Market vs. regulatory responses to corporate fraud: A critique of the Sarbanes-Oxley Act of 2002
The latter part of the 20th century saw a debate about the appropriate approach to regulating the corporation.
Other commentators, notably including Henry Marine, Frank Easterbrook, and Daniel Fischel, building on basic theoretical work by, among others, Ronald Coase, Armen Alchian, and Harold Demsetz, argue that the corporation was appropriately viewed as fundamentally the product of private contractual relationships, and should be entitled to the same presumption of efficiency as contracts generally.5 The twin pillars of this view are that aspects of corporate governance that might seem suspect, including the separation of ownership and control, actually make economic sense; and that public corporate governance arrangements are disciplined in efficient securities markets.
The spectacular crashes and frauds of Enron, WorldCom, and other companies, including Sunbeam, Waste Management, Adelphia, Xerox, and Global Crossing, have reinvigorated this debate. The most public phase of the scandals began with Enron, which was at one time the seventh largest firm based on market capitalization, but in the fall of 2001 was suddenly shown to be a facade created by financial manipulation. WorldCom, which owns among other things the second largest long distance telephone carrier, disclosed in the summer of 2002 that it had created billions in earnings by IMAGE FORMULA 6
capitalizing expenses as needed.6 These firms' managers have become poster boys for the problems of separation of ownership and control. The frauds occurred despite several levels of monitoring by, among others, directors, prominent accounting and law firms, institutional shareholders, debt rating agencies, and securities analysts. Supposedly efficient securities markets adhered to overly optimistic assumptions about firms' business plans in the face of mounting evidence to the contrary.
Some argue that, because of the unreliability of corporate managers, monitors, and the market in these cases, government regulators need to restore confidence in the securities markets.7 The most important legal response was the Sarbanes-Oxley Act of 2002, summarized in the Appendix.8 This Act is the most sweeping federal law concerning corporate governance since the adoption of the initial federal securities laws in 1933 and 1934.9
This Article argues that, despite all the appearances of market failure, the recent corporate frauds do not justify a new era of corporate regulation. Indeed, the fact that the frauds occurred after seventy years of securities regulation shows that more regulation is not the answer.10 Rather, with all their imperfections, contract and market-based approaches are more likely than regulation to reach efficient results. Post-Enron reforms, including Sarbanes-Oxley, rely on increased monitoring by independent directors, auditors, and regulators who have both weak incentives and low-level access to information. This monitoring has not been, and cannot be, an effective way to deal with fraud by highly motivated insiders. Moreover, the laws are likely to have significant costs, including perverse incentives of managers, increasing distrust and bureaucracy in firms, and impeding information flows. The only effective antidotes to fraud are active and vigilant markets and professionals with strong incentives to investigate corporate managers and dig up corporate information.
Part I states the case for reform that emerges from recent corporate frauds. Part II discusses regulatory reform proposals, focusing on the Sarbanes-Oxley Act. Part III discusses the costs of reform, while Part IV discusses the potential for market-based IMAGE FORMULA 9
reforms. Part V presents conclusions and implications.
I. THE PROBLEM OF CORPORATE FRAUD
This Part discusses the problems revealed by Enron and other frauds. Subpart A briefly reviews the relevant facts of Enron and other recent corporate frauds. Subpart B discusses the case for a regulatory response based on the apparent scale of deception and the fact that it escaped detection by several layers of seemingly sophisticated monitors. In general, while this discussion reveals flaws in market and contractual devices designed to catch fraud, it also sets the stage for later discussion purporting the likely ineffectiveness of regulatory responses to Enron.
A. An Overview of Enron and Other Corporate Frauds
The Enron story has been recounted many times, probably most clearly and succinctly by William Bratton,11 and most exhaustively in the report of the Enron Board's Special Committee.12 It will be briefly summarized here. Enron was initially successful in creating an energy market that eliminated the need for utility companies to engage in potentially costly vertical integration.13 It later applied this innovation to other markets, such as fiber optic cable.14 Enron gave the impression of ever-increasing earnings and stable finances through extensive derivatives trading and profitable transactions with special purpose entities (SPEs), which also yielded substantial gains for Enron insiders.15 In fact, the apparent profits were illusory. Among other things, Enron apparently used "marking to market" to book as revenue speculative predictions of years of future sales, and outside investors loaned money to Enron in transactions disguised as revenue from prepay commodities contracts.16
The worst problems centered on Enron's hiding risks through SPEs created for the purpose of keeping debts off of its balance sheet. Enron's disclosure on October 16, 2001 that it was taking a half-billion-dollar after-tax charge against earnings and $1.2 billion reduction of shareholders' equity related to transactions with one of its special purpose entities 17 began the collapse that quickly led to litigation and bankruptcy. The seeds were sown when Enron formed an initial entity, Chewco, to take over CalPERS' investment in IMAGE FORMULA 15
its JEDI joint venture with Enron.18 In order to ensure that Enron would not have to report Chewco's debt on its balance sheet, Chewco's equity needed to be at least three percent of total capital, a requirement that was not met.19 Also, although Barclays provided an outside investment, the loan was secured by a cash reserve account that Chewco itself funded.20 When Andrew Fastow (Enron's chief financial officer) presented this information to the Enron board in November 1997, he did not clarify the nature of the "outside" equity.21
Enron's network of SPEs soon became larger and more complex. Enron formed LJM I in June 1999, which Fastow controlled as the manager of the LLC that served as the entity's general partner, and LJM II, which syndicated investments to outside investors.22 In May 2000, Enron began establishing the Raptor entities, purportedly to hedge against declines in Enron investments.23 The "hedge" consisted of a put pursuant to which Enron could sell its stock back to the Raptor if it declined.24 Thus, the "hedge" was really a bet that securities markets would keep running up Enron's stock price despite its losing investments in the SPEs. Enron was not hedged but was really speculating on derivatives, including the put on its own stock.25 The Enron Finance Committee had recognized the risk, but nevertheless signed off, relying on Arthur Andersen's willingness to do so.26
Enron collapsed quickly after restating its earnings and debts in October 2001.27 The Raptors could no longer cover investment losses as Enron's stock fell. The collapse was hastened by the fact that Enron's basic business depended on its customers trusting its ability to bear market risks, which the falling stock price called into doubt.
In addition to hiding its risks and losses, Enron and its insiders profited from churning of financial assets between Enron and its SPEs, booking gains at both ends that did not reflect the real value of the assets. Although the Enron board understood the problem with potential insider transactions stemming from Fastow's position with the SPEs, it relied on arm's-length negotiations between Fastow and the relevant Enron divisions to protect Enron's interests, as well as on a set of special review procedures.28 However, the Special Committee concluded that the board inadequately probed the transactions given Fastow's involvement and that the very need for exhaustive controls should have suggested to the board the dubiousness of the transactions.29
The blame for Enron's failure has been widespread. The Enron board's Special Committee noted failures at all levels of monitoring within the company: senior IMAGE FORMULA 18
management, including Lay, Enron's chief executive officer at the beginning of its rapid rise, Skilling, the chief operating officer who became chief executive officer, Causey, the chief accountant, and Buy, the senior risk officer; the board, particularly including its audit committee, Arthur Andersen, which was aware of much of the detail but did not insist on clear disclosure; and Enron's main outside counsel, Vinson & Elkins, which helped structure SPE transactions and draft disclosures,30 and whose August 2001 investigation of the famous Sherron Watkins memo concluded that Arthur Andersen's accounting for the Raptor entities was not technically inappropriate.31
There were also failures in the securities markets. Securities analysts continued to give buy recommendations, and all of the major debt rating agencies rated Enron's debt as investment grade, right up until shortly before bankruptcy.32 Arthur Andersen has been virtually shut down by its conviction for Enron-related obstruction of justice,33 Andrew Fastow has been indicted for securities fraud,34 and Michael Kopper has pleaded guilty to fraud.35
Although Enron has been the most publicized of the corporate frauds, it is not the only one.36 In June 2001, Waste Management settled SEC charges arising out of a restatement of $1.43 billion in earnings relating to 1992-1996 financial statements involving a variety of improper accounting techniques, including improper capitalization of expenses, failure to amortize, and improper use of reserves.37 Waste Management's auditor (again, Arthur Andersen), continued to give unqualified audits despite its early recognition that Waste Management was a "high risk client" that "actively managed reported results," had a "history of making significant fourth quarter adjustments" to its financial statements, and was in an industry that required "highly judgmental accounting estimates or measurements," and Waste Management's failure to comply with Arthur Andersen's plan for the firm to get its books in order-a plan that was not communicated to Waste Management's audit committee.38
Sunbeam was another precursor of Enron. In June 1998, a Barrons article revealed that Sunbeam had manipulated its financial statements by, among other things, excessive write-downs in a "big bath" restructuring, booking phony sales and rebates, and not IMAGE FORMULA 20
accounting for accounting and other advertising expenses.39 These irregularities explained most or all of Sunbeam's seemingly remarkable turnaround following "Chainsaw Al" Dunlap's 1996 restructuring. The irregularities were not spotted by Sunbeam's auditor (again, Arthur Andersen), or by Sunbeam's independent directors.
