corporate conflicts of interesthomepages.ulb.ac.be/~plegros/documents/classes/finance/themes... ·...

Corporate Con icts of Interest

Joel S Demski

A con ict of interest arises when an executive an of ceholder or even anorganization encounters a situation where of cial action or in uence hasthe potential to bene t private interest Examples include the physician

who has a nancial interest in a diagnostic laboratory a congressman whose spouseis a lobbyist or corporate director or a professor with a nancial interest in aparticular textbook Empire building nepotism and all manner of in uence activ-ities offer other illustrations Such con ict is as old as history itself and as anhistorical matter it is important to acknowledge that con icts of interest can anddo morph into nancial fraud Consider a few examples from the fairly rich galleryof rogues and abuses over recent decades

The Equity Funding case is noteworthy for the large number of active partic-ipants the head of the rm with the assistance of other top managers andnumerous employees began falsifying insurance policies in 1965 which amongother steps involved company forgery parties where fake policies medical recordsand even death records were created Of course these fake policies brought in norevenue directly But they supported false revenue estimates that pumped up thecompany stock price allowing it to make acquisitions with its stock and the policieswere often resold to other reinsurance companies In turn manufactured deathclaims for ctitious policies that had been sold in the reinsurance market were partof the scam By the time 22 company employees were indicted in 1973 they hadrecorded 64000 false policies and had issued $25 million in counterfeit bondsmdashand $100 million of corporate assets was missing The companyrsquos auditors werefound guilty of fraud for failing to detect the scam and served prison time (Dirksand Gross 1974)

y Joel S Demski is Frederick E Fisher Eminent Scholar Fisher School of AccountingUniversity of Florida Gainesville Florida His e-mail address is joeldemskicbauedu

Journal of Economic PerspectivesmdashVolume 17 Number 2mdashSpring 2003mdashPages 51ndash72

The savings and loan crisis which began in the mid-1970s and lasted through-out the 1980s is noteworthy because it combines regulatory changes nancialfraud audit failures and a parade of con icts of interest that reached all the way tomembers of the US Senate (for one discussion see Pizzo Fricker and Muolo1989) It also ended up costing taxpayers in excess of $150 billion EricksonMayhew and Felix (2000) provide an in-depth analysis of one such audit failure theinfamous case of Lincoln Savings and Loan that ended up with a $2 billion price tagfor taxpayers The controlling owner of Lincoln Savings and Loan was found guiltyof fraudulent land transactions in state and federal court (although the federalconviction was later overturned on the grounds of mistaken jury instructions)while three major auditing rms reached out-of-court settlements totaling $135 mil-lion In 1989 the Senate Ethics Committee censured one senator and cited fourothers for questionable conduct because they had received substantial contribu-tions from the controlling owner of Lincoln Savings and Loan and had interferedwith federal regulators who were investigating the rm

In 1995 Barings Bank of the United Kingdom was driven into bankruptcy after233 years of operation because a 28 year-old trader in Singapore who was supposedto be searching for small-margin arbitrage opportunities lost over $1 billionspeculating with huge purchases of futures and option contracts on the Japanesestock market This case is noteworthy because of the sheer ineptness of Baringsrsquocontrols the rogue trader did his own accounting and reporting (Fay 1996)

The novelty then is not the presence of con icts of interest but theirmanagement Societies rely on various devices to manage these con icts Someactivities are prohibited such as an auditor engaged with an explicit pay-for-performance contract while at other times we rely on disclosure of relationshipssuch as my universityrsquos requirement that the faculty disclose any nancial interestin teaching materials Both prohibitions and disclosure requirements are but-tressed by legal requirements and remedies In between we seek arrangements thatat least dampen the underlying con ict as when the partner in charge of an auditis rotated or when the executiversquos performance is judged in part on the basis of nancial results that have been audited by an independent auditor In all of thesecases as with other economic activities we balance tensions and shy away fromattempting to create a situation of zero failures as a result of con icts of interestsimply because pursuing such an extreme goal would be uneconomic

I propose then to survey con icts of interest in the corporate arena withemphasis on recent events The presence of such con icts is not new We still haverogues Some use modern techniques as in Enronrsquos use of complex structured nance to bolster its apparent performance Others rely on the most mundane oftechniques as in WorldComrsquos capitalization of an expense serving to spread theexpense over many years instead of writing it off in total in the given year

Yet the seemingly steady drumbeat of recent corporate improprieties leads oneto wonder what is new in the corporate setting and what along the way to moderncommunication globalized organizations and markets and structured nancemight have been overlooked Surely complexity has increased and this in itself

52 Journal of Economic Perspectives

puts more strains on our management of various con icts Yet I fear we have failedto appreciate or have forgotten the delicacy of a well-crafted web of controls formanaging con icts of interest We tend to think in terms of a speci c con ict orspeci c control applied thereto for example assuring greater independence forauditors or more outsiders on boards of directors Yet reality is multiple con ictsamong multiple players in the context of an enlarged interactive web of controls(for example Fama and Jensen 1983 Sunder 1997) We have not investedadequately in understanding the role of multiple players and multiple con icts norin understanding how the interplay of shifting institutions and herding behaviormay act to unravel control systems

Recent Changes in the Institutional Background

The institutional environment for corporate governance has evolved substan-tially in recent decades Changes in regulations nancial innovation and the levelof business complexity have surely stressed or even in some cases disrupted thepre-existing web of controls over con icts of interest

Regulatory changes in the nancial sphere in recent decades are far fromminor Until the early 1970s brokerage houses charged xed commissions for eachstock market transaction and some of the revenue from these commissions wasdirected toward stock market analysts who did research on the rms In 1975 theSecurities and Exchange Commission (SEC) abolished xed commissions Theresult was shrinking pro t margins and industry consolidation along with theemergence of discount brokerages that charged for transactions but supportedlittle or no analyst research per se

This shifting of traditional patterns in the nancial industry was furtherenhanced by repeal of the Glass-Steagall Act of 1933 by the Gramm-Leach-Bliley Actof 1999 Glass-Steagall had restricted commercial banks from combining with other nancial services companies such as securities rms insurance companies andinvestment banks Gramm-Leach-Bliley permitted single holding companies tooffer banking insurance securities and other nancial services thus blurringdistinctions across the nancial sector (Barth Brumbaugh and Wilcox 2000)

The traditionally restrictive marketing practices in the audit industry have alsoseen substantial changes The American Institute of Certi ed Public Accountants(AICPA) is the primary professional organization of public accountants and itsCode of Professional Conduct must be observed by its members In 1988 to complywith a Federal Trade Commission consent order dealing with restraints on businesspractices the rules in the AICPArsquos Code of Ethics that prohibited soliciting anotheraudit rmrsquos client were eliminated Competitive bidding for audit services and evenadvertising became commonplace One result was a period of enhanced competi-tion among the large audit rms which saw no growth in domestic auditing feescoupled with dramatic growth in domestic consulting fees to the point thatdomestic audit fees declined to roughly 40 percent of total domestic revenue of the

Joel S Demski 53

large accounting rms This period also saw the major audit rms move fromgeneral partnership to limited liability partnership form to limit individual partnerliability1 Previts and Merino (1998) provide a history of the domestic accountingindustry

A number of non nancial industries also experienced major regulatorychanges including some industries with rms that ended up mired deep incon icts of interest For example the deregulation of natural gas markets dramat-ically altered the business climate for Enron and other energy rms whilethe deregulation of telecommunications shaped the business conditions forWorldCom

Two other factors have combined with the restructuring of nancial and otherindustries in a way that has placed additional stress on the corporate governancefunction greater complexity of business structures and greater emphasis on stockprices In the last decade or so business has experienced a surge of uid organi-zational arrangements as well as a routinization of complex transactions Alliancesjoint ventures multifaceted sale arrangements and hybrid structured nancearrangements have become commonplace The net effect is the economic bound-aries of the rm have become ambiguous and extremely uid a phenomenonre ected in the wonderfully euphemistic phrase ldquooff balance-sheet nancingrdquowhere the rm structures transactions and relationships to avoid their explicitrecognition in traditional accounting displays A typical example is a rm that holdsa portfolio of mortgages It places the portfolio in a free-standing legal entity withdistinctly limited scope a Special Purpose Entity but continues the transactionprocessing and possibly provides credit enhancements In different variationsinventory research and development or even rights to future revenue cash owsare parked in Special Purpose Entities (Hartgraves and Benston 2002)

