1、原文THEAUDITRISKMODEL,BUSINESSRISKANDAUDITPLANNINGDECISIONSIINTRODUCTIONAUDITORSHAVEBEENSHOWNTOADJUSTTHEELEMENTSOFTHEAUDITRISKMODELARMANDINCREASETHEAUDITINVESTMENTINRESPONSETOINCREASINGRISKTHATCLIENTFINANCIALSTATEMENTSCONTAINMATERIALMISSTATEMENTSEG,JOYCE1976GAUMNITZETAL1982KAPLAN1985LIBBYETAL1985OKEEF
2、EETAL1994THEREISALSOEVIDENCETHATAUDITORSRESPONDTOBUSINESSRISKTHERISKOFLOSSORINJURYTOANAUDITORSPROFESSIONALPRACTICEDUETOCLIENTRELATIONSHIPSBYINCREASINGTHEINVESTMENTINTHEAUDITAND/ORCHARGINGFEESABOVETHEAMOUNTSREQUIREDTOCOVERTHECOSTSOFCONDUCTINGAUDITSEG,PRATTANDSTICE1994WALO1995WHILETHISRESEARCHSUGGESTS
3、THATAUDITORSREACTTOBOTHAUDITANDBUSINESSRISK,ITDOESNOTIDENTIFYTHECONDITIONSUNDERWHICHAUDITDECISIONSREFLECTBUSINESSRISKSNOTCAPTUREDBYTHEARMINSUCHCASES,THEARM,ASTANDARDOFPROFESSIONALAUDITPRACTICE,FAILSTODESCRIBEAUDITBEHAVIORTHEARMPRIMARILYADDRESSESTHERISKSASSOCIATEDWITHISSUINGUNQUALIFIEDAUDITOPINIONSON
4、CLIENTFINANCIALSTATEMENTSTHATCONTAINMATERIALMISSTATEMENTSBUSINESSRISK,ONTHEOTHERHAND,ISPRESENTEVENWHENAUDITORSCOMPLYWITHGENERALLYACCEPTEDAUDITINGSTANDARDSGAASANDRENDERAPPROPRIATEAUDITOPINIONSTHEPRIMARYCOSTSASSOCIATEDWITHBUSINESSRISKRELATETOLITIGATION,WHETHERTHEEXPOSURELEADSTOAUDITORSBEINGHELDLIABLEF
5、ORCLIENTSTAKEHOLDERLOSSESOTHERCOSTSRELATETOSANCTIONSIMPOSEDBYREGULATORYBODIES,IMPAIREDREPUTATIONANDFAILURETOCOLLECTFEESBUSINESSRISKENCOMPASSESMORETHANTHERISKSASSOCIATEDWITHISSUINGUNQUALIFIEDOPINIONSONMATERIALLYMISSTATEDFINANCIALSTATEMENTSACLIENTWITHAWEAKINTERNALCONTROLSYSTEMEXPERIENCINGFINANCIALDIFF
6、ICULTY,FOREXAMPLE,INTRODUCESTWOKINDSOFRISKSTHERISKOFAMATERIALMISSTATEMENTANDTHERISKOFFINANCIALFAILURETHEARMREFLECTSONLYTHEFIRSTBUSINESSRISKENCOMPASSESBOTHTHEARMHASBEENDESIGNEDTOSERVEASAGUIDEFORAUDITPLANNINGCUSHINGETAL1995AICPA1997ASOMETIMESTHEGUIDANCEPROVIDEDBYTHEARMISCLOUDEDWHENBUSINESSRISKISBROUGH
7、TINTOJUXTAPOSITIONWITHTHERISKCAPTUREDBYTHEARMARENSANDLOEBBECKE1997,259SUGGESTTHATTHEREISDISAGREEMENTABOUTWHETHERAUDITORSSHOULD,ORSHOULDNOT,ADJUSTACCEPTABLEAUDITRISKFORDIFFERENCESINBUSINESSRISKTHEYREPORTTHATSOMEAUDITORSBELIEVETHATACCEPTABLEAUDITRISKSHOULDBELOWERFORCLIENTSTHATPRESENTHIGHLEVELSOFBUSINE