Other accounting shenanigans in public corporations include Xerox's accelerating revenues from long-term equipment leases,40 Qwest's and Global Crossing's manipulation of revenues and expenses on sales and swaps of fiber optic capacity,41 and apparently rampant looting by the family controlling Adelphia. Most strikingly, WorldCom misstated billions in current expenses (fees the company paid to use transmission networks) as capital expenditures.42 Like the other problems noted above, these irregularities were not uncovered by the firm's board, its initial auditor (again, Arthur Andersen), or by telecom analysts who had long followed WorldCom and continued until virtually the moment before the public disclosure to give it "buy" ratings.43
B. The Case for Market Failure
Modern regulatory theories of corporate governance begin with Adolf Berle and Gardiner Means, who argued in the wake of the 1929 crash that owners of publicly held corporations could not effectively control their corporations.44 This lack of control effectively involves two problems. First, as Adam Smith observed, corporate managers do not watch over "other people's money" with the "anxious vigilance with which the partners in a private copartnery frequently watch over their own."45 In other words, public corporations involve agency costs. Second, agency costs are high because it is impractical for shareholders with small, dispersed interests to invest much time and money in monitoring managers.
The antiregulatory response to Berle and Means is essentially that shareholders cannot be as vulnerable to misappropriation as the regulatory model would suggest because no one can force them to invest in firms that will squander their money. The capital markets therefore can be expected to develop devices that overcome the problems Berle and Means discussed. These devices include hostile takeovers and other devices that aggregate shareholder voting power and facilitate shareholder monitoring, alignment IMAGE FORMULA 24
of managers' and shareholders' interests through incentive compensation, and monitors such as independent directors and large accounting and law firms.
Efficient securities markets, in turn, provide both an effective valuation device to enable these devices to operate and a mechanism for pricing and testing the efficacy of the devices. The takeover market functions because the price of a suboptimally managed firm drops enough to make it worthwhile for better managers to buy the stock and replace the incumbents. Stock compensation is an efficient incentive because it aligns managers' interests with shareholder welfare. A company's stock price can be viewed as measuring the value of its bundle of contracts. Even if the market cannot know the evil that lies within managers' hearts, it can observe the contracts that tend to keep them honest. Investment dollars will tend to flow to the firms with the most efficient governance devices.
Enron and other recent scandals seem to threaten the viability of the antiregulatory model because all of these contracts and markets apparently failed. Enron, for example, had prestigious outside directors, and law and accounting firms who might have been expected to care more about their reputations than about letting Enron insiders get away with shady deals. Enron's stock price failed to reflect even obvious problems until it was too late. Enron's 2000 Annual Report showed large affiliate investments and liabilities.46 Although these disclosures did not clarify that Enron's income essentially depended on derivatives trading and that Enron was not actually hedged, the disclosures were disconcerting enough to lead an adequately inquisitive market to check further and get to the facts. Even more striking is the fact that Enron's ninety dollar share price in 2000 could be justified only by some very unrealistic assumptions, such as a twenty-five percent return on equity forever, revenues of $700 billion within ten years, and an increase thereafter of ten percent per year, more than twice U.S. public firms' historical rate.47 And apparently few people stopped to wonder whether Enron's not paying federal taxes for four of the five tax years through 2001 indicated that it was really not making any money.48
To be sure, Enron's stock price dropped from eighty dollars to forty dollars in the first eight months of 2001 in the face of rising earnings, indicating that Enron's problems may have been seeping into its price.49 But Enron's trading for thirty to forty dollars per share when it was probably worthless suggests that the market had spotted only a small piece of its problems. Moreover, most institutional investors failed to sell Enron until IMAGE FORMULA 27
October 2001.50 This raises serious questions concerning the extent to which the market can be relied on to ferret out facts and to make efficient judgments about appropriate governance devices.
Enron and other accounting frauds may indicate a new category of problems that calls for new regulation. Enron insiders seem to be a new breed of corporate executives who are unconstrained by the traditional devices.51 These executives are hyper-motivated survivors of a highly competitive tournament (that Enron called "rank and yank"52) who have proven their ability to make money while putting on a veneer of loyalty to the firm. At least some of the new breed appear to be Machiavellian, narcissistic, prevaricating, pathologically optimistic, free from self-doubt and moral distractions, willing to take great risk as the company moves up and to lie when things turn bad, and nurtured by a corporate culture that instills loyalty to insiders, obsession with short-term stock price, and intense distrust of outsiders.
Even these new-breed executives would not present a problem if effective monitors watched them. The companies in which major frauds occurred had directors who were not members of senior management and high-priced auditors who missed fundamental problems for a long time. Arthur Andersen was deeply involved in structuring the Enron transactions and cannot plausibly claim lack of expertise. But Arthur Andersen had a strong incentive to hang on to a major buyer of both audit and nonaudit services, and the partner in its Houston office principally responsible for Enron, David Duncan, derived a significant amount of compensation from selling these services to Enron.53 There were similar facts regarding Arthur Andersen's and its partners' relationship with Waste Management.54
The accounting profession seems not to have adjusted to the transition from professional to profit-maximization norms.55 Like other auditing firms, Arthur Andersen pressed the business side, exhorting its partners to sell nonaudit services to audit clients and tying partner compensation to business production.56 In other words, auditing firms have used their auditing services, which firms must buy, as "loss leaders" to sell nonauditing services.57 Auditors' loss of independence in effect may have made them IMAGE FORMULA 30
part of the management team in some cases.58 Years of working for the same client, along with prospects of joining the client's management and participating in its success, may have made auditors subject to the same pathologies that affected client management, including excessive optimism and loyalty, and reduced their concern for their auditing firm's reputation.59 Moreover, as the same people worked for the same clients from year to year, they may have found themselves bound to defend errors from earlier audits.60
Executives, directors, monitors, and the stock markets all arguably were hamstrung in recent cases by the challenges new business methods created for securities valuation and accounting methods. Historical price-earnings multiples did not constrain valuations in the face of arguments that novel business methods created sky-high potential for future earnings. New types of business required determination of appropriate accounting methods. This was certainly true for Enron, which Fortune ranked the most innovative company in America for six straight years, whose chief executive, Jeff Skilling, was named second best in the country (after Microsoft's Steve Ballmer) as recently as April 2001, and which had been the subject of admiring business school studies.61 A more skeptical market, with a more precise metric for measuring the success of "new economy" firms, might have looked more closely at Enron's numbers. This valuation problem was, of course, common in a market that was willing to assume that every new Internet business would produce limitless profits by increasing demand to infinity and reducing costs to zero. Many of these firms were honest failures that made their risks clear to investors. Enron showed how firms could capitalize even more spectacularly on the market's gullibility by inflating earnings and hiding risks.
The recent frauds cannot, however, entirely be attributed to new ways of doing business. Sunbeam and Waste Management were old-economy companies, and the problems in WorldCom and other telecoms apparently involved straightforward mischaracterizations of revenues and expenses. This suggests that the accounting frauds resulted more from fundamental incentive problems than from new business methods.
Whatever the cause, the costs of corporate fraud potentially go beyond owners, employees, and others associated with defrauding firms. If the market cannot distinguish efficient from inefficient firms, investors may, at least at first, put too many resources in the inefficient firms, and ultimately may stay out of the market because they cannot spot the "lemons,"62 with the result that the economy becomes less productive. This suggests that if markets have failed, government must step in. IMAGE FORMULA 32
II. REGULATORY RESPONSES TO CORPORATE FRAUD
Highly publicized corporate frauds engendered a blizzard of reform proposals, mostly in Congress.63 Although current federal securities law does not permit direct federal regulation of corporate governance,64 Congress can change those laws. Moreover, even under the current general disclosure-oriented approach, Congress, and the SEC on its own, can regulate corporate governance indirectly through disclosure laws. The securities exchanges also can regulate corporate governance through their listing requirements. This Part briefly reviews some of the reforms in the Sarbanes-Oxley Act and other responses to recent corporate frauds and discusses how these reforms purportedly respond to the problems discussed in Part I.65 Part III analyzes the benefits and costs of these reforms.
A. Independent Directors
One of the favorite projects of corporate reformers has been the creation of the socalled "monitoring" board. The basic theory is that the corporation's main decisionmaking body should include a majority of "independent" directors who do not work fulltime for the corporation and therefore theoretically are in a position to watch over the insiders, with wholly independent "audit" and "nominating" committees that work with the company's auditing firm and control election of directors.66
The monitoring board model has come into standard usage by large public corporations pursuant to recommendations over the last twenty years by, among others, the American Bar Association's Committee on Corporate Law of the Section of Corporation, Banking & Business Law,67 the SEC,68 the New York Stock Exchange (NYSE),69 and the American Law Institute's Principles of Corporate Governance.70 In 1999, the SEC adopted a rule requiring corporate proxy statements to disclose whether the firm's audit committee has discussed the audited financial statements with management and accounting policies with the auditors, received the auditors' assurances IMAGE FORMULA 37
of its independence, and recommended to the full board that it include in the company's annual report audited financial statements, any written charter the board has adopted for the audit committee, and information concerning audit committee members' independence.71 In the wake of the recent corporate frauds, an American Bar Association task force has made preliminary recommendations on increasing director independence.72
Suggestions and requirements of greater board independence and more board monitoring are predictable responses to Enron and other corporate frauds. Recent events indicate large public companies' managers may need more watching regarding such matters as insider transactions, compensation, and selection and supervision of auditors. The New York Stock Exchange Board of Directors has adopted and submitted to the SEC new listing standards requiring a majority of the board to have no material relationships with the firm, and lengthening to five years the "cooling off ' period for board service by former employees of the issuer or its auditor, requiring that directors meet without management, requiring wholly independent nominating and compensation committees in addition to the independent audit committee, requiring the chair of the audit committee to have accounting or financial management expertise, requiring the audit committee to have sole responsibility for hiring the auditing firm, and prohibiting compensation of audit committee members apart from directors' fees.73 In addition to increasing board monitoring, these rules require ethics and conduct codes and prompt disclosure of waivers, shareholder approval of all equity-based compensation rather than only such compensation to officers or directors as under the prior rules, certification by the chief executive officer of procedures and compliance with those procedures regarding accuracy of information, a program for education of board members, and a provision for public reprimand by the NYSE for violation of its rules.74
Section 301 of the Sarbanes-Oxley Act directs securities exchanges and securities associations to prohibit the listing of any security of an issuer that does not have an audit committee responsible for hiring and oversight of auditors and that is wholly independent.75 Section 407 requires disclosure concerning the audit committee's financial expert.76
B. Other Governance Protections Against Fraud
Corporate responsibility theoretically is enhanced if senior full-time managers have appropriate incentives not to engage in wrongdoing and to ensure that their underlings IMAGE FORMULA 42
are not engaging in wrongdoing. A significant problem in Enron, WorldCom and other cases was that senior managers apparently were involved in the frauds, making them exceptionally difficult to uncover. Also, managers and monitors need to be able to get information about possible fraud from their underlings. Establishment of control systems within the firm and protecting whistleblowers helps ensure the flow of information within the company.