Reporting regulations allow the Special Purpose Entity to be kept off of the rmrsquos formal nancial statements as long as it is disclosed provided substantiverisk has been shifted to an independent third party General Electric (an aggressivepurveyor of these arrangements) for example reports sponsored Special PurposeEntities with assets in excess of $50 billion in its 2001 nancial report Theldquoindependent third partyrdquo must have (among other things) a minimum of 3 per-cent ownership of the Special Purpose Entityrsquos equity and debt although theFinancial Accounting Standards Board (FASB) has recently tightened these rulesBut Special Purpose Entities are only one aspect of this wave of organizational and nancial innovation

This greater degree of complexity has interacted with a corporate governanceenvironment that has been placing heightened emphasis on shareholder value(Holmstrom and Kaplan 2001) including an explosion in the use of option-basedcompensation A substantial portion of the greater complexity appears to bemotivated by a concern for nancial presentation for example ldquobeautifyingrdquo onersquos

1 In addition the 1995 Private Securities Litigation Reform Act replaced the joint and several liabilityrule with a proportionate liability standard in dealing with violations of federal securities statues


balance sheet In some cases the effect may be as simple as a matter of timing forinstance the timing of selected expenditures and shipments can affect currentperiod nancial results just as can the time at which a sale is formally booked or aloan is consummated With the assistance of hybrid nancial and organizationaltransactions a lease can be structured so it does or does not show up on thelesseersquos balance sheet thereby affecting the total debt that a rm reports Enronfor example used a Special Purpose Entity organized by Citigroup to disguisesigni cant amounts of debt as commodity prepay transactions2 Through a series ofcircular or round-trip prepaid transactions this Special Purpose Entity was thecenterpiece in ldquoallowingrdquo Enron to borrow money but to record the amountborrowed as cash generated by operations because prepaid commodity contractsare generally booked as trades not loans (Sapsford and Beckett 2002)

Recent institutional changes then are far from benign and among otherthings have stressed existing arrangements for managing con icts of interestIndeed these institutional changes have outpaced our understanding of the sub-tleties of managing the ow of information within and across organizationalboundaries an understanding that is essential to managing properly the variousplayersrsquo con icts of interest

For example it is appealing and fashionable to call for transparent nancialreporting yet the underlying organizations and transactions have become astound-ingly complex and considerable judgment is required to measure a rmrsquos incomereasonably well in any given period not to mention its stock of assets and liabilitiesThis in turn increases the rmrsquos ability to manage its earnings to exercise thatreporting judgment opportunistically a theme explored by Lev (in this issue) Howthis sorts out depends on the various players in the corporate governance arenaand even then how we would prefer it play out is ambiguous Claiming that fullunfettered disclosure provides a solution is naive but where to draw the linedepends on a variety of con icts among a large number of players3

Players in Corporate Governance Arrangements

The players in the corporate governance game include auditors boards ofdirectors analysts and investment bankers regulators management and still others

2 I will often use Enron in the pages that follow to illustrate various con icts Though that saga continuesto unfold considerable documentation is available in Powers Troubh and Winokur (2002) the ldquoPowersReportrdquo Benston and Hartgraves (2002) Fusaro and Miller (2002) and various Congressional hearingsMcGill and Outslayrsquos (2002) analysis of Enronrsquos federal tax lings is also instructive3 To illustrate using accounting variables to manage earnings has the effect of reporting ldquolessrdquo than the rm knows of garbling the information that is made public relative to the rmrsquos full stock ofinformation In turn the mere fact that compensation contracts for managers can be renegotiatedcreates an interest in endowing the rm with an ability to garble its performance reports becausemanaging the ow of information is the key to ef ciently dealing with the renegotiation option See forexample Demski and Frimor (1999) and Christensen Demski and Frimor (2002) Arya Glover andSunder (1998) link earnings management to settings where the revelation principle is not applicabledue to limited commitment aggregate reporting or less than fully optimal contracts

Corporate Conicts of Interest 55

such as attorneys and investors Con icts of interest are unavoidable among theseparties the issue is how well these con icts are managed Yet the formal analysis ofthese situations is complex One would like to have a framework that incorporatesboth the interactions of the players and the institutional framework for managingtheir con icts of interest With clearly identi ed exogenous variables and endog-enous responses it would then be possible to examine how changes in institutionalarrangements might lead to different manifestations of con icts of interest But inreality both the institutions for managing con icts of interest and the responses tothose institutions are choices and must be regarded as endogenous variables Inaddition observation of the techniques for managing con icts of interest and theactual behavior responses by the players are inherently biased The strength of acontrol system resides in the threat of what might be reported or what might takeplace were off-equilibrium play to occur So to the extent the controls are func-tioning well we do not observe their full import Likewise when we do observetheir failure or their full import it is likely something has gone amiss and we aredealing with an out-of-equilibrium setting as the system copes with exogenousshock or inexorable change4 The discussion that follows will focus on identifyingrecent theory and evidence on key con icts of interest

AuditorsThe auditing industry has received considerable attention in the last few years

both as a bulwark against con icts of interest and as an industry subject to its owncon icts It often isnrsquot recognized how small this industry is relative to the corporatesector Total domestic accounting fees are about $6 billion per year as my col-league Karl Hackenbrack points out the US economy spends more on potatochips than on audit fees each year Similarly auto theft in the United States tops$6 billion annually and shoplifting (ldquoinventory shrinkagerdquo as it is called in politecircles) is routinely estimated at just under 2 percent of sales and US retail salestotal about $3 trillion

Auditors deal with a variety of reporting regulations that fall broadly intoauditing (generally accepted auditing standards or GAAS) and accounting (gen-erally accepted accounting principles or GAAP) categories The Securities andExchange Commission (SEC) in turn has historically relied on closely monitoredself-regulation with the FASB presently the primary source of generally acceptedaccounting principles and the AICPA through its Auditing Standards Board theprimary source of generally accepted auditing standards For a brief period theIndependence Standards Board was in place a roughly parallel arrangement tothat between the SEC and Financial Accounting Standards Board but focused

4 This is why information useful for valuation purposes is not necessarily useful for control purposes(Gjesdal 1982 Kim 1995) Christensen and Demski (2002) stress this possibility in documenting andmanaging the nancial reporting industry More broadly the concern for identifying carefully theendogenous variables and the nature of equilibrium versus off-equilibrium play is reminiscent ofKoopmansrsquo (1947) insistence that measurement be guided by theory


on auditor independence The Sarbanes-Oxley Act of 2002 establishes a PublicCompany Accounting Oversight Board under the auspices of the SEC with ex-plicit responsibility to regularize these various arrangements further alongwith the regulations themselves (for discussion see httpwwwfeiorgadvocacysarbanesoxleycfm )

Accounting rms in turn offer a variety of services broadly grouped intoauditing tax and management consulting Though the practicing auditor is re-quired to ldquobe independent in fact and appearancerdquo and to ldquoserve the publicinterestrdquo in these matters this arrangement has long been contentious A reporting rmrsquos management hires the auditor (usually subject to shareholder approval)Long-term relationships between accounting and client rms are common sale ofnonaudit services is routine and personnel often move from the accounting to theclient rm Indeed the ldquoengagement partnerrdquo the accounting rmrsquos lead orpartner-in-charge on a particular engagement is probably best thought of as a salesengineer with a keen interest in understanding and exploiting a multiplicity ofsales opportunities with the client Arthur Andersenrsquos Enron account for exampledelivered an average weekly billing of $1 million over half being for nonauditservices Andersen also housed a number of its staff in Enron facilities was a routinesupplier of accounting staff to Enron and counted numerous members of theEnron management team among its alumni

That said the auditor is required to be independent and is thus forbidden tohave any explicit economic interest in the client This prohibition even includesfor example stock holdings in a client rm by an accounting rm partner who isotherwise totally disconnected from that particular client Yet the possibility ofemployment by the client as well as the opportunity to sell nonaudit servicesarguably compromise this stark separation On the other hand scope economiesare often alleged to be crucial in understanding why this relationship persists andhas expanded for example understanding the clientrsquos business model and infor-mation system are essential to providing high-quality audit work and at the sametime provide a springboard to productive consulting services just as understandingthe clientrsquos business is essential to providing high-quality consulting service andthus provides a knowledge foundation for higher-quality audit work