8、SSRISK,WHILEOTHERSDONOTWECONTENDTHATDIFFERENCESINTHENATUREOFTHEUNDERLYINGCIRCUMSTANCESTHATGIVERISETOBUSINESSANDAUDITRISKWILLLEADTODIFFERENTRESPONSESBYAUDITORSINSOMECASES,AUDITORSWILLADJUSTTHEIRAUDITPLANSFORTHEUNDERLYINGCIRCUMSTANCESINWAYSCOMPLETELYCAPTUREDBYTHEARM,BUTINOTHERCASESAUDITORRESPONSESWILL
9、NOTBECAPTUREDBYTHEARMINTHEFORMERCASES,THEARMWILLDESCRIBEAUDITORBEHAVIORINTHELATTERCASESITWILLNOTAUDITORSARERESPONSIBLEFORPROVIDINGREASONABLEASSURANCETHATFINANCIALSTATEMENTSAREFREEOFMATERIALMISSTATEMENTSMISSTATEMENTS,HOWEVER,CANARISEFROMEITHEROFTWOSOURCESERRORSUNINTENTIONALMISSTATEMENTSANDIRREGULARIT
10、IESINTENTIONALMISSTATEMENTSAND,ASWEARGUEBELOW,THEABILITYOFTHEARMTODESCRIBEAUDITBEHAVIORAPPEARSTODEPENDUPONTHESOURCEWEMONITORTWOASPECTSOFAUDITORRESPONSESTOAUDITANDBUSINESSRISKTHEAMOUNTOFINVESTMENTPLANNEDFORTHEAUDITANDTHEEXTENTTOWHICHTHEAUDITFEECONTAINSAPREMIUMTOCOMPENSATEAUDITORSFORBEARINGRISKWEHYPOT
11、HESIZETHATTHEVARIANCEINTHEAUDITINVESTMENTEXPLAINEDBYBUSINESSRISK,OVERANDABOVETHATEXPLAINEDBYELEMENTSOFTHEARM,ISGREATERWHENTHERISKSTEMSFROMIRREGULARITIESRATHERTHANFROMERRORSWHENTHEAUDITINVOLVESARISKOFUNDETECTEDIRREGULARITIES,THEARMFAILSTOCAPTUREDIMENSIONSOFBUSINESSRISKTHATAUDITORSCONSIDERWHENPLANNING
12、THEAUDITINVESTMENTWHENTHERISKOFUNDETECTEDERRORSISHIGH,ONTHEOTHERHAND,RELEVANTDIMENSIONSOFBUSINESSRISKAREREFLECTEDINTHEARMWEALSOPREDICTTHATAUDITFEESCONTAINDONOTCONTAINARISKPREMIUMWHENTHERISKOFIRREGULARITIESERRORSISHIGHWHENTHERISKOFIRREGULARITIESISHIGH,STANDARDAUDITPROCEDURESCANNOTREDUCETHERISKSTOTOLE
13、RABLELEVELS,SOAPREMIUMISADDEDTOTHEFEECONVERSELY,WHENTHERISKOFERRORSISHIGH,STANDARDAUDITPROCEDURESARESUFFICIENTTOREDUCERISKSTOTOLERABLELEVELS,ANDAUDITORSWILLNOTPERCEIVEANEEDFORARISKPREMIUMWETESTEDTHESEHYPOTHESESINANEXPERIMENTUSING,ASSUBJECTS,34AUDITPARTNERSANDMANAGERSFROMFOURBIG5ACCOUNTINGFIRMSEACHCO
14、NSIDEREDONEOFTWOHIGHRISKCASESWHEREAMATERIALMISSTATEMENTWASREVEALEDWHILECONDUCTINGPRELIMINARYANALYSESANDINQUIRIESOFAPROSPECTIVECLIENTSACCOUNTINGPERSONNELTHETWOCASESDESCRIBEDTHEDISCOVERYOFEITHERANERRORORIRREGULARITY,THEREBYSUGGESTINGTHATTHERISKOFOTHER,UNDETECTEDERRORSORIRREGULARITIESWASHIGHAUDITORSASS