Sarbanes-Oxley addresses these concerns by requiring executive certification of reports and of internal controls,77 providing for SEC rules prohibiting fraudulently influencing or misleading auditors,78 requiring attorney reporting of evidence of fraud,79 requiring disclosures concerning the firm's internal control structure and code of ethics for financial officers,80 and protecting whistleblowers.81
C. Auditor Oversight and Incentives
Enron and other recent fiascos pointed out weaknesses in the auditing of corporate books. In general, the problems have involved some combination of excessive ties between auditing firms and the companies they are supposed to be scrutinizing, inadequate review of the accounting firm's work by corporate audit committees, discussed in the preceding Subpart, inadequate industry or government scrutiny of accounting firms' work, and excessively lax accounting standards.
There has been much attention focused on tightening standards regarding auditors' nonauditing work for audit clients and other aspects of auditors' relationships with clients that compromise auditors' independence. Auditors theoretically are independent because the highly concentrated structure of the auditing industry means that auditing firms do not depend heavily on audit revenues from particular clients. Moreover, large accounting firms have invested over many years in valuable reputations. They have a strong incentive not to forfeit these "reputational bonds."82
But accounting firms and their members clearly are subject to countervailing pressures. Even the largest accounting firm may have an incentive to overlook misconduct from a client from which it makes significant fees for consulting and other nonaudit work. The auditor may believe that her conclusions are honest, but only because her judgment is affected by a "self-serving" bias to view behavior that serves her interests in the most favorable light.83 Moreover, auditors have incentives to maximize their own income within the firm, which may not match the incentives of the firm as a whole to guard its reputation.84 For example, David Duncan, in the Enron case, derived IMAGE FORMULA 46most of his compensation from Enron.85
In response to concerns about auditor independence and after extensive deliberation, the SEC in late 2000 adopted rules that reflect a compromise between those who believed that substantial reform was necessary and those who believed that major reform would be costly and unnecessary.86 In general, the rules forbid several types of nonaudit services for audit clients and require firms to make proxy disclosures of their aggregate audit and nonaudit fees and whether the audit committee "has considered whether the provision of non-audit services is compatible with maintaining the principal accountant's independence."87
Sarbanes-Oxley overtakes the SEC in this regard by barring auditors from nonaudit work for audit clients,88 requiring rotation of audit partners after five years,89 requiring corporate audit committees to select auditors,90 and restricting auditing of a firm by a member of whose senior management was previously employed by the auditor.91
Part of the problem regarding laxity of accounting arguably has to do with issuer board oversight of auditors' work. This is addressed by the audit committee requirements discussed in Part II.B. Sarbanes-Oxley also requires more detailed reporting by auditors to boards,92 and helps ensure that the audit committee has adequate expertise to perform its oversight function by requiring disclosure concerning the audit committee's financial expert.93
The problems with auditor independence and laxity of auditors' work in uncovering accounting frauds arguably indicate that it may be time to abandon the current model of regulating the public accounting profession. The fact that reputational bonds posted by national accounting firms may no longer be enough to ensure quality work arguably suggests that regulation of the accounting industry should move closer to the model that has applied to the more fragmented securities industry. The current system of peer review within the AICPA obviously has not filled in the gaps, as indicated by the recent corporate frauds themselves and by the fact that no major accounting firm has failed a peer review.
The SEC proposed a rule providing for an independent agency dominated by people who are not accounting professionals that would, among other things, exercise oversight, annually review large accounting firms for independence, objectivity, performance, and methodology, and set standards for audits and quality control.94 This proposal now has been superseded by the Sarbanes-Oxley requirements for a Public Company Accounting IMAGE FORMULA 50
Oversight Board, with oversight by the SEC, to register, inspect, and investigate public accounting firms, directed by five members of which only two have industry ties in the sense of being or having been licensed certified public accountants, and none can be currently working for a major accounting firm.95
Finally, the problems in some cases may have been at least partly attributable to accounting standards for how items must be reported rather than to auditors' noncompliance with these standards. In particular, Enron's obligation to place the debts of its special purpose entities on its own balance sheet hinged on its having total equity amounting to just three percent of capital.96 Although the fall 2001 restatements that triggered the company's downfall were necessitated by the company's noncompliance even with this lax standard, it is not clear why the difference between a minimal and no outside equity should justify significant differences in reporting of debts.97 Also, current standards relating to so-called "pro forma" earnings that exclude "extraordinary" events give firms significant room to manipulate earnings.98 Sarbanes-Oxley addresses these problems by imposing conditions on the structure of the organization that approves accounting standards the SEC deems "generally accepted."99
D. Executive Compensation
The corporate frauds relate in two ways to insider compensation. First, some of the frauds, including, most notoriously, Fastow and others' suspect SPE deals in Enron, Adelphia's dealings with its controlling Rigas family, who have been arrested for looting the company of a billion dollars, and WorldCom's loans to its longtime leader Bernard Ebbers, appear to have involved actual or borderline looting. High levels of executive compensation, though short of outright looting, arguably result from excessive executive control over their own compensation.100
Second, stock-based compensation, particularly compensation based on short-term options, arguably gives insiders incentives to manipulate operations, or earnings alone, to reap large short-run payoffs and then sell before the market adjusts.
Third, inadequate disclosure of the compensation or its effects on earnings may compound these problems. A salient issue in this respect is not requiring firms to record the difference between the market and option price of stock options as an expense just IMAGE FORMULA 54
like most other forms of compensation. Lobbyists defeated a proposal to require expensing of stock options in Sarbanes-Oxley.101 Ironically, Congress threatened to step in to block the move when the Financial Accounting Standards Board (FASB) proposed requiring expensing in 1993.102 Now the FASB is considering the issue again, and the International Accounting Standards Board recently proposed expensing. 103
The NYSE proposal would address some concerns with employee compensation by requiring shareholder approval of all equity-based compensation, thereby expanding its prior rule, which required approval only of stock-based compensation to officers or directors.104 Sarbanes-Oxley deals with the compensation issue mainly by prohibiting insider loans.105 The Act also addresses compensation as a way of penalizing fraud by requiring return of incentive-based compensation, as well as profits from stock sales, following accounting restatements resulting from "the material noncompliance of the issuer, as a result of misconduct, with any financial reporting requirement under the securities laws."106
E. Increasing Disclosures
The corporate frauds entailed firms' inadequate disclosure to shareholders of data relating to the quality of earnings, corporate debts and risks, and insider transactions. The firms themselves are the cheapest sources of this information, but their insiders may lack adequate incentives to be forthcoming, particularly where the disclosures would uncover their own wrongdoing.
The securities laws, of course, already require substantial disclosure, including reporting of important corporate events and extensive quarterly and yearly filings. However, the recent corporate frauds arguably have altered the sorts of things investors would be concerned about. Thus, Sarbanes-Oxley requires the SEC to issue rules requiring, among other things, enhanced disclosure regarding off-balance-sheet transactions and pro forma earnings107 and clear and immediate disclosure of material changes in financial condition.108
The recent corporate frauds may have related less to problems with existing disclosure requirements than to firms' failure to comply adequately with these requirements. This arguably could be remedied by increasing penalties for disclosure failures. Some proposals would focus the penalty on chief executives. The SEC has issued an order requiring executives of certain large firms to certify financial disclosures on penalty of personal liability,109 and the NYSE is considering requiring listed firms' chief executive officers to certify their firms' procedures regarding accuracy of IMAGE FORMULA 58
information and compliance with those procedures.110 Sarbanes-Oxley requires chief executives and chief financial officers to certify financial statements,ill with criminal penalties for reckless certification.112 Increased civil and criminal liability for fraud is discussed in Part II.F.
Finally, increasing SEC surveillance theoretically can reduce corporate fraud. The SEC traditionally has lacked the resources to do intensive review of corporate disclosures. Sarbanes-Oxley addresses this by requiring increased SEC review of public company filings1 13 and increasing the SEC's funding.114
F. Fraud Liability
Recent corporate frauds occurred following 1990s laws scaling back potential liability for corporate fraud, the Supreme Court's 1994 decision in Central Bank eliminating aiding and abetting liability under the 1934 Act's general antifraud provision,115 Congress's scaling back of securities class actions in the Private Securities Litigation Reform Act (PSLRA),116 and state limited liability partnership (LLP) laws eliminating vicarious liability of members in accounting and other partnerships.117 These laws reduced the exposure of gatekeepers such as Arthur Andersen, as well as that of nondisclosing corporations and their insiders. The reduced liability risk may have encouraged fraudulent or shirking behavior in marginal situations where defrauding insiders or lax auditors had persuaded themselves that the likelihood of detection was low. This argues for reversing some aspects of the PSLRA.
