKRIS
20、KWEPERFORMSEVERALADDITIONALANALYSESINSECTIONVTHATINCLUDEEXAMININGWHETHERCREDITSPREADCHANGESREFLECTFIRMRISKCHANGESBETTERWHENWECONSIDERONLYTHERECENTISSUERSOFSNDCHECKINGWHETHERCHANGESINCREDITSPREADSPREDICTFUTURERATINGMIGRATIONSANDEXAMININGWHETHEROURRESULTSAREBANKSPECIFICORWHETHERTHEYAPPLYMOREGENERALLYS
21、ECTIONVIEXAMINESTHEPREVENTATIVEINFLUENCEBENEFITSOFSNDANDSECTIONVIICONCLUDESIDATAOURFIRSTTASKISTOCONSTRUCTCREDITSPREADCURVESATTHEENDOFEACHQUARTERFORASMANYDIFFERENTBANKINGFIRMSASPOSSIBLE,ANDTHENTOREPEATTHISEXERCISEFORACONTROLSAMPLEOFNONBANKINGFIRMSTHEREASONWEUSEQUARTERSFOROURTIMEINCREMENTSISTHATWEWANT
22、TORELATECHANGESINCREDITSPREADSTOCHANGESINFIRMSPECIFICINFORMATION,WHICHISREPORTEDONLYQUARTERLYMOREOVER,BANKSUPERVISORSNEEDTOACTQUICKLYEG,WITHINAQUARTERINRESPONSETOANYSIGNALSTHATABANKSFINANCIALCONDITIONMAYDETERIORATETHEDATAFOROURANALYSISCOMESFROMTHEFIXEDINCOMESECURITIESDATABASEFISDONCORPORATEBONDCHARA
23、CTERISTICSMATCHEDTOTHENATIONALASSOCIATIONOFINSURANCECOMMISSIONERSNAICDATABASEONBONDTRANSACTIONSFROMJANUARY1994TODECEMBER1999THEFISDDATABASECONTAINSISSUEANDISSUERSPECIFICINFORMATIONSUCHASCOUPONRATEANDFREQUENCY,MATURITY,CREDITRATING,CALLABILITY,PUTTABILITY,CONVERTABILITY,ANDSINKINGFUNDPROVISIONS,ONALL
24、USCORPORATEBONDSMATURINGIN1990ORLATERTHENAICDATABASECONSISTSOFALLTRANSACTIONSFROM1994TO1999BYLIFEINSURANCE,PROPERTYANDCASUALTYINSURANCE,ANDHEALTHMAINTENANCEORGANIZATIONCOMPANIESASDISTRIBUTEDBYWARGA2000THISDATABASEISANALTERNATIVETOTHENOLONGERAVAILABLEWARGA1998DATABASEUSEDBYCOLLINDUFRESNEETAL2001ANDEL
25、TONETAL2000,2001ANDISTHEONEUSEDBYCAMPBELLANDTAKSLER2002THENAICDATABASEHASANADVANTAGEOVERTHEWARGADATABASE,SINCETHEFORMERCONTAINSTRANSACTIONPRICESRATHERTHANQUOTESWERECORDTHETRANSACTIONPRICESANDALLTHECHARACTERISTICSOFEACHTRADEDBONDWESEPARATEALLDATAINTOTWOBROADCATEGORIESTRADESBYBANKINGANDTRADESBYNONBANK
26、INGFIRMSFORBANKS,WEHAVE18,776TRADESACROSS185DIFFERENTBANKSTHENUMBEROFTRADESANDFIRMSARESHOWNINTHEFIRSTTWOCOLUMNSOFPANELAOFTABLEIFORNONBANKS,WEHAVE240,876TRADESINVOLVING3,265DIFFERENTFIRMSOURFIRSTSCREENELIMINATESALLBONDSOTHERTHANFIXEDRATEUSDOLLARDENOMINATEDBONDSINTHEINDUSTRIAL,BANKING,ANDSERVICESSECTO