15、ESSEDTHEELEMENTSOFTHEARMANDBUSINESSRISKANDMADERECOMMENDATIONSFORTHEAUDITINVESTMENTANDFEEBOTHBEFOREANDAFTERDISCOVERINGTHEMISSTATEMENTTHERESULTSSUPPORTTHEHYPOTHESES,SUGGESTINGTHATBOTHTHEABILITYOFTHEARMTODESCRIBEPLANNEDAUDITINVESTMENTSANDAUDITORINCLINATIONSTOINCLUDERISKPREMIUMSINAUDITFEESDEPENDUPONTHEN
16、ATUREOFTHERISKSPRESENTINTHEAUDITINTHEERRORCONDITION,THEARMEXPLAINEDSIGNIFICANTVARIANCEINTHEAUDITINVESTMENTANDBUSINESSRISKADDEDNOADDITIONALEXPLANATORYPOWERINTHEIRREGULARITYCONDITION,BUSINESSRISKDOMINATEDTHEARMINTHEEXPLANATIONOFTHEAUDITINVESTMENTFURTHER,THEAUDITFEEDIDNOTCONTAINARISKPREMIUMINTHEERRORCO
17、NDITIONTHEFEEREFLECTEDONLYTHECOSTSOFTHEAUDITINVESTMENTINTHEIRREGULARITYCONDITION,AUDITORSPROPOSEDAFEETHATEXCEEDEDTHEAMOUNTNEEDEDTOCOVERTHEPLANNEDAUDITINVESTMENTTHESERESULTSQUESTIONWHETHERCURRENTAUDITSTANDARDSAREPROVIDINGUSEFULGUIDANCEINTHEPRESENCEOFIRREGULARITIESSEESHIBANO1990FORARELATEDARGUMENTANIM
18、PORTANTPURPOSEOFPROFESSIONALSTANDARDS,LIKETHEARM,ISTOPROMOTECONSISTENCYACROSSAUDITENGAGEMENTSCUSHINGETAL1995,BUTINTHECASEOFIRREGULARITIESWHEREAUDITORSAPPEARTODEPARTFROMTHEARM,ITISUNLIKELYTHATTHESTANDARDISHELPINGTOACHIEVETHATGOALTHERESULTSALSOARECONSISTENTWITHMENONANDWILLIAMS1994,WHOSTATETHATITISIMPO
19、RTANTTHATAUDITFEESREFLECTTHEINSURANCESERVICEAUDITORSPROVIDEFORTHEIRCLIENTSESPECIALLYRISKYCLIENTSAUDITFEESAPPEARTOINCLUDERISKPREMIUMSWHENTHERISKOFIRREGULARITIESISHIGHPRIORRESEARCHEXAMININGRELATIONSHIPSAMONGAUDITRISKS,INVESTMENTANDFEESHASUSEDBOTHBEHAVIORALANDARCHIVALMETHODSSEVERALBEHAVIORALSTUDIES,FOR
20、EXAMPLE,HAVEFOUNDRELATIONSHIPSBETWEENINHERENTAND/ORCONTROLRISKANDAUDITINVESTMENTEG,JOYCE1976GAUMNITZETAL1982LIBBYETAL1985MORERECENTLY,PRATTANDSTICE1994RELATEDCLIENTCHARACTERISTICSTOAUDITINVESTMENTANDPROVIDEDEVIDENCEOFRISKPREMIUMSINTHEPRESENCEOFBUSINESSRISKHOWEVER,THEAUTHORSNEITHERMEASUREDTHEELEMENTS
21、OFTHEARMNORIDENTIFIEDCONTEXTUALFACTORSTHATCAUSERISKPREMIUMSPRIORARCHIVALRESEARCHGENERALLYHASFAILEDTOFINDRELATIONSHIPSBETWEENINHERENTAND/ORCONTROLRISKANDAUDITINVESTMENTEG,BEDARD1989MOCKANDWRIGHT1993OKEEFEETAL1994HACKENBRACKANDKNECHEL1997SUCHRELATIONSHIPSMAYBEDIFFICULTTOIDENTIFYINARCHIVALSETTINGSBECAU