Bazerman Morgan and Loewenstein (1997) and Bazerman Loewenstein andMoore (2002) argue that the very idea of auditor independence under currentarrangements is a myth that auditorsrsquo information processing and judgments arebiased away from the public interest simply because close af nity with the clientrenders the desired independence psychologically impossible Yet context alsomatters and accounting rms have developed elaborate internal structures topolice and manage this apparent con ict (even to the point many have becomeof cial AICPA quality control standards) For example it is routine (and required)to rotate a partner in charge on an SEC client engagement just as thorough reviewof an engagement is routine and a central authority inside the auditing rm isrelied upon to have nal word on reporting issues at the client level InexplicablyArthur Andersen weakened the authority of this central authority in its own rm

Joel S Demski 57

and allowed the Enron partner in charge to have nal say on a number of reportingissues5

Portions of the Sarbanes-Oxley Act of 2002 deal with this independence issueby prohibiting some activities such as bookkeeping or actuarial services andinternal audit outsourcing and allowing others provided they are preapproved bythe clientrsquos audit committee (The act also mandates that the audit committeeexplicitly manage the relationship between the rm and its auditor) Tax serviceremains acceptable presuming the necessary preapproval despite the fact this isclearly a client advocacy service and thus potentially in con ict with the publicinterest perspective one hopes is at work in an audit engagement

Explicit evidence on how well this con ict of interest is or has been managedis dif cult to come by because such evidence is not directly observable Someindirect evidence is provided by litigation and by SEC enforcement actions Forexample Heninger (2001) links litigation to deteriorating nancial condition andunusual patterns in the accounting accruals Bonner Palmrose and Young (1998)connect the ldquotyperdquo of fraud reported in an SEC enforcement action with subse-quent auditor litigation More broadly auditors are sued at the relatively low rateof about three cases per 1000 audits and SEC enforcement actions against auditorsoccur at a rate of about two cases per 10000 lings Also about 1 percent of these lings are restated (Elliott 2000) though their number has grown in recent years6

Alternatively survey data routinely document a concern for auditor independence(for example Swanger and Chewning 2001) and the contextual nature of thisconcern has been demonstrated in experimental settings (for example Hacken-brack and Nelson 1996) Studies of whether audits are unduly in uenced by clientpurchases of nonaudit services have found mixed results Frankel Johnson andNelson (2002) nd that unusual patterns in the reported nancial numbers them-selves at times appear to be associated with unusually large purchases of nonauditservices In contrast Antle et al (2002) simultaneously estimate audit fees non-audit fees and unusual patterns in the reported nancial numbers and nd no suchproblematic connection (They do however document a connection betweenaudit fees themselves and unusual patterns in the reported numbers)

5 Moving the nal authority closer to the client has the competitive advantages of faster client responseand in principle more client in uence It has the disadvantage of weakening the force of reputationand thus the audit rmrsquos control system as the central authority manages that reputation from the pointof view of the rm as a whole while the decentralized authority sees but a portion of the rm-wide issues6 Palmrose (2000) provides summary data on auditor litigation Also another possible source of indirectevidence on audit failure might be found in data on auditor replacement DeFond and Subramanyam(1998) for example link unusual patterns in the accruals to auditor replacement Divestiture ofconsulting groups and failure per se should also be noted In fact two large audit rms have failed inrecent years Arthur Andersen (following a series of high pro le audit failures and document shreddingassociated with the Enron investigation) and Laventhol amp Horwath (in 1990) Laventhol amp Horwath led for bankruptcy in the face of numerous lawsuits and at that time was the seventh largest accounting rm in the United States Ironically a major issue in its demise was Arthur Andersenrsquos claim that LampHrsquosperforming payroll services for the client in question led to a lack of independence yet a decade laterwe nd that Enron had outsourced its internal audit function to its auditor Arthur Andersen (Cren-shaw 1990)


The economy of scope issue though is central to understanding the strainbetween auditor independence and nonaudit services Here Antle et al (2002)also nd economies of scope in the auditor-client relationship a nding consistentwith a continuing tension between independence and breadth of services

Boards of DirectorsBoards of directors perform an important oversight function and thus take us

into corporate governance issues more broadly (for example Bushman and Smith2001 Shleifer and Vishny 1997) Important potential con icts here concernwhether the board as an entity is suf ciently independent of management andwhether it is suf ciently dedicated in pursuing its oversight responsibilities Oftenthis concern focuses on the ldquooutsiderdquo or ldquoindependentrdquo board members who haveno other direct connection to the rm But even an outside director is not totallyindependent for example the continued presence of an outside director on aboard depends on that directorrsquos behavior in the board room Interlocking outsidedirectors where top executives from two different rms serve on each otherrsquosboards are far from uncommon Moreover outside directors usually own sharesand options in the rm Equally clear the outside director is a part-time overseerand faces other demands on time and attention Again the question in such asetting is not whether con icts of interest exist but how well they are managed

The Conference Boardrsquos annual survey for example reports that the mediansize of a board is nine members that they meet about six times per year four hoursper meeting and that they receive basic compensation (excluding bene ts andstock compensation) ranging from about $10000 to $70000 depending on size ofthe rm and committee assignments (all for non nancial rms) Stock compensa-tion is also the norm with about 90 percent receiving some form of stock com-pensation usually in the form of options (Peck and Silvert 2001)

Hermalin and Weisbach (2002) review the governance issues associated withboards of directors As they emphasize the evidence is dif cult to interpret becausechoices about board governance are both interdependent and endogenous and willalso re ect a mix of equilibrium and out-of-equilibrium behavior With that caveatthe data are consistent with the claim that larger boards are less effective duepresumably to free rider issues In general issues with the independence of theboard do not jump out in their review of the data However Peasnell Pope andYoung (2002) report that less independent boards of UK rms do appear to beassociated with rms that display unusual patterns in their nancial data RelatedlyKlein (2002) reports lessened independence of the boardrsquos audit committee isassociated with more unusual patterns in the reported nancial data

A deeper issue here is the question of con ict with respect to what It iscommon to worry whether the board is doing an adequate job of protecting theshareholders (for example Shleifer and Vishny 1997) and a commonly proposedmechanism to address this potential con ict has been stock-based compensationfor outside directors However one can question whether short-term shareholdervalue is the appropriate goal of corporate governance or as Tirole (2001) stresses


the protection of other stakeholders should play a role In a mixed economy wesurely have concern for employees suppliers and so on But the very notion ofmeasuring value is problematic in a mixed economy indeed market value maxi-mization is not even guaranteed to be in the best interest of the shareholdersthemselves in a mixed economy simply because with limited markets the share-holders are not likely to be optimally insured (Ekern and Wilson 1974)

Enronrsquos board of directors exempli ed many of the potential dif culties It wasrelatively large (with 17 members) and used equity-based compensation for itsoutside directors This board knowingly allowed a member of senior managementto act as general partner in a venture doing business with Enron a clear con ict ofinterest that according to the boardrsquos Code of Conduct required explicit approvalTo help manage the con ict Enronrsquos board put in place a variety of controls aimedat monitoring transactions between Enron and the venture But there is no evi-dence the board subsequently investigated whether those controls were eitheradequate or even functioning Moreover the same interlocking pattern was appar-ently repeated in other transactions without the boardrsquos knowledge let aloneapproval It also turns out Enron employees who held partnership positions in therelated Special Purpose Entities collected unusual to say the least compensationfrom their partnership positions While it appears that Enron management with-held considerable information from its board I can nd no evidence this possibilitywas ever of concern among the individuals on the board Of course while a boardof directors can be more or less diligent in demanding information from manage-ment it may ultimately have little defense against a management that is committedto subterfuge