27、RSTHATHAVENODERIVATIVEFEATURESINPARTICULAR,WEFOCUSONBONDSTHATARENONCALLABLE,NONPUTTABLE,NONCONVERTIBLE,NOTPARTOFAUNITEG,SOLDWITHWARRANTSANDHAVENOSINKINGFUNDWEALSOEXCLUDEBONDSWITHASSETBACKEDANDCREDITENHANCEMENTFEATURESTHISENSURESTHATOURCREDITSPREADSRELATEMOREDIRECTLYTOTHECREDITWORTHINESSOFTHEISSUERRA
28、THERTHANTOTHECOLLATERALFURTHER,WEELIMINATEALLDATATHATHAVEINCONSISTENTORSUSPICIOUSISSUE/DATES/MATURITY/COUPONETC,OROTHERWISEDOESNOTLOOKREASONABLECOLUMNS3AND4OFPANELAOFTABLEISHOWTHENUMBEROFTRADESANDFIRMSTHATREMAINAFTERAPPLYINGTHISFILTERFORBANKSANDNONBANKSFORBANKS,WEARELEFTWITH14,660TRADESOVER144DIFFER
29、ENTBANKSTHISCULLINGOFDATAREPRESENTSJUSTOVER20OFTHETRANSACTIONSFORNONBANKS,WEARELEFTWITH26,608TRANSACTIONSFROM245FIRMSTHISREPRESENTSAMUCHMOREDRAMATICCULLINGOFDATABYAFACTOROF88AMAJORREASONFORTHISDIFFERENCEISTHATTHENUMBEROFNONBANKSISSUINGCONVERTIBLE,CALLABLE,ORPUTTABLEDEBTANDDEBTWITHSINKINGFUNDPROVISIO
30、NSISMUCHHIGHERTHANTHENUMBEROFBANKSISSUINGSUCHDEBTOURSECONDSCREENELIMINATESALLTHOSEFIRMQUARTERCOMBINATIONSFORWHICHWEHADFEWERTHANSEVENTRADESFORTHEQUARTERTHEREASONFORTHISSCREENISTOENSURETHATWECOULDOBTAINRELIABLEESTIMATESFORTHECREDITSPREADCURVEFORAFIRMATTHEENDOFEACHQUARTERFORBANKS,THISLEFTUSWIT9,167TRAN
31、SACTIONSOVER81DIFFERENTBANKS,WHILEFORNONBANKSWEWERELEFTWITH16,480TRANSACTIONSFROM210DIFFERENTFIRMSCOLUMNS5AND6OFPANELAOFTABLEISHOWTHERESULTINGNUMBEROFFIRMSANDTRANSACTIONSUSINGTHISCRITERIONOFTHEBANKANDNONBANKTRANSACTIONS,62SURVIVEDTHISCULLINGBYELIMINATINGDATAFROMFIRMSWITHINFREQUENTTRANSACTIONSINANYQU
32、ARTER,WERUNTHERISKOFBIASINGOURRESULTSAGAINSTFINDINGLIQUIDITYPREMIAHOWEVER,INORDERTOESTIMATECREDITSPREADCURVESATTHEENDOFEACHQUARTERWEDONEEDAMINIMUMNUMBEROFDATAPOINTSOURTHIRDANDFINALSCREENREMOVESTHETRANSACTIONSOFFIRMSFORWHICHWECOULDNOTCOLLECTFIRMSPECIFICRISKMEASURESTHEFIRMSPECIFICRISKDATAARECOLLECTEDF
33、ROMTHEFEDERALRESERVEY9REPORTSANDFEDERALFINANCIALEXAMINATIONCOUNCILSREPORTSOFINCOMEANDCONDITIONCALLREPORTSFORBHCSANDBANKS,ANDFROMCOMPUSTATQUARTERLYFORTHENONBANKSWENEEDEDDATATOCOMPUTEALLOURFIRMRISKMEASURESFORALLTHE24QUARTERSOFOURDATASETANDONEQUARTERBEFOREOURDATABEGINSANDONEQUARTERAFTERITENDSTHISENABLE
34、SUSTOCOMPUTETHECHANGESINFIRMRISKRATIOSTHATWILLBEUSEDASINDEPENDENTVARIABLESINOURREGRESSIONONCREDITSPREADCHANGESTHEEXACTNATUREOFTHESERISKMEASURESWILLBEDISCUSSEDLATERFORBANKS,OURFINALDATABASECONSISTSOF6,590TRANSACTIONSFROM535ISSUESMADEBY50BANKINGFIRMSOFTHESE,17AREBANKSAND33AREBHCSFORNONBANKS,WEHAVE9,70