22、SEDIRECTANDTIMELYMEASURESOFRISKARENOTREADILYAVAILABLETHISLIMITATIONFORCESARCHIVALRESEARCHERSTOINFERRISKASSESSMENTSFROMTHEDATAAND/ORELICITTHEMAFTERCOMPLETIONOFTHEAUDITINSOMECASESFROMAUDITORSWHODIDNOTMAKETHEACTUALAUDITPLANNINGJUDGMENTSINADDITION,ARCHIVALSTUDIESMUSTRELYUPONRELATIVELYCRUDEPROXIESFORINHE
23、RENTANDCONTROLRISKEG,CATEGORICALVARIABLESCODEDAS0,1THATAREUNLIKELYTOREFLECTTHELEVELOFRISKSINSPECIFICAUDITCONTEXTSFINALLY,WHILEARCHIVALSTUDIESHAVESHOWNTHATAUDITFEESINCREASEINTHEPRESENCEOFFACTORSASSOCIATEDWITHBUSINESSRISK,THEYHAVEBEENUNABLETODETERMINEWHETHERTHEFEEINCREASESAREATTRIBUTABLETOGREATERAUDIT
24、INVESTMENTSORRISKPREMIUMSSIMUNICANDSTEIN1996OUREXPERIMENTALAPPROACHELICITSDIRECTMEASURESOFAUDITRISK,BUSINESSRISK,ANDPLANNEDAUDITINVESTMENTANDFEES,HOLDINGCONSTANTFACTORSSUCHASCLIENTATTRIBUTES,INDUSTRY,AUDITORBACKGROUNDANDTHEAUDITENVIRONMENTWEALSOMANIPULATETHENATUREOFTHEEVENTUNDERLYINGTHERISKDISCOVERY
25、OFANERRORORIRREGULARITYENABLINGUSTOEXAMINEWHETHERTHERELATIONSHIPSAMONGAUDITRISK,BUSINESSRISKANDAUDITINVESTMENTDEPENDUPONWHETHERTHEBUSINESSRISKSARISEFROMPOTENTIALUNDETECTEDMATERIALERRORSORIRREGULARITIESINADDITION,WETESTTHEEXTENTTOWHICHAUDITFEESAREEXPLAINEDBYTHEAUDITINVESTMENTAND/ORBUSINESSRISKFORERRO
26、RSANDIRREGULARITIES,INORDERTOEXAMINETHECONDITIONSUNDERWHICHFEESCONTAINRISKPREMIUMSRATHERTHANEXTRACOMPENSATIONFORADDITIONALAUDITEFFORTTHENEXTSECTIONCONTAINSADESCRIPTIONOFTHEAUDITRISKANDAUDITPLANNINGMODELS,FOLLOWEDBYTHEDEVELOPMENTOFTHEHYPOTHESESWETHENDESCRIBETHEEXPERIMENTALDESIGNANDPRESENTTHERESULTSCO
27、NCLUSIONSANDSUGGESTIONSFORFUTURERESEARCHCOMPLETETHEPAPERIITHEORYANDHYPOTHESESINTHISSECTION,WEUSETHEARMANDANAUDITPLANNINGMODELTOPROVIDEAFRAMEWORKFORDEVELOPINGTHEHYPOTHESESWEPREDICTDIFFERENCESBETWEENERRORSANDIRREGULARITIESINTERMSOFWHETHER1BUSINESSRISKPROVIDESEXPLANATORYPOWERFORTHEAUDITINVESTMENTBEYOND
28、THEELEMENTSOFTHEARM,AND2THEAUDITFEECONTAINSARISKPREMIUMAUDITRISKMODELANDTHEEXPLANATIONOFAUDITINVESTMENTSASNO47REQUIRESAUDITORSTOUSETHEARMASPARTOFTHEAUDITPLANNINGPROCESSAICPA1997ATHEFOLLOWINGPRESENTATIONOFTHEARM,INCLUDINGVARIABLEDEFINITIONS,ISADAPTEDFROMSASNO471DETECTIONRISKACCEPTABLEAUDITRISK/INHERE
29、NTRISKXCONTROLRISKWHEREACCEPTABLEAUDITRISKAARISTHEPROBABILITYTHATAUDITORSAREWILLINGTOACCEPTTHATTHEYWILLRENDERUNQUALIFIEDOPINIONSONMATERIALLYMISSTATEDFINANCIALSTATEMENTSAARISSETBYTHEAUDITORINHERENTRISKIRISTHEPROBABILITYTHATANACCOUNTBALANCEORCLASSOFTRANSACTIONSCONTAINSAMATERIALMISSTATEMENTBEFORECONSID
