Accounting For Bad Accounting: An Interview with John Coffee

John Coffee is the Adolph A. Berle Professor of Law at Columbia Law School, and a leading expert on corporate crime. He was Reporter for the American Bar Association for its Model Standards on Sentencing Alternatives and Procedures and for the American Law Institute's Principles of Corporate Governance. He has been listed by the National Law Journal as one of "The 100 Most Influential Lawyers in the United States." He is the author and editor of numerous articles and case books. .

Multinational Monitor: What is the auditor's accountability in the Enron case?

John Coffee: That is the most significant question in the Enron case. It will always happen that some managements will overstate their income, or understate their liabilities, or otherwise engage in earnings management. The question is not why they do it. The question is why the gatekeepers have failed to detect it.

This looks like a case where both the auditors and the other gatekeepers had plenty of warning that they were entering very treacherous waters. Both the auditors and the securities analysts failed the shareholders of Enron.

The auditors did not withdraw their certificate until just about a month before the company failed, when they belatedly decided that three years of earnings had to be restated. And of the 20 or so securities analysts following Enron, up until the date of bankruptcy, only one at the last minute put a sell recommendation on the stock.

MM: Isn't that generally the case now, that the securities analysts say, "Buy, buy, buy?"

Coffee: It is generally the case, but it is the product of conflict of interests.

Analysts today are chiefly compensated based on a transactional basis. Their earnings come from the investment banking side of the business. The investment banking firm pays analysts because they are useful in helping the investment banking firm attract and retain clients for underwritings.

Part of the underwriters' marketing pitch may be an implicit promise that their "star" analyst will bless the company and put a strong buy recommendation on its stock.

Once you create a competitive process where one investment bank firm can obtain a client from another, so that the investment banking firm with the top analysts can steal clients, then you create a world in which the analysts are very highly paid, but paid in terms of the their ability to please the management of companies, and cause those companies to switch investment banking firms.

That's not a world that we used to see. Twenty years ago, analysts were paid by the "buy" side of the business, namely, the brokerage side. Today, the source of funding comes not from the buy side, but from the investment banking side, which is trying to sell deals. That makes the analyst much more subservient to the corporate management over which he is supposed to watchdog.

You are not going to be a very effective watchdog, if the person you are supposed to watch feeds you.

MM: On the auditing side, the key gatekeeper was Arthur Andersen. What was their conflict?

Coffee: Arthur Andersen is characteristic of the big five accounting firms. They are not that different. I don't want to suggest that this is a firm that is grossly different from or worse than the other big five accounting firms. All of them now receive income both from auditing and from management advisory services. In the case of Arthur Andersen, during the last year that they represented Enron, they received something like $27 million for auditing services, and more than that, $28 million for management advisory services, which often involves software consulting and other high growth income.

If you are looking at Enron as a client 末 from a standpoint of a chief executive of the auditor, or that of the local branch chief of Arthur Andersen down in the Houston office 末 what you see is a client that is producing fairly staid, straight line revenues for auditing and also very high growth revenues for management consulting services. You may have seen that grow from a few million dollars to $28 million this year. And you may anticipate the prospect in the future that this could grow to $100 million or more.

Management consulting services is a very high growth market, whereas auditing is a fairly limited market, that doesn't have much growth potential. There are only so many large companies you can service. The auditing income doesn't change much year to year.

In that kind of world, you may see much more profit potential in trying to develop your advisory relationships with the company than as serving it as an auditor. To maximize the income in that relationship, you have to be extremely friendly. And you have an incentive at least in many cases to be willing to subordinate your responsibility as an auditor to your desire to market your consulting services. That's the conflict that bothered the SEC when the SEC put out a fairly tough proposal to regulate the independence of auditors.

That came out early in 2000. Under federal securities laws, the auditor has to be independent of the client to provide the annual auditing of the firm necessary to be both listed on an exchange and to file its 10Ks and 10Qs with the SEC 末 to continue as a publicly held company.

To be an independent auditor, the SEC said, you should not have other relationships with the company that would give you this basic conflict of interest, this desire to subordinate the auditing role to other business relationships.

Let me explain in simpler terms. The classic auditor of 20 or 30 years ago was a firm that might have had 1,000 clients, each of them paying between one tenth of one percent to one percent of the firm's overall revenues. In that world, where you had many clients, each paying a small annual revenue flow, there was very little incentive to compromise your own auditing judgment, to subordinate what you thought was the necessary standards required by generally accepted accounting principles, to what the company wanted.

You had very little incentive, because at most you were going to gain on the profit side the continuation of that one tenth of one percent to 1 percent of your revenues each year. Whereas, if you made a mistake, or if you acquiesced in improper accounting, it could blow up in your face and damage your reputation, which was far more valuable to you than the prospective income from that client.

Today, the client is a much more lucrative asset, capable of providing much more income. And the auditing relationship is not nearly as important.

Indeed, auditors today, in their own professional journals, often view the auditing role as a kind of portal of entry through which they gain the attention of senior management, such as the treasurer, the comptroller and the chief financial officer. Through them, the auditors try to market more lucrative services, such as software consulting, management advice and so on. In that kind of world, you are inherently subject to a greater conflict of interest.