Re ecting provisions of the Sarbanes-Oxley Act of 2002 the New York StockExchange enacted new corporate governance standards in August 2002 that re-quire independent directors to comprise a majority of the board (as opposed toearlier independence requirements that applied only to members of the auditcommittee) and also ban performance-based compensation for those on the auditcommittee The standards also call for the rm to have an explicit policy prohib-iting con icts of interest related to say improper personal bene ts that accrue toone as a result of their position in the rm

Analysts Brokers and Investment BankersAnalysts report on and forecast earnings for selected rms Brokers play a

more direct role in trading activities Investment bankers provide services indesigning and oating securities Con icts of interest can arise in how these partiesrelate to each other After all the brokerage rm bene ts from trading recom-mended by the analyst just as investment bankers and their clients bene t fromfavorable coverage by the analyst

The parallels between a full service nancial services rm and a full serviceaccounting rm are striking (Schipper 1991) In both cases the issue centers onthe information functionmdashwhether auditing or analyst workmdash becoming entangledwith other parts of the rm For example might the prospect of a rm selling


investment banking services to a client cloud the objectivity of its analyst who israting the same client Just as in the case with auditors numerous regulationsare in place such as a mandatory ldquoquiet periodrdquo when an initial public offeringof securities is being conducted during which the investment bankrsquos analystsare precluded from issuing new forecasts or recommendations for a client ldquoinregistrationrdquo

A number of studies have examined analystsrsquo forecasts of corporate earningscontrolling for different effects One general nding is that analystsrsquo forecasts areupward biased though still more accurate than forecasts derived from simple timeseries models (for example OrsquoBrien 1988 Abarbanell 1991) Analyst recommen-dations are also typically skewed toward the ldquostrong buyrdquo and ldquobuyrdquo categoriesrather than to ldquoholdrdquo or ldquosellrdquo For example almost 61 percent of the recommen-dations in Lin and McNicholsrsquo (1998) sample of seasoned equity offerings andalmost 96 percent of those in Bradley Jordan and Ritterrsquos (2003) sample of recentinitial public offerings are in the ldquostrong buyrdquo or ldquobuyrdquo categories Censorship alsoappears to be present when an analyst begins to cover an additional rm theassociated recommendation tends to be high while poorly performing rms tendto be either not added to the list in the rst place or simply removed from that list(McNichols and OrsquoBrien 1997)7 In addition analysts routinely follow rms withwhom their rm has an investment banking relationship and when that relation-ship is present they tend to issue more favorable growth forecasts and recommen-dations (Lin and McNichols 1998) Moreover an important reason for switchingunderwriters appears to be access to coverage by a ldquostarrdquo analyst (Krigman Shawand Womack 2001)

Digging deeper into the institutional fabric we also nd potential for con ictof interest in the reputation and career concerns for investment analysts (fordiscussion see Hong Kubik and Solomon 2000 Li 2002 Michaely and Womack1999) For example the annual Institutional Investor poll based on opinions ofvarious money managers and institutions appears to be a well-watched form ofrelative performance evaluation for investment analysts

Recent regulatory changes beginning with the Sarbanes-Oxley Act of 2002 andextending to rule changes by the National Association of Securities Dealers imposelimits on communication between a rmrsquos investment banking and research groupThey also impose new disclosure requirements forbid analyst compensation beingtied to explicit investment banking transactions and also forbid offering a favor-able research rating while soliciting a client The National Association of SecuritiesDealers is a self-regulating organization under the auspices of the SEC thatregisters governs monitors and disciplines virtually all securities rms doingbusiness in the domestic economy

7 Evidence based on market reactions to these recommendations is mixed Michaely and Womack(1999) for example nd less reaction to a lead underwriterrsquos positive recommendation while Lin andMcNichols (1998) nd similar reactions to lead and nonlead underwriter recommendations

Joel S Demski 61

RegulatorsRegulatory institutions provide an additional set of players and potential

con icts The Securities and Exchange Commission which has statutory authorityto prescribe nancial reporting deals with over 170000 reporting companies700000 brokers and 8000 brokerage rms and the fees it collects vastly exceed itsown expenditures even with the recently legislated increase in its operating bud-get Although the SEC exercises considerable oversight the task of setting nancialreporting standards has been delegated to the seven-member FASB which in turnrelies on a 40-person staff The Financial Accounting Foundation (FAF) exercisesoversight over the FASB appoints its members and provides its nancial supportthough the Sarbanes-Oxley Act of 2002 calls for user fees collected by the newPublic Company Accounting Oversight Board to fund any board whose standardsare to be recognized by the SEC There are also more focused subgroups within theFASB For example the Emerging Issues Task Force (EITF) is an interpretive bodythat supplies authoritative answers to reporting issues under the auspices of theFASB and as such serves as a stopgap between practice and the FASB itself Forexample the rule that a Special Purpose Entity needed to have 3 percent outsideownership arose from an EITF ruling (with SEC concurrence) and gradually spreadin use beyond the original ruling

For all the talk of ldquodelegatingrdquo authority to the SEC or how the SEC delegatesauthority to the FASB or AICPA congressional involvement in their decisionsexplicit and implicit is a routine event (Zeff 2002) The most recent examplementioned earlier is the Sarbanes-Oxley Act of 2002 that establishes an ldquoindepen-dentrdquo board between the SEC and the audit industrymdashthough regulation of bothreporting and auditing activities were and remain the province of the SEC But theagenda and priorities of nancial regulators are often in uenced by politicalpressures For example politics may well explain why the SEC chose not to vetvarious Enron lings despite the rmrsquos growth glamour status and its unprece-dented use of structured nancial transactions Several years ago the US Senateplayed an important role in discouraging the FASB from passing a rule that wouldrequire treating stock options as an expense now key senators have reversed theirpositions and the FASB may well enact such a rule Johnson and Swieringa (1996)provide a fascinating glimpse into the FASBrsquos extensive due process and attendantlobbying including that by the Federal Reserve

Regulators face several interrelated con icts of interest One is that regulatorsrely on industry for both information and often for future job opportunitiesthough FASB members are typically appointed late in their careers Moreover loudindustry complaint may make a regulatorrsquos position politically untenable FASBmembers for example are drawn from professional accounting nance industryand academia Balancing the need for their technical skill with a broader socialperspective is an ever-present concern In fact the SEC recently orchestrated anenlarged public interest membership in the FASBrsquos oversight body the FinancialAccounting Foundation (Zeff 1998) The FAF has also altered its trustee appoint-ment procedures in an attempt to increase its independence


Much of the recent concern over nancial regulation has centered onaccounting standards especially dealing with the Special Purpose Entities thathave allowed rms to move debt and risk off their books or whether stockoptions should be treated as a current-year expense Thus attention has focusedon the FASB and two general criticisms have surfaced slowness in reachingdecisions with some agenda items lasting over a decade and overly detailedcomplex reporting standards Both traits might be interpreted in terms ofregulatory capture Slowness may arise where various interests are successful inblocking a particular reporting requirement The complexity of FASB rulesresults from a high level of detail and various exceptions and with the verycomplexity they implicitly provide (desired) guidance on how to structuretransactions to achieve some particular reporting objective While the FASB ispresently considering a move toward less detailed standards and thus moreresponsibility for the reporting rms and their auditor the reporting industryand the large rms who are their clients often clamor for detailed rules Forexample the aforementioned EITF is heavily in uenced by large rms andoffers a steady supply of detailed guidance and only recently has the FASBdecided to sign off explicitly on EITF decisions Consolidated databases of thevarious rules and regulations are regarded as proprietary and two-thirds of theFASBrsquos budget is currently nanced by sale of its own publications

This slowness and complexity might also be interpreted as re ecting a self-preservation motive The FASB and its staff exist at the sufferance of the SEC Manyof those involved with the FASB have a strong desire to retain the regulatoryfunction in the ldquoprivate sectorrdquo and bold moves I suspect are tempered by thisself-preservation concern

Recently the FASB has encountered the threat of entry with the emer-gence of the London-based International Accounting Standards Board or IASB(Dye and Sunder 2001) The IASB though similarly structured but with a muchlarger board is focused on reporting standards on a global basis and its standardsmust be followed by for example European Union members (beginning in 2005)The two boards are coordinating various projects with the hope that a single set ofstandards will eventually emerge So-called ldquoharmonizationrdquo would increase com-parability on a global basis To date however the IASB offers reporting require-ments of a more general nature so-called ldquoprinciples-based standardsrdquo