Sarbanes-Oxley makes several moves toward increasing penalties for corporate fraud, including increased criminal penalties,118 reducing the standard for officer-- director bars,119 increasing the statute of limitations for fraud claims,120 and providing for bars on professional practice before the SEC.121 It also indirectly penalizes fraud by requiring the return of incentive based compensation or profits from stock sales following accounting restatements resulting from "misconduct" whether or not by the IMAGE FORMULA 62
executive whose compensation or profits had to be returned.122
G. Securities Analysts
The markets did a bad job of uncovering fraud at least partly because securities analysts and other market professionals were not assiduous in finding and reporting problems with firms they were covering. The underlying problem was analysts' increased involvement in the sale of securities, particularly following the deregulation of commissions, which made it harder for securities firms to compensate analysts for research.123 Analysts' research function came to be compromised by their own investments and their ties to the investment banking arms of their securities firms.124 These problems were addressed in part by recent NASD rules approved by the SEC that regulate and require disclosure of ties between analyst research and investment banking activities of analysts' employers and analysts' ownership of recommended stocks.125 Sarbanes-Oxley addresses these conflicts by ordering significant new regulation of securities analysts, including rules ensuring that analysts are not subject to pressure on account of their firms' investment banking activities.126
III. PROBLEMS WITH THE REGULATORY APPROACH
This Part shows that the case for additional regulation in response to Enron and related scandals is weaker than might appear from the litany of market failures discussed in Parts I and II. As discussed in Part III.A, corporate fraud has deep-seated causes, including the regulatory environment itself, and is accordingly not amenable to simple regulatory solutions.
Part III.B discusses the inherent limitations on what the new regulation can accomplish. The basic problem is that post-Enron reforms hope to reduce fraud mainly by making outside monitors more independent. But independence reduces access to information, and therefore is not likely to be an effective counter to the high-powered incentives of the perpetrators.
Even if increasing the level of regulation reduces the risk of future fraud, Part III.C shows that the costs of increased truthfulness may outweigh the benefits, including deterring beneficial transactions, increasing the adversarial nature of corporate IMAGE FORMULA 68
governance, reducing executives' incentives to increase firm value, and diverting executive talent to closely held firms. In general, as Alan Greenspan warned, "[w]e have to be careful... not to look to a significant expansion of regulation as the solution to current problems, especially as price/earnings ratios increasingly reflect the market's perception of the quality of accounting."127
Most of this Part discusses inherent limitations on what government is able to accomplish. Beyond these limitations, political considerations usually cause regulation to fall short of its theoretical potential. As discussed in Part III.D, this is particularly true of Sarbanes-Oxley, which was passed in a hectic environment in which politicians played on public misperceptions of risk and eschewed careful balancing of costs and benefits.
A. Explaining the Frauds
Before trying to make the markets safer, it is necessary to consider why the latest frauds occurred and to place them in historical context. This is not the first time that widespread financial chicanery has occurred in the context of rampant market speculation. For example, some of the speculation preceding the 1929 Crash has a familiar ring. J.K. Galbraith recounts Goldman Sachs' launching of a series of "exiguous" trading companies whose assets consisted largely of their own stock, rose sharply with their own value, and fell just as fast.128 After seventy years of regulation, Enron did much the same thing. Indeed, almost 300 years ago the South Sea Bubble, the high tech of its day, lured investors with, among other things, the hope of riches from the new world, only to collapse amid recriminations against directors and "stock-jobbers."129 This history raises at least preliminary questions as to whether new regulation will work any better to prevent tomorrow's frauds than yesterday's regulation did to prevent those of today.
Part III.A discusses some of the conditions that may have given rise to the corporate frauds. It also shows that regulation itself may have contributed to creating conditions that are conducive to fraud. This discussion indicates the difficulty and danger that await regulators who attempt to eradicate fraud.
1. Insiders' Incentives and Heuristics
In order to fix the markets, it is necessary to understand why the insiders who pulled accounting scams at major public corporations thought they could get away with them in efficient and regulated securities markets. A thorough understanding of the perpetrators' motives would seem to be essential in designing regulation that has a significant chance of preventing future frauds. It is too simplistic to ascribe these frauds to "greed" without accounting for the risk of detection. Notably, in contrast to notorious crooks such as Robert Vesco, none of the main characters in the recent scandals tried to flee. Moreover, the alleged perpetrators were not shady criminals but seemingly responsible business IMAGE FORMULA 74
people who had earned the trust of their, even more respectable, monitors. For example, Scott Sullivan, who is accused of manipulating WorldCom's books in order to meet earnings targets, was regarded as "one of the best chief financial officers around" and "the key to WorldCom Inc.'s financial credibility."130 How could such a man have engaged in large-scale financial manipulation if this is proved to be the case? Similar questions arise regarding the seemingly more blatant behavior of some Enron insiders, particularly Andrew Fastow. Indeed, the insiders' conduct seems particularly puzzling, at least at first glance, given agents' usual incentives. Since agents bear severe penalties in firms if they fail, including loss of job and reputation, but normally do not get the full benefit of success, it follows that they would tend to be more cautious than their employers would want them to be, rather than the reverse.
To begin with, there is a large literature on judgment biases that lead at least some people to tend to be more optimistic about the future and more confident in their judgment and ability to control future events, than would an actor who objectively processed the relevant data.131 For example, traders generally overestimate their ability to judge the true value of the company-i.e., the value of their private information.132 These biases may be bolstered over time by the self-esteem-maximizing device of emphasizing positive returns as an indication of ability and downplaying trading losses as irrelevant. 133 Moreover, even rational people arguably would be more likely than a third party observer to attribute their own failures to luck and their own successes to skill given their tendency to select actions they think will be successful. 134
Actors' judgment biases may depend somewhat on their self-esteem, in the sense that those with the highest self-esteem are the most likely to misjudge their control and skill. But managers' attributes in this respect are not randomly distributed. Donald Langevoort argues that successful firms tend to reward and promote a particular type of individual--one who is highly, perhaps unrealistically, optimistic about the firm's prospects, confident in his abilities,135 seemingly loyal to the firm and its senior management, and distrustful of outsiders.136
These judgment biases and miscalculations might be enhanced by external cues. In particular, legal rules hold that managers are better able to judge the value of their companies than markets. This view emerges most clearly in cases involving takeovers or sale of the company, in which courts have given managers the power to defend against IMAGE FORMULA 76
above-market-value takeovers.137 The courts' acceptance of managers' arguments that market prices are systematically too low is particularly striking given managers' power to release positive information and ability to delay the release of negative information. The law's disregard of markets may have helped confirmed managers' beliefs that their actions were benefiting the firm, regardless of what markets, or earnings, might be saying at the moment.
The above story seems to fit some recent corporate frauds. New methods of doing business such as the provision of alternative markets (Enron), consolidation of long distance telephone service (WorldCom), or the power of downsizing (Sunbeam) produced initial successes and high market valuations based on optimistic estimates of future earnings. Stock prices built on hope were highly susceptible to negative earnings shocks, providing an incentive to prevent these shocks at all costs. Hypermotivated and superoptimistic insiders might be able to persuade themselves that any setbacks were temporary, so that cover-ups need only work for a little while to be successful. On the one hand, they might conclude that markets that spiked in defiance of modest, or no, earnings confirmed their firms' high inherent value, and therefore the validity of their business plans. On the other hand, stocks that fell on bad earnings did not reflect even publicly available information about the firms' abiding value. Thus, hyperoptimistic insiders might be able to convince themselves that earnings manipulations "corrected" the market's misimpressions of their companies.
But even these misjudgments do not seem fully to explain why insiders would risk jail and loss of all of their wealth and future business prospects by engaging in fraud that a rational person would surely realize was likely to be detected, all without apparently having a Vesco-type end game strategy. It has been argued that, once having begun their conduct, insiders managed to deceive themselves that their actions were right.138 But surely at some point insiders would realize that the probability discounted cost of severe sanctions outweighs the potential benefit. Indeed, it would seem that insiders who disregarded the risk of punishment because they were convinced they were right were behaving altruistically rather than greedily.
The solution to the puzzle may lie in the shift of agent incentives that occurs when agents perceive the risk that they may lose everything. This is probably before the agents have committed any wrongdoing, which helps explain why they would engage in wrongdoing in the first place. Insiders face punishment in the form of job and reputation loss even for lawful conduct that fails to meet investor expectations-that is, for their firm's failure to meet investors' earnings expectations.139 At this point insiders may enter a final period in which they are no longer susceptible to potential discipline by their firms or the employment market because failure to distort earnings also will result in loss of IMAGE FORMULA 79
their job and reputation.140 Since insiders are convinced that they are doing the right thing in defending their company's value from destruction by misguided markets, they are also not subject to a significant moral constraint.