35、3TRANSACTIONSFROM2,335ISSUESMADEBY133FIRMSWEARE,FINALLY,LEFTWITHADATABASETHATCONTAINSTHETRANSACTIONPRICES,TRADINGDATES,ANDTHESPECIFICTERMSOFSNDS,ORDEREDBYFIRMANDBYQUARTERTHISISTHEFIRSTOFTHETHREEDATABASESWEUSEINTHISPAPERPANELBOFTABLEIPROVIDEDETAILSONMATURITYANDCOUPONOFSNDSASWELLASFIRMRATINGSOFOURFINA
36、LSAMPLEOFBANKINGANDNONBANKINGFIRMSFORBANKINGFIRMSINOURFINALSAMPLE,33OFSNDISSUESHAVEMATURITIESBETWEEN1AND5YEARS,AND26HAVEMATURITIESBETWEEN5AND10YEARSWEFINDTHAT8OFALLBANKINGFIRMSISSUESWERERATEDAAANDABOVE,62WERERATEDA,AND17HADLOWERRATINGSTHEDESCRIPTIVESTATISTICSOFBONDSISSUEDBYNONBANKINGFIRMSINOURFINALC
37、ONTROLSAMPLEAREROUGHLYSIMILARTOTHOSEISSUEDBYTHEBANKINGFIRMSWITHTHEEXCEPTIONOFRATINGSWEWEREUNABLETOFINDTHECREDITRATINGSFORMANYISSUESOFNONBANKINGFIRMSWEUSETHEFINALSAMPLEOFBANKINGANDNONBANKINGFIRMSTOCONSTRUCTCREDITSPREADCURVESFOREACHFIRMTHEAVERAGENUMBEROFISSUESTRANSACTIONSPERFIRMQUARTERUSEDTOCONSTRUCTC
38、REDITSPREADCURVESFORBANKINGFIRMSWAS5011367ALITTLEOVER23OFTHESEBONDSHAVEBETWEEN1AND5YEARSREMAINNGTOMATURITY,ANDALITTLEOVER52HAVEATIMETOMATURITYBETWEEN5AND10YEARSTHUS,THENUMBEROFTRANSACTIONANDISSUESANDTHEIRRANGEOFMATURITIESALLOWUSTOBECONFIDENTTHATMEANINGFULQUARTERLYCREDITSPREADCURVESATTHEFIRMLEVELCANB
39、EDERIVEDOURSECONDDATASETCOMPRISESDAILYESTIMATESOFTHEZERORISKLESSYIELDCURVEUNLIKEFORCORPORATEBONDS,THEREISMUCHINFORMATIONONTREASURYRATESTOSETTHISUP,FOREACHDAYWEUSETHEWEEKLY3MONTH,6MONTH,1,2,3,5,7,10,20,AND30YEARCONSTANTMATURITYTREASURYRATEDATAFROMJANUARY1993TODECEMBER2000OBTAINEDFROMTHEWEBSITEOFTHEFE
40、DERALRESERVEBANKOFSTLOUISWEUSEACUBICSPLINESMOOTHINGPROCEDURETOEXTRACTTHEPARRATESFOR3AND6MONTHMATURITIES,ANDTHENFORALLREMAININGMATURITIESAT6MONTHINTERVALSFROMTHISPARCURVE,WETHENEXTRACTTHEZEROCOUPONRATESFOR3AND6MONTHMATURITIESANDFORALLMATURITIESTHEREAFTERATINTERVALSOF6MONTHSTHEFINALSAVEDOUTPUTFOREACHD
41、AYISTHEANNUALIZEDCONTINUOUSLYCOMPOUNDEDZEROCOUPONYIELDSFORTHE3AND6MONTHRATES,ANDFORTHE1,2,3,5,7,10,20,AND30YEARMATURITIESOURTHIRDDATABASECONSISTSOFQUARTERLYDATAONFIRMSPECIFICRISKRATIOS,MARKETVARIABLES,ANDLIQUIDITYVARIABLES,ASWELLASSTOCKRETURNSANDFIRMRATINGSTHEEXACTNATUREOFTHISDATABASEISDISCUSSEDINSE
42、CTIONIIISOURCEKRISHNAN检查信用利差的变化预测未来的评级标准，以及研究我们的结果是银行的具体或者它们是否适用于更普遍。第六节考察了SND高新区和预防性的影响,第七节总结。1、数据我们的首要任务是利用信用利差曲线建立每个季度末为尽可能多的不同的银行公司，然后重复的非银行公司行使控制样品。我们之所以使用我们的时间增量的四分之三，我们希望有关公司利用特定的信息，报告季度信贷利差的变化改变。此外，银行监管机构必须迅速采取行动，反应（如一个季度内）任何信号，银行的财务状况可能恶化。我们的分析数据来自于公司债券相匹配的保险专员（NAIC）提供债券交易数据库从1994年1月至1999年1