The issues here however ultimately come down to independence By analogyimportant data sources in the economy such as cost-of-living indices or economicstatistics from the Bureau of Economic Analysis are managed by professionals whoare reasonably well shielded from heavy lobbying or political intervention Man-agement of corporate reporting is much less independent and the Sarbanes-OxleyAct of 2002 does not offer improvement on this score

ManagementManagement of course is the nexus of corporate con icts of interest We

worry about shareholders debtholders employees customers suppliers stark


self-interest and so on We also must contend with the usual litany of ldquoconsumptionat workrdquo or ldquovalue diversionrdquo as exempli ed by empire building career concernspower politics and prestige Into this usual milieu the recent debates over corpo-rate governance add concern for managing the rmrsquos risk pro le marketing the rmrsquos securities including issues of opportunistic disclosures and market expecta-tions of earnings and rationalizing executive compensation

The compensation numbers themselves are eye catching Data from Standardamp Poorrsquos Compustatrsquos ldquoExecuComprdquo database (based on proxy statement data fortop executives in nearly 2000 companies) indicate median total executive com-pensation has quadrupled in the last decade while that of the largest rms hasincreased nearly eightfold thanks in part to heavy use of stock options in compen-sation packages and the robust stock market during this period In additionmedian total compensation is about $2 million in 2001 though much larger forlarge rms and options now dwarf either base salary or bonus as a source ofexecutive compensation In turn Conyon and Murphy (2000) report compensa-tion for US chief executive of cers is nearly double that of their UK counter-parts with the bulk of the difference attributable to share option grants

Understanding these patterns and documenting the possible role played bycon icts of interest is a dif cult and unsettled task For example are the datare ective of carefully designed ef cient incentive programs that are carefullymanaged by the compensation committee of the board of directors or are theyre ective of a compensation function that has been captured by managementAgain both the methods chosen to address managerial self-interest and the ways inwhich those interests are expressed are endogenous and the behaviors and nan-cial outcomes observed will be a mix of equilibrium and off-equilibrium outcomesTo this ever-present list we also encounter serious measurement error issues hereWhile theory stresses the connection between ldquopayrdquo and ldquoperformancerdquo neither isreadily observable Performance for example includes private information in thehands of the compensation committee Executive pay comes in a wide variety offorms at a variety of times (such as deferred cash compensation and retirementperquisites) and involves complex issues of risk from a market perspective from theexecutiversquos perspective and from the perspective of the incentives that the com-pensation provides Varying adjustments for these risks can produce remarkablydifferent patterns in the data (Abowd and Kaplan 1999 Hall and Liebman 1998Hall and Murphy 2002)

Tax issues also enter For example current tax law disallows a rm fromdeducting compensation in excess of $1 million at least if performance goals aredetermined by a compensation committee that includes inside directors Presum-ably this rule invites substitution from salary to performance-based pay like stock-options but for a variety of reasons the data are ambiguous (Rose and Wolfram2000) As one example Enronrsquos chief executive of cer received total compensationin 2000 in excess of $140 million and his base salary was slightly in excess of$1 million with the amount above the $1 million deductibility cap deferred

The primary focus of research on managerial con icts of interest has been the


possibilities for managers to use their positions to enrich themselves This mayhappen in a number of interrelated ways through a not-very-independent com-pensation committee on the rmrsquos board of directors through opportunistic use ofoption instruments through overemphasis on personal career enhancement orthrough the manner in which rmrsquos prospects are marketed through forecasts proforma announcements earnings management and the timing of disclosures8

The evidence itself is mixed Consider the use of options as one exampleFocusing on oil and gas rms where we expect risk to be a major issue Rajgopaland Shevlin (2002) document a positive relationship between exploration risk anduse of employee stock options in the compensation package a nding consistentwith the idea that options play a useful idiosyncratic role in connecting the risksperceived by managers and shareholders On the other hand Aboody and Kasznik(2000) provide evidence consistent with management opportunistically timingdisclosures to in uence market expectations around the time of stock optionawards Alternatively earnings management suggests self-serving manipulation ofreported results (Lev this issue) yet measurement issues exactly what we mean byearnings management and endogeneity issues cloud the documentation

Despite extensive work on career concerns turnover incentive intensity rel-ative performance evaluation shareholder expropriation and so on the patternsand trends in executive compensation and the documentation of potential conicts of interest and their management continue to offer many open issues Abowdand Kaplan (1999) Bushman and Smith (2001) Lambert (2001) and Prendergast(1999) provide recent reviews and assessments of this literature

Other PlayersOther players in corporate governance beyond those already named include

management consultants attorneys employees investors and academics Consult-ants including compensation consultants and attorneys have presumably aninterest in maintaining and expanding their relationship with the buying rm andthis depends on keeping the rmrsquos management happy and maintaining theperceived nancial health of the customer rm A rmrsquos employees may oftenshare with its managers an incentive to pump up short-term nancial performanceFor that matter numerous accounting professors bene t from the largess ofaccounting rms holding endowed chairs and bene ting from departmental slushfunds supplied by major accounting rms

The Enron debacle provides a glimpse into how this web of corporate gover-nance with multiple players can go off track Enron carried out countless highlycomplex and carefully crafted nancial transactions These all involved selling ofadditional nancial services by consultants attorneys and investment banks In

8 We also have concern for the ldquolevelrdquo of incentive intensity for example the connection betweenchanges in wealth of shareholders and of chief executive of cers It is also important to remembercompensation consultants are active players in the larger governance game and that regulations are farfrom minor for example see the analysis of value diversion statutes in Bebchuk and Jolls (1999)

Joel S Demski 65

many cases these transactions were designed with no apparent purpose other thanmanipulating recorded debt and earnings and often provided an opportunity for a nancial institution to collect fees on both sides of a transaction The ability tocontinue to sell these services rested on the reputation and prestige of all theparties involved that is these various transactions re ected the joint work ofmanagement the board the accountants the attorneys and the nancial institu-tions (Healy and Palepu this issue) Arthur Andersen had scores of staff workingfull time at Enron and ldquoopinion lettersrdquo provided by Andersen were critical inlegitimizing the various Special Purpose Entity transactions Many Enron employ-ees must have had some personal evidence on what the rm was doing yet manyof these same Enron employees also chose to skew their 401(k) holdings and ridethe stock price bubble (testimony of the Comptroller General before the SenateFinance Committee February 27 2002) Enronrsquos outside council Vinson amp Elkinswas paid in excess of $30 million in 2001 roughly 7 percent of its total revenues (ascompared to Enronrsquos payments to Andersen being less than 1 percent of totalAndersen revenues) Commercial lenders were also involved in providing nancein many cases

Both institutional and individual investors were also players The Californiapension retirement system Calpers for example simultaneously held Enron sharesand positions in at least one of its off-balance sheet ventures which could easily betaken as an implicit endorsement that the off-balance venture was a reason-able institutional innovation from the point of view of a sizeable shareholderLikewise increased use of the Internet by various individual investors accessinganalyst reports and trading aggressively arguably reinforced less attention tofundamentals

In short any attempt to blame the Enron meltdown solely on secretive or evenfraudulent behavior by a handful of top Enron and Arthur Anderson executivesdoes not hold water Surely deceit and obfuscation were in play Yet just as surelythe breadth of Enronrsquos shortcomings and nancial obfuscations was known bymore than a select few Certainly many of Enronrsquos activities were almost indescrib-ably complex yet they were shrouded in a plethora of hints such as the complexityitself the pure number of Special Purpose Entities the lavish management com-pensation and the arrogance of its visible management Yet no one wanted to knowIt was as if a cult of Enron had emerged one that believed in hype growth the neweconomy and that taking items off a balance sheet could make shareholders rich