As soon as insiders begin engaging in fraud, their incentives change. At this point, insiders risk loss of wealth and even personal freedom unless they continue the cover-up. Indeed, the consequences of discovery may be so severe that even a small chance of success might lead a rational actor to cover up. This calculus may be reinforced by a psychological tendency to prefer risk when choosing between present loss and a chance to avoid loss.141
This brief summary should be enough to indicate that strong measures may be necessary to significantly reduce the risk of future fraud. Insiders who think that they are doing the right thing may be harder to detect and deter than those who are simply greedy. Deterrence that is effective also may be very costly. Moreover, given the shift in incentives discussed above when the end seems near, increasing punishment may actually increase the risk of a cover-up, even as it has little effect on the fraud itself. All of this suggests significant uncertainty about how best to craft the law to prevent future frauds.
2. Investor Heuristics
Why did not securities markets reflect skepticism about the companies' numbers and basic business plans even after public signs of problems emerged?142 For example, no one seems to have considered the implications of WorldCom's meeting earnings projections by small fractions of a penny per share, or why Enron did not owe any taxes.143 These phenomena are at least partly attributable to investor judgment biases that lead them to underestimate the risk that bad things will occur. Investors, like others, may be overly optimistic in the sense of discounting risks, including the risk of fraud. 144 This optimism may have been exacerbated in the present circumstances by a confirmation or conservatism bias that tended to discount evidence contrary to the long-running bubble IMAGE FORMULA 83
market. 145
Assuming these observations are accurate, their implications are ambiguous. On the one hand, given investor biases, perhaps stock prices do not efficiently reflect inherent value, thereby increasing investors' need for disclosure.146 On the other hand, even if factors other than inherent value influence stock prices, this should not matter to investors unless they can consistently outguess the market.147 There is little, if any, evidence that they can.
Moreover, even if investors are irrational, it is not clear what disclosure law can or should do about it. Why would investors want more information about inherent value if their more emotional colleagues will ignore this evidence and take the price in a different direction? Indeed, critics of market efficiency argue that even well-financed arbitrageurs would not want to bear the risk of investing contrary to investor momentum and, for this reason, do not move the market toward efficiency.148 More information alone cannot cure investors of the judgment biases that supposedly lead them to misuse the information.
Perhaps clear disclosures of risks would provide the sort of salient warnings necessary to break through investors' excessive optimism or conservatism bias. But requiring such disclosures carries the cost of forcing firms to make characterizations that are not necessarily supported by the available facts. This could make stock prices more volatile and expose firms and insiders to liability for excessive pessimism. Moreover, mandatory disclosure designed in light of investors' judgment biases may present different problems in bear and in bull markets. As discussed below in Part IV.A, in a bear market investors may be overly skeptical, suggesting that in this context firms need to be more careful with negative than with positive information. It is not clear how a single set of rules can deal with creating considerable uncertainty.
In short, regulation that simply ensures that markets will have more accurate information will not solve the problem of corporate fraud if, as many commentators suggest, investors do not know what to do with the information when they get it. A more promising approach is encouraging investors to be more skeptical of firms' disclosures and more alert to fraud than they seem to have been in the recent corporate frauds. Yet, as discussed in the next part, regulation may actually contribute to these problems. IMAGE FORMULA 86
3. Trust
A few individuals probably could not alone have perpetrated and sustained frauds in large corporations in an atmosphere of skepticism and distrust. Large-scale, successful frauds may require collusion among employees and others.149 For example, there have been reports that WorldCom employees knew of the accounting manipulations years before they were disclosed1 50 and tried to block efforts to expose the fraud.151
Investor complacency in the face of signals of insider fraud could be explained simply as investor trust in insiders' honesty based either on altruistic motives or on a realistic assessment of insiders' basic honesty rather than the judgment biases discussed in Part III.A.2. At least some of the defrauding firms seem to have been helped by a high level of employee loyalty. This may be why, for example, Enron's workers kept their retirement funds invested in the company's stock when they could have diversified, even after problems started appearing and before the company had locked down the retirement funds. One WorldCom worker was quoted as saying after the earnings debacle, "I have never worked for a better company."152 Of course the loyalty may have been instilled, in part, by fear of reprisals within the firm or loss of pensions or stock investments if the firm went under. But some of the loyalty may have been attributable to firms' efforts, in response to arguments by corporate commentators and reformers, to instill a corporate 11 culture" of loyalty in workers and others.153 This provides, among other things, a rationale for managerial opposition to takeovers that would threaten the carefully developed relationship between workers and incumbent managers. 154
The recent frauds reveal a potential dark side to corporate trust. Good corporate governance may require distrustful employees to watch over insiders for the same reason that it needs distrustful auditors, outside board members, securities analysts, and others. This is not to suggest that employee trust is bad. As discussed below, too much monitoring may be counterproductive precisely because it instills distrust.155 Moreover, the recent corporate scandals indicate that monitoring alone cannot prevent fraud and that an atmosphere of hypercompetition can encourage fraud.156 Thus, Enron may have succumbed to a conflict between, or combination of, trust and competition.157 Reducing IMAGE FORMULA 89
the level of trust might have either reduced the risk of fraud by adding more monitoring to the mix, or increased the risk of fraud by exacerbating the culture of competition. The only clear lesson is that the incentives and heuristics of corporate actors are highly complex and not yet well enough understood that they provide a clear basis for any particular regulatory regime.
Investor trust in insiders may be even more complex. It seems unlikely that investors would have the same sort of altruistic motives for trust that workers would have. Rather, investor trust more likely resulted from a calculation that, given high levels of regulation of securities markets, insiders were unlikely to lie or steal.158 Instead of using "trust" to describe both of these disparate phenomena, it would seem clearer to refer to the second phenomenon as investor reliance on, or belief in, the efficacy of regulation.159 The possibility of such reliance suggests the difficulty not only of defining trust, but also of distinguishing judgment biases from reasonable, but ultimately frustrated, expectations. Investors might assume that widespread fraud could not be occurring because the corporate governance framework so highly touted by corporate reformers-a vigilant SEC, public corporations' "monitoring" boards and audit committees, and public accounting firms-was effectively guarding against fraud. This assumption might have been reasonable given what investors had been told, or might have resulted from a potent mix of facts with the confirmation bias discussed above. 160
This view of investor complacency suggests a possible cost of regulation in addition to those discussed below in Part III.C. The market may have been misled not only by defrauding insiders, but also by years of regulators' and reformers' exaggerated claims about the efficacy of regulation. In other words, regulation sends a signal to investors that helps shape their behavior and, specifically, that may mislead them into inaction.161 If this hypothesis is correct, then additional regulation, accompanied by new exaggerated claims for its efficacy, might inhibit markets from self-adjusting to fraud by giving investors a reason for continued complacency.162 For example, Sarbanes-Oxley provisions calling for increased SEC review of corporate filings and a significantly increased SEC budget163 may give investors the impression that the SEC is effectively guarding against fraud. This is an additional reason for concern about the effectiveness of these and other proposed regulatory responses to corporate fraud, discussed in more detail in Part III.B. IMAGE FORMULA 92
B. Effectiveness of Regulatory Proposals
As discussed in Part III.A, the executives and other insiders who perpetrated the frauds apparently acted because of acute judgment biases as well as strong incentives to protect themselves and colleagues from exposure. In other words, these were probably not cases of simple, calculating greed. Some may argue that these circumstances support stronger forms of regulation.164 But this Part shows that past approaches to regulation, monitoring, and liability are seriously flawed. This suggests that intensifying regulation is unlikely to be effective. 165
1. Independent Directors as Watchdogs
As discussed in Part ILA, corporate reformers have emphasized independent directors as a way to curb insider abuse. However, the emphasis on the monitoring board over the last thirty years has demonstrated the inherent limitations on independent directors' effectiveness. Myles Mace, in his famous 1971 study of directors, summarized these limitations as constraints on time, information, and inclination to participate effectively in management.166 Outside directors lack the time to do more than review, rather than make, business decisions. They also must depend on insiders for critical information. With respect to inclination, independent directors traditionally are nominated by insiders and, in any event, generally are selected from the business community to ensure that they will have adequate expertise and, therefore, usually will be unwilling to second guess managers.
Not surprisingly, board independence has done little to prevent past mismanagement and fraud. For example, thirty years ago the SEC cast much of the blame for the collapse of the Penn Central Company on the passive nonmanagement directors.167 No corporate boards could be much more independent than those of Amtrak, which have managed that company into chronic failure and government dependence. Enron had a fully functional audit committee operating under the SEC's expanded rules on audit committee disclosure. 168
The substantial data on boards of directors that has been compiled over the last twenty years offers little basis for relying on regulation of board composition as the solution to corporate fraud.169 The evidence shows that there is no overall positive relationship between various measures of firm welfare, including earnings, Tobin's q, IMAGE FORMULA 97
and stock price, and the degree of independence of corporate boards.' ,u While there is evidence that independent boards may be better at some tasks, such as removing poorly performing managers,171 there is also evidence that independent directors are correlated with worse corporate performance.172 This evidence indicates that insiders may have some value on boards, perhaps in adding important expertise.173 In general, firms seem to be making the right decisions as to how much board independence is appropriate. If anything, evidence of a negative correlation between corporate performance and board independence may indicate that, even prior to the post-Enron regulation, corporations were being forced to err on the side of independence. 174
This data does not necessarily mean that board independence is irrelevant to corporate fraud. First, independent directors arguably are better at certain types of decisions, perhaps including supervising their firms' financial disclosures and relationships with auditors. Enron and related scandals arguably make the data cited above obsolete because they uncovered pervasive fraud that increases the need for this type of supervision. Second, although the overall proportion of independent directors may not affect corporate performance, the independence of certain "trustee" committees such as audit, nominating and compensation committees, may be particularly important.175 Third, post-Enron regulation might usefully tweak the definition of independence so that it precludes at least some directors, particularly those on sensitive "trustee" board committees, from receiving favors such as donations to pet charities with which insiders can buy director loyalty.176 For example, Ross Johnson at RJR-Nabisco sought to buy board member Juanita Kreps by endowing two chairs at her school, Duke, one of them named after Kreps.177 Nevertheless, it seems unlikely that a relatively minor donation could influence a director with a strong reputation to protect.