43、2月全国协会的固定收益证券数据库（FISD）特点。该数据库包含诸如FISD票面利率和频率，期限，信用等级，通话能力，推杆能力转化的，和偿债基金规定对所有在1990年或以后到期的美国公司债券，发行及发行人的具体信息。该数据库包括所有NAIC交易从1994至1999年人寿保险，财产和人身意外伤害保险，并按照WARGA（2000）分布式健康维护组织的公司将这个数据库由科林杜弗兰等人使用的代替。该NAIC数据库在WARGA数据库的优势，因为前者包含价格，但不是交易价格。我们会记录所有的交易价格和交易债券的每个特征，并所有数据分为两大类银行业机构和非银行业机构。对银行来说，我们有18776家不同的银行在
44、185个行业。对行业和企业数量都显示在列表中，一列是针对非银行，我们有240876行业涉及不同的公司3,265。我们的在第一个屏幕上消除了所有的债券比固定利率美元债券的工业，银行和服务行业有没有其他衍生工具的功能。我们尤其关注那些不可赎回债券，不可兑换，而不是一个单位的一部分（例如，用认股权证出售），没有偿债基金。我们也排除与资产抵押债券和信用增强功能。这将确保我们的信用利差，而不是更直接涉及到抵押品及发行人的信誉。此外，我们消除了所有的数据不一致或有可疑问题/日期/到期/优惠券等问题，否则是不合理的。表A申请后，我要让这对银行和非银行机构筛选的行业和仍然存在的企业。对银行来说，我们只剩下14
45、660多144不同的银行交易。这种数据只占20的交易。对于非银行机构，我们只剩下26608交易从245公司。这是一个数据更以88的占有交易量。造成这种差别的主要原因是，非发行与偿债基金规定兑换，赎回或卖回银行债务和债务的数量远远超过了多家银行发行此类债券。我们的第二个屏幕上消除了所有这些公司上季度的组合，而我们曾经每季度少于七个行业。消除此屏幕的原因是为了确保我们可以得到一个公司的信用利差在每个季度结束曲线可靠的估计。对于银行来说，这留给我们超过81不同银行9167交易，而对非银行我们与1648万来自210个不同公司的交易了。表A我展示和交易的公司使用这一标准产生的数量，对银行和非银行交易，有
46、62存活。通过消除来自各方面与公司频繁的交易数据，我们就会发现流动性溢价的偏置对我们结果的风险有巨大的影响。然而，为了估计需要在每个季度末，我们做信用利差曲线的数据点的最小数目。我们的第三个和最后一个屏幕上消除的企业，我们不能收集公司特定风险措施的交易。该公司特有的风险数据收集从美联储的Y9份报告与收入条件，银行美国联邦金融检查委员会的报告，并从COMPUSTAT对非银行季刊。我们需要的数据来计算所有24个数据集，我们一个季度采取的所有风险的措施之前，我们坚定我们的数据的开始和结束后的一个季度。这使我们能够在公司风险比率计算的，将用于在我们的回归为自变量对信贷利差的变化而变化。这些风险的措施确
47、切性质将在后面讨论。对银行来说，我们的最终数据库包括6590家银行公司从50535个问题作出交易。对于非银行机构，我们有9703由133个公司所制造的2,335问题的交易。我们终于在同一个数据库，它包含的交易价格、交易日期，以及SND的具体条款，由企业按季度排序。这是我们的三个数据库，本文使用的表B第一小组提供成熟和SND的详细信息以及我们的银行和非银行企业最终样品公司的评级。在我们最后的样本银行公司，33的问题，高新区1至5年期限，26的5至10年到期。我们发现，所有银行公司8的问题评为AA及以上，62被评为A，17的低收利率。由非银行机构发行的债券，我们的最终控制样品大致的描述性统计相类似
48、的银行事务所出具的利率，我们都无法找到非银行机构的信贷评级的许多问题。我们使用的银行和非银行机构最后的样本来构造每家公司的信用利差曲线。了用于构建银行公司信贷利差曲线，每家公司上季度的问题（交易）的平均人数为501（1367）。对这些债券的23多一点有1至5年将走向成熟，一超过52多一点有5至10年的时间来成熟。因此，交易和问题，其期限范围内的数字让我们有信心，有意义的季度信贷利差在企业层面曲线可以计算出来。我们的第二个数据集包括零无风险收益曲线每日估算值。不同的公司债券，国债利率上有很多信息。我们要将其进行设置，每一天我们每周使用3个月，6个月，1、2、3、5、7、10、20和30年期国库券
49、利率数据从1993年1月至2000年12月获得从联邦储备银行圣路易斯的网站。我们使用三次样条平滑程序，提取率标准杆3，然后于6个月的时间间隔和所有剩余期限6个月到期。从这个标准杆曲线，然后提取3零息率在此后每隔6个月和6月到期的所有到期。最后保存的每一天的输出是连续复利年零息票收益率的36个月的利率，为1、2、3、5、7、10、，20、和30年到期。我们的第三个数据库包括对公司特定风险的比率，市场变数和流动性变量，以及股票回报和公司评级季度数据。这个数据库的确切性质在第三节将加以讨论。来源KRISHNANRITCHKENTHOMSON，财政部，FED2005，VOL60第1期，P343378，36P