30、ERINGTHEEFFECTIVENESSOFTHEINTERNALCONTROLSYSTEMIRISASSESSEDBYTHEAUDITORCONTROLRISKCRISTHEPROBABILITYTHATAMATERIALMISSTATEMENTISNOTPREVENTEDORDETECTEDONATIMELYBASISBYTHEINTERNALCONTROLSYSTEMCRISASSESSEDBYTHEAUDITORDETECTIONRISKDRISTHETOLERABLELEVELOFRISKTHATAUDITINGPROCEDURESWILLNOTDETECTMATERIALMISS
31、TATEMENTSMATERIALITYISTHEMAGNITUDEOFANOMISSIONORMISSTATEMENTOFACCOUNTINGINFORMATIONTHAT,INLIGHTOFSURROUNDINGCIRCUMSTANCES,MAKESITPROBABLETHATTHEJUDGMENTOFAREASONABLEPERSONRELYINGONTHISINFORMATIONWOULDHAVEBEENCHANGEDORINFLUENCEDBYTHEOMISSIONORMISSTATEMENTBECAUSEAAR,IR,CRANDDRALLDEPENDUPONAPRESETLEVEL
32、OFMATERIALITY,MATERIALITYAFFECTSTHELEVELSOFALLOTHERELEMENTSINTHEARMACCORDINGTOTHEARM,AUDITORSSETMATERIALITYANDCHOOSEAARIRANDCRARETHENASSESSED,WHICHLEADSTOATOLERABLELEVELOFDRINVESTMENTSINAUDITINGDECREASEDR,ANDTHEPLANNEDLEVELOFAUDITINVESTMENTREFLECTSTHATPOINTWHEREATOLERABLELEVELOFDRISACHIEVEDAUDITPLAN
33、NINGMODELANDTHEEXPLANATIONOFAUDITFEESSIMUNIC1980DEVELOPEDAMODELTOEXPLAINAUDITFEES,LATERUSEDBYPRATTANDSTICE1994ASABASISFOREXAMININGAUDITPLANNINGDECISIONSWEMODIFYTHISMODELASFOLLOWSTOREPRESENTTHEEXPECTEDCOSTSOFAUDITING2ECCQEDXEREFXEPWHEREECTOTALEXPECTEDCOSTCTHEPERUNITFACTORCOSTOFEXTERNALAUDITORRESOURCE
34、STOTHEAUDITOR,INCLUDINGALLOPPORTUNITYCOSTSQTHEQUANTITYOFRESOURCESUTILIZEDBYTHEAUDITORINPERFORMINGTHEAUDITEXAMINATIONEDEXPECTEDPRESENTVALUEOFPOSSIBLEFUTURELOSSESINCURREDBYCLIENTSTAKEHOLDERSTHATAREASSOCIATEDWITHUNDETECTEDMATERIALMISSTATEMENTSINTHISPERIODSFINANCIALSTATEMENTSEREXPECTEDLIKELIHOODTHATTHEA
35、UDITORWILLBEHELDRESPONSIBLEFORCLIENTSTAKEHOLDERLOSSESASSOCIATEDWITHUNDETECTEDMATERIALMISSTATEMENTSINTHISPERIODSFINANCIALSTATEMENTSEFEXPECTEDPRESENTVALUEOFPOSSIBLEFUTURELOSSESCAUSEDBYBEINGIDENTIFIEDWITHTHISPERIODSFINANCIALSTATEMENTSDUETOFACTORSOTHERTHANUNDETECTEDMATERIALMISSTATEMENTSEPEXPECTEDLIKELIH
36、OODTHATTHEAUDITORWILLSUFFERLOSSESASARESULTOFBEINGIDENTIFIEDWITHTHISPERIODSFINANCIALSTATEMENTSBECAUSEOFFACTORSOTHERTHANUNDETECTEDMATERIALMISSTATEMENTSINTHEPLANNINGMODEL,EDXERANDEFXEPREPRESENTTHEEXPECTEDCOSTSOFBUSINESSRISKEDXERREPRESENTSCOSTSASSOCIATEDWITHUNDETECTEDMATERIALMISSTATEMENTSANDEFXEPREPRESE