Even given these caveats, more independence is not necessarily correlated with better monitoring. In order to avoid suspect relationships and connections, corporations may have to appoint more directors from outside the business community. Board members such as law professors,178 with little hands-on business experience and no IMAGE FORMULA 99
formal connection with a company may not be sophisticated enough to spot problems or be able or willing to stand up to a powerful executive. Moreover, there are significant limits on what even the best audit committee can do if, as is typically the case, it meets only a few times a year.179
These problems of board independence may be exacerbated by other proposals to reform the board, particularly including proposals to have directors represent multiple constituencies in the company, such as workers, rather than the shareholders exclusively.180 Encouraging or requiring directors to focus on goals other than financial performance increases the risk that directors will miss signs of misbehavior, if only because of the limitations on directors' time discussed above. Directors who are specifically selected to represent particular constituencies may be useless in protecting against insider fraud because of their lack of business sophistication or their interest only in looking after a particular constituency.181 To be sure, firms might minimize these problems by delegating financial monitoring to a specific audit committee that is focused on this task. But even in this situation, a multiconstituency board might interfere with monitoring by nominating financially unsophisticated directors or by impeding full disclosure to and discussion by the full board. 182
It has been argued that, other things being equal, independent directors would be more attentive to corporate interests if they held stock in their companies.183 Indeed, there is evidence that firms with independent boards that get incentive compensation are more likely to fire bad managers.184 On the other hand, the WorldCom directors were heavy investors in WorldCom, having received both their stock and board memberships in WorldCom's earlier acquisitions of MCI and other companies. 185
Perhaps the problem with board independence is not the specific links between board members and their companies but how they are selected and monitored. Thus, IMAGE FORMULA 102
commentators have suggested that corporations should have professional directors who serve on several boards at the insistence, and subject to the supervision, of institutional shareholders.186 Although this idea has gained little ground in the more than ten years since its publication, perhaps Enron will make the time ripe. It has also been suggested that directors should be more active in their firms' affairs.187 The question is whether these ideas will prove to be more successful than the basic principle of the monitoring board.
Finally, it is important to keep in mind that all of the proposals for better board incentives and greater independence can do no more than improve the prospects for the board's hiring of specialists, particularly including the auditors, to catch fraud. Thus, the discussion of auditors' capabilities in the Part III.B.2 is critical. The board itself, as a part-time supervisory body, is inherently unsuited to reviewing the minutiae of corporate transactions closely enough to spot fraud by committed and astute insiders. Justice Holmes noted this problem more than eighty years ago in holding that outside directors could not be held liable for failing to stop a cashier from taking essentially the entire assets of a bank by drawing checks on the bank and falsely charging the checks to various accounts, including the president's.188 Although the outside directors accepted the cashier's statement of liabilities without inspecting the deposit ledger, the Court said that these directors could trust the cashier because his work had been validated "by the semi annual examinations by the government examiner" and by the president "whose responsibility, as executive officer; interest, as large stockholder and depositor; and knowledge from long daily presence in the bank, were greater than theirs."189 Accordingly, the court concluded that the outside directors "were not bound by virtue of the office gratuitously assumed by them to call in the pass-books and compare them with the ledger, and until the event showed the possibility they hardly could have seen that their failure to look at the ledger opened a way to fraud."190 This seems similar to the Enron directors' failure to see through a complex maze of SPEs, hedging and derivatives, and the WorldCom directors' failure to see that trusted insiders were turning expenses into assets. 191
2. Auditors and Other Detectives
Public accountants undoubtedly contributed to the recent corporate frauds by certifying financial statements that ultimately proved to be fraudulent or at least defective. Lax accounting standards, as in the case of the three percent equity rule for IMAGE FORMULA 107
moving liabilities to off balance sheet entities, arguably contributed to the problem. These rules could be tightened, but this might lead to excess conservatism that causes as many problems as excessively lax standards, including misleading the market and inviting evasion.192 In any event, standards could not have been the whole problem because as long as the basic transactions are recorded, the market ultimately can see through how the transactions are reported. This was certainly the case with Sunbeam, in which most of the bad accounting was uncovered by a Barron's reporter based almost entirely on the company's own disclosures,193 and to some extent was the case even with Enron. 194
The main problem in many cases was that outside auditors did not discover or report the problems. Auditors' failure to report problems that they discover would seem to be amenable to changes in auditors' incentives and to more intensive monitoring of the profession. As discussed above,195 auditors' independence arguably has been compromised by, among other things, their dependence on nonaudit work from audit clients. The current system of self-regulation proved insufficient to counteract these strong economic motives. Accordingly, a strong system of regulation, including prohibition on certain types of arrangements, monitoring by a non-industry-dominated agency, and more independent corporate audit committees, would seem necessary and appropriate. The need for strong measures is indicated by the fact that auditing firms have proven willing to cast aside valuable reputations196 for short-term profits. Notably, a $7 million fine levied against Arthur Andersen because of its mishandling of Waste Management a few months before the Enron scandal broke197 did not provoke Arthur Andersen to significantly change its ways.
Even if the sale of nonaudit services affected auditors' reporting of fraud,198 it is not clear that restricting accounting firms' sale of such services, as under Sarbanes-Oxley, will solve the problem. Such a restriction, standing alone, probably cannot reverse the strong profit-oriented culture that now seems to pervade accounting firms.199 In any event, the restriction is porous. Although the Act forbids the sale of nonaudit services 11 contemporaneously" with audit services by the same accounting firm,200 clients may still have some leverage over auditors that hope to sell nonaudit services to them in the IMAGE FORMULA 110
future or to others with whom clients have contractual or ownership ties.201
For present purposes, the more serious issue is whether even strong regulation will change auditors' practical ability to find corporate fraud when determined corporate insiders want to hide it. In the wake of the WorldCom disclosure, an accounting expert pointed out that accountants do not do "forensic audits" designed to uncover wrongdoing, but rather only sampling audits that may entirely miss the problem.202 The AICPA draft standard on auditing for fraud observes that "[i]dentifying individuals with the requisite attitude to commit fraud, or recognizing the likelihood that management or other employees will rationalize to justify committing the fraud, is difficult."203 The draft notes that "[c]haracteristics of fraud include concealment through (a) collusion by both internal and third parties; (b) withheld, misrepresented, or falsified documentation; and (c) the ability of management to override or instruct others to override what otherwise appear to be effective controls."204
To be sure, there is much auditors can do to spot fraud, including developing crosscheck procedures and identifying risky situations, as is made clear by the extensive discussion in the AICPA's Exposure Draft.205 However, requiring auditors to do significantly more than they are doing now may involve more than just changing their incentives and making them more independent, but also may involve changing the basic scope of what they do. The benefits of increased auditing may not exceed the costs.
If investors cannot rely on auditors to find fraud, it is even less realistic for them to rely on government regulators. Sarbanes-Oxley establishes a Public Company Accounting Oversight Board to scrutinize auditors.206 However, as indicated by the controversy over picking the chair of the board,207 simply designating a new regulatory overseer is unlikely to be a panacea. Sarbanes-Oxley also instructs the SEC to increase its review of financial statements and increases the SEC's budget.208 However, the SEC faces formidable problems in monitoring for fraud.209 The SEC is charged with a wide range of tasks in addition to spotting fraud in financial statements, including oversight of IMAGE FORMULA 112
securities firms, exchanges, investment advisors, and mutual funds, and of market trading, including insider trading. Its staff is perennially too small for these mammoth tasks.210
Sarbanes-Oxley hopes to enlist others to help in the fight against fraud. Lawyers will now be required to report "evidence of a material violation of securities law or breach of fiduciary duty or similar violation by the company or any agent" to executives and possibly to the board.211 The Act also includes strong protection for whistleblowers.212 As discussed below in Part III.C, these rules may be costly because they inhibit efficient information flows within the firm and perversely affect the relationship between corporations and their lawyers. The main point for present purposes is that these rules are also ineffective for purposes of uncovering fraud. Those involved in a fraudulent scheme are unlikely to discuss it with nonparticipants. The new rules may inhibit even innocuous conversations that might have helped indirectly in uncovering frauds by making them fodder for federal litigation and investigations.
Sarbanes-Oxley also includes provisions designed to make executives more vigilant. Section 302 provides for new SEC rules requiring executive certification of facts in firms' securities filings,213 and section 304 requires chief executives and senior financial officers to return their incentive or equity-based compensation or profits from stock sales during a year following the issuance of a financial document that had to be restated due to "misconduct," whether or not they were personally involved in the misconduct.214 As discussed below in Part III.B.4, at least the latter provision may be a significant change in insider liability. As discussed below in Part III.C, the benefits of any such increase in liability must be weighed against potential costs of holding executives responsible for subordinates' acts.