Herding in a Market with Multiple Players

Were the case of Enron a singular anecdote it could be led on the list ofinevitable occasional failures that attend any system for balancing con ict ofinterest concerns After all it contains many familiar elements being a story with alarge number of players as in the Equity funding story less-than-aggressive con-trols as in the Barings story and regulatory changes as in the savings and loan


story that ends in disaster But the larger picture is a sea of possible con icts ofinterest combined with numerous checks and balances checks and balances thatfailed in a number of other cases which brings a suspicion that we are not dealingwith an isolated case or two but rather with a systemic problem9 Moreover thissystemic problem may involve an element of contagion where poor businesspractices spread to otherwise healthy rms For example Enronrsquos so-called ldquopre-payrdquo transactions which were structured to make a loan appear like a sale weremarketed by the investment banker to other customers When one rm usedaggressive accounting others felt pressure to use the same techniques or to inventeven more aggressive approaches of their own

In this setting the theory of herding is a natural lens to apply It has been usedin structuring our understanding of such diverse phenomena as clustering bycompetitors crime medicine valuation newsletters forecasts bank runs andexecutive compensation10 At the risk of herding on the use of herding modelsherding has something to teach us about corporate con icts of interest (Hirshleiferand Teoh 2001) The basic idea in a herding model is learning from the behaviorof others learning that leads to coordinated behavior or clustering actors actsequentially and their actions are observed and interpreted by those who followEmulating others thereby substitutes for acquiring and analyzing information onprivate account In this sense emulation substitutes for circumspection

As a simple example of herding imagine that a series of people privately andsequentially observe the ip of a potentially biased coin and then announce to agroup based on both the prior group announcements and their own privateobservation whether in their opinion the coin is more likely to be biased towardheads or tails If several people have announced that the coin is biased towardheads then even if you privately observe a ip of ldquotailsrdquo the string of priorannouncements favoring heads will outweigh your individual observation Ofcourse your announcement that a bias toward heads is likely even though youobserved tails will in turn cause others who observe ldquotailsrdquo to announce to thegroup that ldquoheadsrdquo is a more likely bias too Hans Christian Andersenrsquos fable ldquoTheEmperorrsquos New Clothesrdquo is perhaps the classic fable about herding First everyone(including the emperor) observes that everyone else is praising the emperorrsquos newclothes and emulates that praise rather than believing their own eyes But when one

9 I also sense an episodic if not cyclical pattern in these types of events For example the 1977 ForeignCorrupt Practices Act was passed in response to corporate behavior concerns in an earlier era Amongother things it introduced an explicit internal control requirement that may well have led to substitu-tion of internal for external audit services (Previts and Merino 1998)mdashin effect rebalancing the mix ofclient and audit rm personnel more in the direction of client personnel performing the overall auditfunction And the Sarbanes-Oxley Act of 2002 mandates an audited internal control report now beincluded in the annual report10 For an introduction to the theory of herding in this journal see Bikhchandani Hirshleifer and Welch(1998) For other useful introductions see Hirshleifer and Teoh (2001) and Devenow and Welch(1996) Evolutionary analysis provides a complementary if not parallel model of these events (forexample Ania Troger and Wambach 2002)


person speaks the truth that the emperor has no clothes at all everyone immedi-ately switches positions

Thus if it is known that several players have tested the waters with an innova-tive highly complex structured nance transaction and concluded it is ldquowithin therulesrdquo and provides bene ts to the rmrsquos shareholders a subsequent player maywell follow suit After all the earlier choices reveal information that those who wentbefore apparently concluded the transaction was appropriate Moreover it takesenviable sophistication to design and execute these transactions so emulationconveys sophistication and is far easier than inventing a new transaction fromscratch Diverging from prior choices has the effect of rejecting the informationthey carry as well as signaling the requisite sophistication may be in short supplyAs a result the choices of followers cluster together and without the bene t ofadded scrutiny

A web of corporate governance arrangements with multiple parties may exac-erbate and reinforce this pattern of herding For example it is not only a situationof investment bankers learning from a transaction pioneered by other investmentbankers The learning and herding will also take place among analysts fundmanagers commercial banks and investors in their interpretations of these trans-actions as well as by business schools and accounting programs The multipleparties may cause the herding effect to spread more quickly and to be reinforcedmore powerfully but not to be examined any more carefully

Putting together a model of herding with multiple parties and a web ofinterlocking devices for managing con icts of interest has two important implica-tions First as is usual with these types of models fragility is an issue This fragilityis displayed in the declining stock market and its effects on certain large rms suchas Enron in recent years11 But the fragility also refers to how institutional arrange-ments may shift rapidly Once the stock market bubble burst we witnessed astampede to alter corporate reporting requirements executive compensation andthe regulatory apparatus itself Even if this rearrangement of corporate governanceinstitutions addresses some existing con icts of interest it may prove in a few yearsunder different stresses to be just as fragile as the previous set of arrangements hasproven to be A second implication is that the convergence on behavior anddownplaying of local information as a cascade developsmdashfor example the aggres-sive use of Special Purpose Entities and their nancial interpretationmdashleads un-wittingly to an upward-biased assessment of the control systemrsquos effectiveness Thisconclusion is partly driven by the inef cient behavior associated with a herdingequilibrium which is not being detected by the control system But the situation isarguably made worse because herding can mean that the web of monitors and

11 For that matter Enronrsquos implosion was accelerated by the fact it used its own stock as a ldquocreditenhancementrdquo or guarantee in various Special Purpose Entities so when the stock price weakenedfollowing a restatement of an SEC ling the bank run was on so to speak It was also forced to restateits 1997ndash2000 nancial reports because some of its Special Purpose Entities did not warrant off balancesheet treatment in the rst place


controls is dependent rather than independent This argument is a variation onHarrisonrsquos (1977) identi cation of calibration error in reliability assessmentsSuppose a transaction must be vetted by three gatekeepers call them the accoun-tant the attorney and the investment banker Further suppose each has an errorrate of 5 or 15 percent with equal probability If the gatekeepers are statisticallyindependent the overall failure rate where a poor transaction is not spotted is 1 in1000 (that is the expected value of the error rate is 10 percent at each level and0103 5 001 overall) But if the error rates are dependent as would happen in acascade the expected failure rate is 75 percent higher (that is the error rate has a50-50 chance of being either 0053 5 000125 or 0153 5 003375 and the averageof these two is 00175) This higher error rate is driven by the uninvited coordina-tion among the partiesrsquo assessments

Thus I interpret the Enron saga and other recent failures of corporategovernance in terms of a governance system with multiple parties unavoidablecon icts of interest a web of controls for addressing these con icts of interest andherding issues that inadvertently tax and disrupt this web of controls Howeversuch a framework begs a number of questions For example can empirical methodsdifferentiate this explanation from competing interpretations such as an unex-pected group of random rogues What are the potential public policy implicationsof this perspective such as the role of disclosure regulations increased indepen-dence requirements or institutional training programs The task of putting to-gether and exploring such a model largely remains to be accomplished but it doeshighlight the key facts that corporate con icts of interest are widespread that theyare managed with an interlinked web of control arrangements and that the systemis susceptible to a correlated failure of those arrangements

Conclusion

Con icts of interest in the corporate arena are neither new nor avoidable Theissue is how they are managed As a society we rely on an elaborate web of controldevices and we are long experienced in the importance of re ning and redirectingthese devices as relationships and technology change We should also be accus-tomed to the reality that these control mechanisms involve striking various balancesand thus will never be perfectly effective Yet our understanding of the largerpicture is limited

The shallowness in our understanding of these phenomena is in my minddriven by our lack of investment in adequately understanding the role of multipleplayers multiple con icts and potential cascades in unraveling control systems Tobe sure we have examined certain combinations of controls (like board structureand turnover of the chief executive of cer) as well as substitutes for formalcontrols such as socialization But we have not invested adequately in understand-ing the role of multiple players multiple con icts and potential cascades inunraveling control systems We know fragility is part of the story after all the

Joel S Demski 69

current gallery of corporate governance rogues was identi ed only after the eco-nomic downturn took place We know rogues may bene t from temporary voids inthe web of controls We know empirical assessment is unusually dif cult in this areabecause of endogeneity and off-equilibrium issues But we do not understand howthe pieces t together

y Helpful comments by Rick Antle Brad De Long George Drymiotes John Fellingham HansFrimor Karl Hackenbrack Bill Kinney Baruch Lev Maureen McNichols David Sapping-ton Geoff Sprinkle Mary Stone Timothy Taylor Tom Tallarini and Michael Waldman aregratefully acknowledged