37、NTSCOSTSASSOCIATEDWITHFACTORSOTHERTHANUNDETECTEDMATERIALMISSTATEMENTSEG,REPUTATIONLOSSESCAUSEDBYASSOCIATIONWITHCLIENTSOFQUESTIONABLEINTEGRITYACCORDINGTOTHEMODEL,AUDITSAREPLANNEDINTHEFOLLOWINGWAY1ASSESSTHELEVELOFBUSINESSRISKAND2INVESTINAUDITING,Q,TOTHEPOINTWHERETHEMARGINALREDUCTIONINTHECOSTSOFBUSINES
38、SRISKFROMANADDITIONALUNITOFAUDITINGISEQUALTOTHEMARGINALCOSTOFTHATUNITOFAUDITINGINEQUILIBRIUM,AUDITFEESARESETTOCOVERTHETOTALEXPECTEDCOSTSOFAUDITING,WHICHINCLUDEANORMALPROFITN1THEARMANDBUSINESSRISKTHEARMISDESIGNEDTOHELPAUDITORSMANAGERISKSASSOCIATEDWITHISSUINGUNQUALIFIEDOPINIONSONFINANCIALSTATEMENTSTHA
39、TCONTAINUNDETECTEDMATERIALMISSTATEMENTSCONSEQUENTLY,THEELEMENTSOFTHEARMCANBEADJUSTED,ANDTHEAUDITINVESTMENTINCREASED,TOREFLECTTHATPORTIONOFBUSINESSRISKASSOCIATEDWITHEDXERBRUMFIELDETAL1983MESSIER1997ARENSANDLOEBBECKE1997CHANGESINAUDITINVESTMENTSATTRIBUTABLETOCHANGESINAAR,IR,CRAND/ORMATERIALITYARECONSI
40、STENTWITHBOTHTHEARMANDCHANGESINBUSINESSRISKASSOCIATEDWITHEDXERINOTHERWORDS,MANYFACTORSTHATRELATETOTHELIKELIHOODOFMATERIALMISSTATEMENTSSIMILARLYAFFECTBUSINESSRISKEXAMPLESINCLUDEHIGHLEVELSOFACCOUNTSRECEIVABLEANDINVENTORY,ANDEXCESSIVESALESGROWTHSTICE1991BUSINESSRISKENCOMPASSESANADDITIONALFACTOR,EFXEP,T
41、HATISNOTASSOCIATEDWITHTHERISKOFUNDETECTEDMATERIALMISSTATEMENTSANDTHUSISNOTEXPLICITLYREFLECTEDINTHEARMPOORFINANCIALCONDITIONANDHIGHSTOCKPRICEVARIABILITY,FOREXAMPLE,HAVEBEENSHOWNTOINTRODUCEDIMENSIONSOFBUSINESSRISKTHATAREINDEPENDENTOFTHERISKOFUNDETECTEDMATERIALMISSTATEMENTSBRUMFIELDETAL1983STICE1991INT
42、HEPRESENCEOFTHESEFACTORS,AUDITSTANDARDSPROVIDELIMITEDGUIDANCESASNO47AICPA1997ASTATESONLYTHAT,INTHEPRESENCEOFLOWBUSINESSRISK,AUDITORSSHOULDNOTREDUCETHEAUDITINVESTMENTBELOWTHATIMPLIEDBYTHEARMTHISAMBIGUITYMAYEXPLAINWHYAUDITFIRMSSETRELATIVELYRIGIDGUIDELINESINTHEIMPLEMENTATIONOFTHEARM,PERHAPSINANATTEMPTT
43、OACHIEVECONSISTENCYACROSSAUDITSSTUDIESHAVESHOWN,FOREXAMPLE,THATAUDITFIRMSHAVEPOLICIESTHATSETAARATPREDETERMINEDLEVELSJIAMBALVOANDWALLER1984DANIEL1988KNECHEL1998RICCHIUTE1998,WHETHERAARISSETQUALITATIVELYEG,“LOW“ORQUANTITATIVELYEG,5PERCENTN2SOURCERICHARDWHOUSTON,MICHAELFPETER,JAMIEHPRATTTHEAUDITRISKMOD