3. Increased Disclosure
The recent corporate frauds were attributable less to firms' silence or misleading than to the falsity of their disclosures. Thus, it is not clear how much difference the Sarbanes-Oxley requirements concerning disclosure of off-balance-sheet transactions, pro forma earnings, and material changes in financial condition215 will make in preventing future fraud. To be sure, burying information in financial statements can make it difficult for individual investors to determine a firm's financial condition. But IMAGE FORMULA 117
misleading legions of analysts, reporters, and others in an active market requires greater opacity. In any event, these provisions deal with yesterday's problem. Recent events have cast so much light on these specific matters that additional wattage is unlikely to make any difference in these particular areas. The next great fraud probably will occur elsewhere.
4. The Marginal Deterrent Effects of Liability
Subpart ILF discusses the notion that Enron and other frauds occurred because of rules restricting securities class actions in the PSLRA and the Supreme Court's Central Bank decision. However, the marginal deterrent effect of pre-PSLRA rules or post-Eiron reforms on corporate actors' conduct, as compared with corporate actors' incentives under current law, is unlikely to be significant.
First, there is no indication that the above law changes significantly reduced federal securities law liability. To begin with, both the number of suits and the size of settlements have increased since the enactment of the PSLRA.216 This is consistent with the incremental nature of the law's changes. The PSLRA clarified the standard for pleading scienter by requiring the complaint to "state with particularity [the] facts giving rise to a strong inference that the defendant acted with the required state of mind."217 These pleading standards did not affect liability for fraud, were the same as that previously adopted by the Second Circuit,218 and nowhere significantly affected plaintiffs' ability to bring claims in the cases of blatant misstatements that have attracted recent attention. The PSLRA also modified joint and several liability by providing that parties who do not "knowingly" violate the securities laws in most cases can be held liable for the portion of the judgment that cannot be collected from other defendants, up to fifty percent of the party's proportionate share of the total damages.219 But this leaves the threat of considerable liability in cases like Enron in which the damages potentially IMAGE FORMULA 121
are in the billions. Central Bank clarified that there was no civil liability for "aiding and abetting" under section 10(b),220 but this liability was unclear before Central Bank, and the clarification leaves the coast clear for liability for misstatements, including accountant certifications.
Second, any effects of reducing securities liability must be considered in the context of the whole set of incentives and constraints facing corporate actors. Even without securities liability, accountants and corporate insiders still face state law liability for fraud and breach of contract. Although state law privity rules may block direct investor suits against auditors, a corporation or bankruptcy trustee can sue a corporation's auditors for failing to spot or report a corporate fraud. For example, Adelphia has sue its former auditor, Deloitte & Touche.221 Although there is authority for letting the auditors off the hook when the insiders lied to them,222 auditor liability in this situation has been recognized when the firm suffered from the misleading certification.223 Moreover, with or without criminal or civil liability, accusations of fraud could ruin insiders' or gatekeepers' reputations.224
Third, any increased likelihood of fraud from reducing securities law liability must be compared with the costs of liability, including those of placing the risks of doing business on corporate agents.225 The market's assessment of these costs may have contributed to the positive stock price effects of the enactment of the PSLRA.226
To be sure, some may question the wisdom of leaving sanctions to state law or extra-legal devices. In particular, some commentators are skeptical of the power of reputational sanctions,227 and the threat of such sanctions obviously did not prevent the recent corporate frauds. But it is important to keep in mind that the issue under consideration is not whether federal liability for securities violations should be reduced, but whether it should be increased from present levels. The recent corporate frauds do not demonstrate that more penalties are appropriate. Rather, they demonstrate that corporate insiders are willing to proceed in the face of potential reputational or other injury because they are driven by strong impulses of loyalty, greed, or fear, and failed realistically to assess the risks of their conduct.228 It is unclear how more liability would have succeeded where other constraints failed. IMAGE FORMULA 124
Whatever changes in liability might have been effective to stop future frauds, it is unlikely that the marginal changes in Sarbanes-Oxley will accomplish this. Lengthening the statute of limitations229 may open the door to a few good cases that otherwise would have been blocked, but also may permit prosecution of suits that should not have been brought because of stale evidence. If business people were not deterred from willful fraud by the thought of substantial jail sentences or fines available under prior law, increasing jail terms or fines for mail, wire, or securities fraud230 and imposing penalties for knowingly certifying false reports231 would not seem to hold much promise.
The requirement of executive certification of issuers' reports in section 302 of the Act,232 which probably has received more publicity than any other provision, seems to provide an important incentive given its potential effect in triggering civil liability for misstatements made with the requisite scienter. But even under prior law, chief executives and chief financial officers had to sign the annual 10-K and the latter had to sign the quarterly 10-Q, with potential direct liability based on scienter.233 Thus, the rule's main innovation is requiring certification of, and therefore creating a basis of liability for, not only financial information, but also the firm's internal controls and the executives' candor with the firm's auditors and audit committee.234
Section 304 includes a kind of vicarious liability provision by requiring executives to return compensation or stock profits following a misconduct-induced accounting restatement.235 Notably, although the provision requires a showing of "misconduct," it does not require that the reimbursing executive have participated in the misconduct, or that the amounts reimbursed relate to the accounting misstatement. Although damages are limited to stock profits or compensation, this does not negate the penalty aspect of the reimbursement. Thus, Section 304 expands executive liability. Part III.C considers whether the costs of greater agent liability exceed the benefits.
C. Costs of Increased Liability and Regulation
Even if the regulatory moves discussed in Part II are more effective than might be expected from the analysis in Part III.13, it is important to consider the potential costs of increased liability and regulation. The potential costs include increasing agency costs by skewing executives' incentives to engage in value-maximizing transactions, encouraging executives to move to less monitored firms and activities, increasing firms' costs of obtaining information about executives' fraudulent activities, and increasing friction in the organization by reducing trust. IMAGE FORMULA 128
1. Agency Costs
Contracts in firms are designed to a significant extent to maximize the benefits and minimize the costs of hiring nonowner agents. In order to operate efficiently, large firms must separate capital raising and control functions. The tradeoff is that nonowner agents who control property have incentives to use their control to benefit themselves rather than the owners. This generates what have been referred to as "agency costs," which include the owner's costs of monitoring the agent, the agent's cost of posting a bond to protect the owner, and residual losses that agents impose on owners despite monitoring and bonding.236 Optimal contracting in the firm involves minimizing these costs without unduly reducing the benefits of using nonowner agents. Even if residual costs are eliminated, total agency costs, including monitoring and bonding costs, might be quite high. Agency costs could be reduced to zero by making the agents owners, but this would involve sacrificing the benefits of specializing functions.
Designing firms' contracts to maximize the net benefits of employing agents obviously is a complex, multidimensional task that requires consideration of each firm's characteristics, and the fact that regulation can affect firms in unpredictable ways. This is apparent from understanding the role of increased liability and regulation in the general context of agents' incentives.
A main benefit of hiring nonowner agents and thereby specializing ownership and management functions is to permit passive owners to invest in diversified portfolios. Owners can tolerate risks associated with specific firms and care only about the risks of their whole portfolios, which if adequately diversified are similar to those of the market as a whole. Thus, a diversified investor should not be concerned about the prospects of his umbrella investment in a dry year because that investor also owns a bathing suit firm. This investor accordingly would not want his agent to be too concerned about firmspecific risks. On the one hand, agency costs can be associated with agent behavior that is too cautious, as well as with behavior that is too risky from the principal's perspective. On the other hand, nonowner agents who do not bear even the small burden of failure that is borne by diversified owners might have an incentive to invest the firm's assets in projects with poor expected returns. Optimal agency contract design involves encouraging the agent to take the owners' interests into account, but not forcing the agent to bear so much of the firm's risks that she is more cautious than the owners would want her to be.
This general discussion suggests a potential problem with Sarbanes-Oxley provisions that encourage executives to police their firm's fraud, including section 302, which forces chief executives to vouch for their firms' financial statements and internal controls,237 and section 304, which requires reimbursement of compensation and stock profits following accounting misstatements.238 The former provision may permit liability on the basis of a court's ex post judgment that the executive certified controls that proved to be inadequate. The latter explicitly provides for liability up to the amount of compensation or stock profits for misconduct by others in the organization regardless of IMAGE FORMULA 132
scienter, and, indeed, even if the executives exercised all reasonable care in monitoring and instituting controls. The provisions therefore require executives to bear some of the risk of fraud formerly borne more cheaply by diversified investors. This may increase rather than reduce agency costs in the sense of causing agents to act more conservatively than owners would prefer.
Executives may respond in various ways to increased liability. They can bear the risk themselves and seek compensation from the company or have the company insure them for the risk. If the executives are protected and reimbursed in full, the liability may have little effect on their incentives. Nevertheless, even fully reimbursed executives may act overly cautiously because of the risk of reputational harm. Moreover, even if the executives' incentives remain the same, the risk shifting may be an extra cost to the firm to the extent that the executives or their insurers are less efficient risk bearers than the diversified investors. This will probably be true for the executives, and may even be true for the insurers if the firm itself would be better able than the insurer to monitor its own executives. Thus, it is not surprising that directors' and officers' liability insurance is becoming significantly more expensive and less available in the wake of Enron and WorldCom.239
Laws that impose extra risks on executives who are not in a position readily to spot fraud may cause executives to respond in several ways to reduce their risks of liability.240 First, they may manage the firm to reduce the potential for liability. One possible approach is to reduce the variance in its expected returns, thereby reducing the chance of an earnings "surprise" that could trigger massive liability. The liability also may affect the categories of transactions executives seek to engage in on behalf of the firm. Because monitors and courts cannot determine with certainty whether a transaction or the corporation's monitoring and approval procedures are efficient, they must rely on signals that may not be completely accurate. For example, derivatives, insider transactions, SPEs, and incentive compensation may have been abused in Enron but may serve valuable purposes in other settings.