References

Abarbanell Jeffery 1991 ldquoDo Analystsrsquo Earn-ings Forecasts Incorporate Information in PriorStock Price Changesrdquo Journal of Accounting andEconomics June 142 pp 147ndash65

Aboody David and Ron Kasznik 2000 ldquoCEOStock Option Awards and the Timing of Corpo-rate Voluntary Disclosuresrdquo Journal of Accountingand Economics February 291 pp 73ndash100

Abowd John and David Kaplan 1999 ldquoExec-utive Compensation Six Questions that NeedAnsweringrdquo Journal of Economic Perspectives Fall134 pp 145ndash68

Ania Ana Thomas Troger and Achim Wam-bach 2002 ldquoAn Evolutionary Analysis of Insur-ance Markets with Adverse Selectionrdquo Games andEconomic Behavior August 402 pp 153ndash84

Antle Rick et al 2002 ldquoThe Joint Determina-tion of Audit Fees Non-Audit Fees and Abnor-mal Accrualsrdquo Unpublished working paper YaleUniversity

Arya Anil Jonathan Glover and Shyam Sun-der 1998 ldquoEarnings Management and the Rev-elation Principlerdquo Review of Accounting Studies31 2 pp 7ndash34

Barth James R R Dan Brumbaugh Jr andJames A Wilcox 2000 ldquoPolicy Watch The Re-peal of Glass-Steagall and the Advent of BroadBankingrdquo Journal of Economic Perspectives Spring142 pp 191ndash204

Bazerman Max George Loewenstein andDon Moore 2002 ldquoWhy Good Accountants DoBad Auditsrdquo Harvard Business Review November8011 pp 96 ndash102

Bazerman Max Kimberly Morgan and

George Loewenstein1997 ldquoThe Impossibility ofAuditor Independencerdquo Sloan Management Re-view Summer 38 pp 89 ndash94

Bebchuk Lucian and Christine Jolls 1999ldquoManagerial Value Diversion and ShareholderWealthrdquo Journal of Law Economics and Organiza-tion July 152 pp 487ndash502

Benston George and Al Hartgraves 2002ldquoEnron What Happened and What We CanLearn From Itrdquo Journal of Accounting and PublicPolicy Summer 212 pp 105ndash27

Bikhchandani Sushil David Hirshleifer andIvo Welch 1998 ldquoLearning From the Behaviorof Others Conformity Fads and InformationalCascadesrdquo Journal of Economic Perspectives Sum-mer 123 pp 151ndash70

Bonner Sarah Zoe-Vonna Palmrose and Su-san Young 1998 ldquoFraud Type and Auditor Liti-gation An Analysis of SEC Accounting and Au-diting Enforcement Releasesrdquo Accounting ReviewOctober 734 pp 503ndash32

Bradley Daniel Bradford Jordan and Jay Rit-ter 2003 ldquoThe Quiet Period Goes Out with aBangrdquo Journal of Finance February 581 pp1ndash36

Bushman Robert and Abbie Smith 2001 ldquoFi-nancial Accounting Information and CorporateGovernancerdquo Journal of Accounting and EconomicsDecember 321ndash3 pp 237ndash333

Christensen John and Joel Demski 2002 Ac-counting Theory An Information Content PerspectiveBoston McGraw-Hill Irwin

Christensen Peter Joel Demski and Hans Fri-mor 2002 ldquoAccounting Policies in Agencies


with Moral Hazard and Renegotiationrdquo Journalof Accounting Research September 401 pp1071ndash090

Conyon Martin and Kevin Murphy 2000ldquoThe Prince and the Pauper CEO Pay in theUnited States and United Kingdomrdquo EconomicJournal November 110 pp F640ndashF671

Crenshaw Albert 1990 ldquoFirmrsquos Last DayHugs Wry Smiles Laventhol amp Horwath ClosesUp Shopsrdquo The Washington Post November 21 pE1

DeFond Mark and K R Subramanyam 1998ldquoAuditor Changes and Discretionary AccrualsrdquoJournal of Accounting and Economics February251 pp 35ndash67

Demski Joel and Hans Frimor 1999 ldquoPerfor-mance Measure Garbling Under Renegotiationin Multi-Period Agenciesrdquo Journal of AccountingResearch 37 Supplement pp 187ndash214

Devenow Andrea and Ivo Welch 1996 ldquoRa-tional Herding in Financial Economicsrdquo Euro-pean Economic Review April 403ndash5 pp 603ndash15

Dirks Raymond and Leonard Gross 1974The Great Wall Street Scandal New York McGraw-Hill

Dye Ron and Shyam Sunder 2001 ldquoWhy NotAllow FASB and IASB Standards to Compete inThe USrdquo Accounting Horizons September 153pp 257ndash72

Ekern Steiner and Robert Wilson 1974 ldquoOnthe Theory of the Firm in an Economy withIncomplete Marketsrdquo Bell Journal of EconomicsSpring 51 pp 171ndash80

Elliott Robert 2000 Testimony Before theSecurities and Exchange Commission Septem-ber 21

Erickson Merle Brian W Mayhew and Wil-liam I Felix Jr 2000 ldquoWhy Do Audits Fail Evi-dence From Lincoln Savings and Loanrdquo Journalof Accounting Research Spring 381 pp 165ndash94

Fama Eugene and Michael Jensen 1983ldquoAgency Problems and Residual Claimsrdquo Journalof Law amp Economics June 262 pp 327ndash50

Fay Stephen 1996 The Collapse of BaringsLondon Richard Cohen Books

Frankel Richard Marilyn Johnson and KarenNelson 2002 ldquoThe Relation Between AuditorsrsquoFees for Non-Audit Services and Earnings Qual-ityrdquo Accounting Review Supplement 77 pp 71ndash105

Fusaro Peter and Ross Miller 2002 WhatWent Wrong at Enron Hoboken NJ Wiley

Gjesdal Froystein 1982 ldquoInformation and In-centives The Agency Information Problemrdquo Re-view of Economic Studies July 493 pp 373ndash90

Hackenbrack Karl and Mark Nelson 1996ldquoAuditorsrsquo Incentives and Their Application of

Financial Accounting Standardsrdquo Accounting Re-view January 711 pp 43ndash59

Hall Brain and Jeffrey Liebman 1998 ldquoAreCEOs Really Paid Like Bureaucratsrdquo QuarterlyJournal of Economics August 1083 pp 653ndash91

Hall Brian and Kevin Murphy 2002 ldquoStockOptions for Undiversi ed Executivesrdquo Journal ofAccounting and Economics February 331 pp3ndash42

Harrison J Michael 1977 ldquoIndependenceand Calibration in Decision Analysisrdquo Manage-ment Science November 243 pp 320ndash28

Hartgraves Al and George Benston 2002ldquoThe Evolving Accounting Standards for SpecialPurpose Entities and Consolidationsrdquo AccountingHorizons September 163 pp 245ndash58

Heninger William 2001 ldquoThe Association Be-tween Auditor Litigation and Abnormal Accru-alsrdquo Accounting Review January 761 pp 111ndash26

Hermalin Benjamin and Michael Weisbach2002 ldquoBoards of Directors as an EndogenouslyDetermined Institution A Survey of the Eco-nomic Literaturerdquo FRBNY Economic Policy ReviewForthcoming

Hirshleifer David and Siew Hong Teoh 2001ldquoHerd Behavior and Cascading in Capital Mar-kets A Review and Synthesisrdquo Unpublishedworking paper Ohio State University

Holmstrom Bengt and Steven N Kaplan2001 ldquoCorporate Governance and Merger Activ-ity in the United States Making Sense of the1980s and 1990srdquo Journal of Economic PerspectivesSpring 152 pp 121ndash44

Hong Harrison Jeffrey Kubik and Amit Sol-omon 2000 ldquoSecurity Analystsrsquo Career Con-cerns and Herding of Earnings Forecastsrdquo RandJournal of Economics Spring 311 pp 121ndash44

Johnson L Todd and Robert Swieringa 1996ldquoAnatomy of an Agenda Decision Statement No115rdquo Accounting Horizons June 102 pp 149ndash79

Kim Son Ku 1995 ldquoEf ciency of an Informa-tion System in an Agency Modelrdquo EconometricaJanuary 631 pp 89ndash102