44、EL,BUSINESSRISKANDAUDITPLANNINGDECISIONJTHEACCOUNTINGREVIEW,19993281298译文审计风险模式、业务的风险和审计规划的决定一、导言审计员表明,调整的因素包括审计风险模式ARM、增加投资的审计日益增加的风险、客户财务报表包含重大误报例如,JOYCE1976;GAUMNITZETAL1982;KAPLAN1985;LIBBYETAL1985;OKEEFEETAL1994,也有证据表明,审计业务风险损失由于客户增加投资审计,对收取的审计费用进行审计例如,PRATTANDSTICE1994;WALO1995。这项研究表明虽然审计员应当及时
45、对审计和业务的风险作出反应,但在未确定的条件下,审计风险具有很大的不确定性。在这种情况下,ARM这个标准的专业审计做法,能说明一定的审计行为。主要涉及的风险审计意见提供不合格的客户财务报表,带有重大误报。从另一方面来说,业务风险是目前审计员遵守的公认的审计标准GAAS以便提供适当的审计意见。相关费用的主要业务风险以及有关的诉讼,是否造成风险这些是审计员哟啊对客户的攸关利益负责的。其他费用主要涉及制裁的监管机构,其受损声誉以及收取的费用。业务风险比提供无保留意见的财务报表更加困难。如果一个客户内部控制系统遇到财政困难,例如,存在两种风险一种是重大错报风险,另一种是金融失败风险,包括业务的风险。A
46、RM是为了指导审计规划而服务的CUSHINGETAL1995;AICPA1997A。有时候,特别是当商业风险与其他风险并存的时候,ARM常常起到指导的作用。阿伦斯和勒贝克1997,259建议,无论审计员赞成或者不赞成,都应当接受商业风险中不同的审计风险。他们报道说,一些审计员认为可接受的审计风险能够降低目前的高水平的业务的风险,但是其他人都不赞同。我们认为,不同的性质的基本情况会引起商业和审计风险,这将导致不同审计员有不同的差别反应。在某些情况下,审计员将运用ARM来调整他们的审计计划以便适应周围的情况,但在其他情况下审计反应将不被逮捕的手臂。在前面的情况下,ARM将描述审计行为,后一种情况下
47、则不会。审计员要负责提供合理的保证,以确保财务报表无重大误报。然而,错报产生的两个可能来源错误无意误报和违规行为蓄意误报,如同我们的认为,ARM描述审计行为依据有关资源的。当大量的审计投资计划实施以及审计费用(审计费用包括溢价补偿)达到一定的程度时,我们对审计员提出的审计风险和业务风险的两方面进行监督。我们假设业务风险能够解释审计投资导致的不同差异,除那之外,一小部分人认为ARM比前者更好,尤其是当风险来自不规范的数据的时候。当对不规范的数据进行审计时,ARM不利于搜集业务风险,这是审计员在进行审计投资计划是应当考虑的。当审计风险很高而又未发现错误的时候,另一方面,有关业务风险的大小已经反映在
48、ARM中了。我们还预测,当违规风险错误很高的时候,审计费用还包括不含的风险溢价。当违规风险错误很高的时候,审计程序不能降低可容忍的风险水平,因此,增加了审计费用。相反,当错报的风险很高的时候,审计程序的标准足够可以降低审计可容忍的水平,审计员就不需要特别注意风险溢价因素。我们通过实验来测试这种假设,这个实验由34个审计合作伙伴和5大会计师事务所联合进行。根据每一个高风险情况虚报材料显示,进行初步分析和调查。两个案件中发现了一种错误或违规,从而暗示其他的风险,而未被发现错误或违规现象依然很高。审计员运用ARM对业务的风险审计之前和之后发现的虚报进行分析,并提出相关建议。这个实验表明,在审计方面,
49、投资计划的审计和审计取向是包括风险溢价在内的,其审计费用取决于不同风险的性质。其显著的差异以及投资带来的审计业务的风险是由于错误的条件所引起的;在正常情况下,企业风险控制需要进行投资审计。此外,审计费用不包括风险溢价这一错误条件只反映在成本的投资审计中;在不正常情况下,审计员提出的费用,事实上超过了所需的投资计划的审计。这些结果是否会为目前不规范的数据资料提供一个审计标准,有利于对其进行指导呢库欣等人1995年。但在审计员出现违规行为的时候,是不可以通过这个标准来帮助以便达到目标的。这个结果还符合梅农和威廉姆斯1994的观点,他门认为,重要的是,审计费用是反映审计员为客户提供审计服务的,特别是有风险的顾客。似乎包括风险溢价的审计费用进行审计时更高。事先研究彼此之间的关系在对投资和收费使用方法和档案进行审计分析是有用的。几个实验研究证实了结果,例如,已找到控制风险和审计投资之间固有的关系。最近,普拉特等人就有关客户特点、审计投资和提供证据的风险溢价以及业务的风险提出疑问。然而,审计员既不能通过ARM来衡量风险也不能通过上下之间的关系来辨别造成风险溢价的因素。通常在档案研究之前一般还不能找到控制风险和审计投资之间固有的关系例如,贝达德1989;模拟和赖特1993;奥基夫等人1994年;HACKENBRACK和KNECHEL1997年。这种关系可能在档案设置阶段很难确定