Second, liability may perversely affect the disclosure policies executives set for the firm. The general agency problem is that, while executives do not get the benefits of minimizing disclosure costs or of extra clarity of disclosure, they bear the costs of failing to disclose fraud. For example, executives may under-report earnings on the theory that they are less likely to be held liable for overly conservative than for exaggerated earnings reports,241 cover themselves by inundating investors with information, or surround IMAGE FORMULA 135
disclosures with obfuscating hedges and qualifiers. These options would not necessarily better serve investors' interests than managerial inattention to fraud. Apart from the information that is actually disclosed, executives may institute very costly information-- getting procedures in the firm that produce less value for investors than protecting executives from the risk of fraud liability.
The perverse effects of increased agent liability are exacerbated by regulation of agent compensation. Stock-based compensation may have the beneficial effect of aligning agents' and shareholders' incentives, as indicated by researchers' finding positive share-price effects associated with the adoption of stock-based compensation.242 Conversely, regulating such compensation could cause corporate executives to behave more like bureaucrats and less like entrepreneurs by reducing their benefits from risky decisions that pay off for the firm243 and their incentive to work hard to produce profits. This is particularly a problem regarding outside directors, who must take on new governance responsibilities while facing new restrictions on compensation. Executive compensation always has been a tempting target for regulators because concerns about harm to shareholders combine with populist antipathy to wealth disparities.244 This concern was muted when markets were rising, but is emphasized in falling markets during which people seek to assign blame. For example, there have been calls for increased scrutiny and regulation of executive stock options and particularly for requiring firms to account for these as expenses, like other compensation.245 Also, Sarbanes-Oxley IMAGE FORMULA 137
bans certain loans to executives, including loans for buying the company's stock.246 On the other hand, enabling insiders to share in the upside without taking downside risk could help align their interests with those of the shareholders. Excessive risk-aversion is always a problem with agents because, unlike shareholders, agents invest their nondiversifiable human capital in the firm.247 Now it is especially a problem in light of executives' increased liability risks discussed above in this part.
2. Resource Allocation
Increased liability and regulation not only may increase agency costs, but also may affect the flow of resources to particular firms. Firms whose earnings are more variable, that are in lines of business in which the accounting standards are more uncertain, or that use now-suspect business practices such as hedging and derivatives, are all subject to increased liability risk. This effect is exacerbated by the possibility of increased securities law and other liability for insiders, auditors, and outside directors of firms whose disclosures are now subject to greater scrutiny. Other things being equal, increased liability and regulation will reduce "suspect" firms' value and ability to attract financial and human capital. This is inefficient to the extent that these firms do not, in fact, pose a significantly higher risk of fraud.
Increased liability risk may have two types of effects on the flow of human capital. First, if reward remains constant while the risk increases, the affected jobs will attract less risk-averse parties for whom the cost of risk is relatively low. Second, the best executives who now head large firms may be enticed to companies that can offer greater risk-adjusted rewards because they are not subject to full-fledged disclosure regulation. This may inefficiently attract managers to nonpublic firms even if they would be better suited to larger, more specialized firms. Also, firms in riskier industries, with higher likelihood of liability, may be less able to attract managers.
Increased securities law liability risk might also affect firm structure. Sarbanes-- Oxley's enhanced disclosure and other requirements effectively impose a tax on public ownership of stock. Firms can avoid this tax by buying their shares and "going private," thereby freeing themselves of 1934 Act reporting requirements. Depressed share prices make such transactions even more attractive. If the costs of being private would outweigh benefits for a particular firm but for the liability risk, these can be considered deadweight costs of the liability rules.248 Moreover, a trend toward going-private transactions could reduce the available investment options. The effect might be exacerbated if going-private transactions were concentrated in particular industries that will have particularly high liability and auditing costs under Sarbanes-Oxley. This effect would be ironic in light of the law's intent to lure investors back into the market.
Increased liability and regulation also might have significant effects on auditing firms, with indirect effects on their clients. Forcing accounting firms to stop using audit services to sell nonaudit services to clients might increase the price publicly traded firms IMAGE FORMULA 141
must pay for audit services. Some might respond by hiring small firms, including auditonly subsidiaries of traditional accounting firms, whose prices are lower because they do not internalize the full cost of liability. This might lower overall auditing standards, unless the new oversight of auditing firms required by Sarbanes-Oxley really is effective. Also, new liability and regulatory constraints might make it harder for auditing firms to attract the best people from jobs in investment banking, consulting, and other fields that pay more for the same skills.249 Reinstituting vicarious partner liability for firm debts would help ensure auditor responsibility but would exacerbate the potential brain drain from the auditing profession.250
3. Information Costs
Post-Enron regulation has two kinds of effects on information costs. On the one hand, the regulation directly increases firms' costs in part by requiring them to spend more to get information. In particular, new auditor regulation significantly increases firms' audit fees as well as their costs of dealing with and producing information for auditors.251 Although firms may get more and better information, the question is whether the increased benefits outweigh the costs.
The regulation also may indirectly increase firms' costs of obtaining the same quantity and quality of information by prohibiting business practices that, in effect, subsidize information gathering and disclosure. The moves toward auditor and outside director independence are intended to reduce job-preservation incentives for fraud by severing other links between monitors and the monitored firm, such as by prohibiting consulting work by auditors or revolving doors between auditing and client firms, requiring periodic change of auditors, or forbidding independent directors from having other associations with or receiving other benefits from the company. The problem is that monitors' other links with firms increase their access to information. If the fully independent monitor can duplicate the connected monitor's information, requiring greater independence just increases the firm's cost of obtaining information. Regulation and liability that impose higher standards of investigation on public firms forces these firms and their investors to incur the cost. In that case, whether investors are better off depends on whether higher investigation costs exceed the benefits of better monitor incentives. In other cases, prohibiting some links between monitors and firms, such as the performance of nonaudit services, may block "knowledge spillovers" that give monitors access to valuable information.252 Also, insiders might have less incentive to shade the truth when dealing with consultants who help them do their jobs than when dealing with IMAGE FORMULA 145
auditors whose sole function is disclosure.253
There may be similar effects at the director level. The closer the relationship between the director and the particular firm or industry, the more insight the director is likely to have into the firm's problems and the quality of the information the board is receiving. This relates to the board's ability not only to advise managers, but also to uncover fraud. For example, directors with inside knowledge of the company may be better able than outside directors to see through ambiguous, opaque, or misleading financial statements because they have enough background to understand the kinds of tricks insiders might be playing.
The costs and benefits of independence may vary from one situation to another. First, some types of independence may have higher net benefits than others. Monitors may have better or cheaper access to information if they perform other tasks for the company, but not if they simply receive more compensation. Second, the amount of information provided by monitors' other links with firms may vary according to the complexity or uniqueness of the firm's business. Third, the cost-benefit tradeoff may depend on how many levels of monitors the firm has. For example, it may make sense to require complete independence at the auditor or director level, but the total costs of independence may exceed the benefits if it is required at both levels. Thus, a fully independent audit committee might provide the optimal mix of independence and access without prohibiting nonaudit services.254 These variables and uncertainties support a flexible and contractual approach to regulation,255 including requiring disclosure of, but not otherwise regulating, auditor independence.256
4. Distrust
The level of trust among those working in a firm can significantly affect the firm's operating costs by, among other things, increasing the flow of information among personnel and the extent to which people in the firm are willing to rely on informal assurances of reciprocal fair play, rather than insisting on costly regulatory and contractual protection.257 As discussed above,258 high levels of trust may disarm IMAGE FORMULA 150
monitors. Conversely, mandating complete independence of monitors risks creating an adversarial relationship between insiders and outsiders that may reduce both the efficiency of day-to-day management and the monitors' access to information.
First, requiring staffing of boards of directors by people with no other ties to the company may remove insiders who can assure executives that they will actually receive the rewards the firm explicitly and implicitly promises them.259 Insiders confronted by adversarial, outsider-dominated boards may insist on upfront, nonperformance-oriented compensation or move to insider-dominated firms. Also, executives who work for outsider-dominated boards may spend time unproductively on self-promotion and political activities in order to curry favor.260
Second, more monitoring and liability may work counter to their intended goals by inhibiting the detection and reporting of fraud. Discovering fraud depends on communication among various levels of the organization which, in turn, depends to some extent on trust. For example, a worker whose conduct was at least arguably innocent or defensible in the light of applicable rules, but nevertheless hurt the firm, might reasonably fear punishment by overly zealous monitors or whistleblowers and therefore may be reluctant to communicate with them.261
Third, insiders who are closely monitored may become less trustworthy, thereby increasing the need for costly legal sanctions and constraints. Legal sanctions may "crowd out" parties' motivations to engage in trustworthy or benevolent behavior in the absence of such sanctions.262 More monitoring and penalties also may change the "social meaning" of behavior, with a similar result of making insiders less trustworthy and increasing the need to rely on legal sanctions.263 For example, forcing insiders to deal with adversarial outsiders might induce them to see their jobs as a kind of game, or Dilbert comic strip, in which they must outwit clueless outsider directors, courts, and overly scrupulous auditors. Fostering cooperation between insiders and outsiders alters the connotation of behavior such as shading accounting numbers from succeeding at a game to impeding the attainment of the firm's legitimate objectives. Finally, establishing stringent liabili