Klein April 2002 ldquoAudit Committee Boardof Director Characteristics and Earnings Man-agementrdquo Journal of Accounting and EconomicsAugust 333 pp 375ndash400

Koopmans Tjalling 1947 ldquoMeasurementWithout Theoryrdquo Review of Economic Statistics Au-gust 293 pp 161ndash72

Krigman Laurie Wayne Shaw and Kent Wo-mack 2001 ldquoWhy Do Firms Switch Underwrit-ersrdquo Journal of Financial Economics May 602ndash3pp 245ndash84

Lambert Richard 2001 ldquoContracting Theory


and Accountingrdquo Journal of Accounting and Eco-nomics December 321ndash3 pp 3ndash87

Li Xi 2002 ldquoCareer Concerns of EquityAnalysts Compensation Termination andPerformancerdquo Unpublished working paperVanderbilt University

Lin Hsiou-Wei and Maureen McNichols1998 ldquoUnderwriting Relationships AnalystsrsquoEarnings Forecasts and Investment Recommen-dationsrdquo Journal of Accounting and Economics Feb-ruary 251 pp 101ndash27

McGill Gary and Edmund Outslay 2002 ldquoDidEnron Pay Taxes Using Accounting Informa-tion to Decipher Tax Statusrdquo Tax Notes 968 pp1125ndash136

McNichols Maureen and Patricia OrsquoBrien1997 ldquoSelf-Selection and Analyst CoveragerdquoJournal of Accounting Research 35 Supplementpp 167ndash99

Michaely Roni and Kent Womack 1999ldquoCon ict of Interest and the Credibility of Un-derwriter Analyst Recommendationsrdquo Review ofFinancial Studies Special 124 pp 653ndash86

OrsquoBrien Patricia 1988 ldquoAnalystsrsquo Forecasts asEarnings Expectationsrdquo Journal of Accounting andEconomics January 101 pp 53ndash83

Palmrose Zoe-Vonna 2000 Auditor LitigationConsiderations and Data Sarasota Florida Amer-ican Accounting Association

Peasnell Ken Peter Pope and Steven Young2002 ldquoBoard Monitoring and Earnings Manage-ment Do Outside Directors In uence Abnor-mal Accrualsrdquo Unpublished manuscript Lan-caster University

Peck Charles and Henry Silvert 2001 Direc-torsrsquo Compensation and Board Practices in 2001New York The Conference Board

Pizzo Stephen Mary Fricker and Paul Muolo1989 Inside Job The Looting of Americarsquos Savingsand Loans New York McGraw-Hill

Powers William Jr Raymond Troubh andHerbert Winokur Jr 2002 ldquoReport of Investiga-

tion by the Special Investigative Committee ofthe Board of Directors of Enron Corprdquo

Prendergast Canice 1999 ldquoThe Provision ofIncentives in Firmsrdquo Journal of Economic Litera-ture March 371 pp 7ndash63

Previts Gary and Barbara Merino 1998 AHistory of Accountancy in the United States ChicagoOhio State University Press

Rajgopal Shivaram and Terry Shevlin 2002ldquoEmpirical Evidence on the Relation BetweenStock Option Compensation and Risk TakingrdquoJournal of Accounting and Economics June 332pp 145ndash71

Rose Nancy and Catherine Wolfram 2000ldquoHas the lsquoMillion-Dollar Caprsquo Affected CEOPayrdquo American Economic Review May 902 pp197ndash202

Sapsford Jathon and Paul Beckett 2002ldquoCitigroup Deals Helped Enron Disguise itsDebts as Tradesrdquo Wall Street Journal July 22CCXL p 15

Schipper Katherine 1991 ldquoAnalystsrsquo Fore-castsrdquo Accounting Horizons December 54 pp105ndash21

Shleifer Andrei and Robert Vishny 1997 ldquoASurvey of Corporate Governancerdquo Journal of Fi-nance June 522 pp 737ndash83

Sunder Shyam 1997 Theory of Accountingand Control Cincinnati Ohio South-WesternPublishing

Swanger Susan and Eugene Chewning Jr2001 ldquoThe Effect of Internal Audit Outsourcingon Financial Analystsrsquo Perceptions of ExternalAuditor Independencerdquo Auditing A Journal ofPractice and Theory September 202 pp 115ndash29

Tirole Jean 2000 ldquoCorporate GovernancerdquoEconometrica January 691 pp 1ndash35

Zeff Stephen 1998 ldquoIndependence andStandard Settingrdquo Critical Perspectives on Account-ing October 95 pp 535ndash43

Zeff Stephen 2002 ldquolsquoPoliticalrsquo Lobbying onProposed Standards A Challenge to the IASBrdquoAccounting Horizons March 161 pp 43ndash54


The savings and loan crisis which began in the mid-1970s and lasted through-out the 1980s is noteworthy because it combines regulatory changes nancialfraud audit failures and a parade of con icts of interest that reached all the way tomembers of the US Senate (for one discussion see Pizzo Fricker and Muolo1989) It also ended up costing taxpayers in excess of $150 billion EricksonMayhew and Felix (2000) provide an in-depth analysis of one such audit failure theinfamous case of Lincoln Savings and Loan that ended up with a $2 billion price tagfor taxpayers The controlling owner of Lincoln Savings and Loan was found guiltyof fraudulent land transactions in state and federal court (although the federalconviction was later overturned on the grounds of mistaken jury instructions)while three major auditing rms reached out-of-court settlements totaling $135 mil-lion In 1989 the Senate Ethics Committee censured one senator and cited fourothers for questionable conduct because they had received substantial contribu-tions from the controlling owner of Lincoln Savings and Loan and had interferedwith federal regulators who were investigating the rm

In 1995 Barings Bank of the United Kingdom was driven into bankruptcy after233 years of operation because a 28 year-old trader in Singapore who was supposedto be searching for small-margin arbitrage opportunities lost over $1 billionspeculating with huge purchases of futures and option contracts on the Japanesestock market This case is noteworthy because of the sheer ineptness of Baringsrsquocontrols the rogue trader did his own accounting and reporting (Fay 1996)

The novelty then is not the presence of con icts of interest but theirmanagement Societies rely on various devices to manage these con icts Someactivities are prohibited such as an auditor engaged with an explicit pay-for-performance contract while at other times we rely on disclosure of relationshipssuch as my universityrsquos requirement that the faculty disclose any nancial interestin teaching materials Both prohibitions and disclosure requirements are but-tressed by legal requirements and remedies In between we seek arrangements thatat least dampen the underlying con ict as when the partner in charge of an auditis rotated or when the executiversquos performance is judged in part on the basis of nancial results that have been audited by an independent auditor In all of thesecases as with other economic activities we balance tensions and shy away fromattempting to create a situation of zero failures as a result of con icts of interestsimply because pursuing such an extreme goal would be uneconomic

I propose then to survey con icts of interest in the corporate arena withemphasis on recent events The presence of such con icts is not new We still haverogues Some use modern techniques as in Enronrsquos use of complex structured nance to bolster its apparent performance Others rely on the most mundane oftechniques as in WorldComrsquos capitalization of an expense serving to spread theexpense over many years instead of writing it off in total in the given year

Yet the seemingly steady drumbeat of recent corporate improprieties leads oneto wonder what is new in the corporate setting and what along the way to moderncommunication globalized organizations and markets and structured nancemight have been overlooked Surely complexity has increased and this in itself








Joel S Demski 53

























Joel S Demski 57


























Joel S Demski 61


























Joel S Demski 65




















Conclusion



Joel S Demski 69



References






















































































Joel S Demski 53

























Joel S Demski 57


























Joel S Demski 61


























Joel S Demski 65




















Conclusion



Joel S Demski 69



References








































































































Joel S Demski 57


























Joel S Demski 61


























Joel S Demski 65




















Conclusion



Joel S Demski 69



References









































































































Joel S Demski 61


























Joel S Demski 65




















Conclusion



Joel S Demski 69



References









































































































Joel S Demski 65




















Conclusion



Joel S Demski 69



References



































































































Conclusion



Joel S Demski 69



References


















































































References
















































































corporate conflicts of interesthomepages.ulb.ac.be/~plegros/documents/classes/finance/themes... ·...

Documents