In the Old World, an auditor had to be irrational to risk its reputational capital, its personal prestige, simply to appease a client that was responsible for only less than one percent of its revenue.

In the New World, that client is the source of so much income, that the temptation is so much greater.

MM: Why the change?

Coffee: It's very profitable and the risks in this practice declined over recent years. Inherently, as an auditor, you have your foot in the door with the auditing client, and its much easier to sell once you have your foot in the door.

Also, at the same time, there was a decline in the risk of liability. Auditors face lesser legal risk because of a Supreme Court decision, Central Bank of Denver, that abolished aiding and abetting liability. And they face lesser risk because of the Private Securities Litigation Reform Act (PSLRA), which, through its pleading rules, made it much harder to even state a cause of action against an auditor.

And ultimately auditors face less liability because that same statute (PSLRA) substituted proportionate liability for joint and several liability. That means that even if an auditor were found liable for securities fraud, in most cases, it would only be liable for a relatively small proportion of the total judgment 末 10 to 20 percent on average. All of those factors coalesce to mean that auditors don't have to fear the risk of liability as much as they did in the past.

Next, couple the legal side with the marketing side. It is now possible to sell much more lucrative services to the audit client than you could ever sell as simply its auditor in the past. Audit revenues are essentially limited 末 they grow at a modest rate per year. Whereas software consulting revenues are not only more lucrative, but they can increase at a hyperbolic rate as the company becomes more and more dependent upon the auditing firm as its source of software.

So, in the world of auditing, the benefits of acquiescence have gone up, and the potential legal liabilities, or costs, have gone down. And that produces a world in which the auditor is rationally more likely to acquiesce to what the client wants to do than it would in the past when the costs were higher and the potential benefits were lower.

MM: Chairman Levitt had effective control of the SEC. Why didn't he just impose his rule, barring provision of consulting services to the audit client, over the objections of Congress?

Coffee: He would have lost his budget. There was incredible pressure placed on the SEC by both houses of Congress.

The chairmen of both of the Congressional committees to which the SEC reports expressed their strong displeasure with Chairman Levitt's proposed, fairly tough audit independence rule. Chairman Levitt was essentially told that he was going to lose his budget if he pushed it.

MM: Is this Enron specific, or systemic?

Coffee: The industry wants to contain the scandal and limit it to being about bad actors at Enron. There may well have been bad actors at Enron.

I'm not sure that we can yet fully apportion the blame or state from afar who should be indicted, and who shouldn't be, or who should be sued and who shouldn't be.

Still, the public policy issue is broader than Enron. The reasons that gatekeepers become acquiescent and lax are structural. They don't just involve individual moral failures, although there may have been that element too.

The circumstances that caused Arthur Andersen to fail at Enron may have also caused it to fail, as the auditors for Sunbeam and Waste Management, two other egregious cases in which they were involved recently. And it is not a unique position involving Arthur Andersen. I'm not in a position to say that they are better or worse than Ernst & Young or Pricewaterhouse or Deloitte and Touche.

All of these firms are approaching the point where they are going to be highly conflicted in the future and they are going to have much reduced incentives to in effect call a halt to "gray" or dubious practices by their audit clients. They instead are in a position where, if there is any way to say that these practices comply with an accounting principle, they have a strong incentive to do that.

Over the last three years, an Arthur Andersen study found, the number of earnings restatements by reporting companies has risen by 47 percent. Between 1998 and 2000, the numbers go from 158 to 233. That is a 47 percent rise over three years.

Restatements generally suggest that the accountants have made an error. Now, on an individual case, I can't tell you whether that error was the product of simply an honest difference of opinion or an honest mistake, but on an aggregate basis, I have to say that that much of an increase in failures by auditors or the company to file their results in accordance with generally accepted accounting principles looks to me like a statistically important trend.

While I can't talk about individual cases easily, I can say that on an aggregate basis, it seems to me like the simplest explanation for why auditors are making more and more mistakes is that they are more and more conflicted.

Errors will occur. But errors occur most when you have a conflict of interest 末 one that gives you an overriding desire to be cooperative with management. Auditors down deep are not simply supposed to be cooperative with management. They are supposed to protect the interests of investors.

MM: So, what should be done?

Coffee: While we should certainly investigate this individual case, and dispense justice, we have to expand our range of vision beyond dealing with the Enron management, and recognize that the gatekeepers of corporate governance are increasingly conflicted.

The simplest things to do are to have a much more effective auditor independence rule that at some point would not consider an auditor to be independent of management if it had extensive business dealings with that management 末 such as by serving as a consultant on a host of non-audit issues.

Reasonable persons can disagree over how to define that rule. But the rule that the SEC settled under pressure for last year was too weak. It was a pale imitation of what had been originally proposed.

MM: On the question of private litigation, what are the chances that these companies will be held liable by the class action bar?

Coffee: Enron clearly could be forced to settle for a high number because you are in a very weak position once you have restated three years or more of earnings and claimed $500 million in earnings that you now acknowledge you never earned. There clearly were mistakes. I think it would be difficult for Enron to escape liability if it could still be sued.

But Enron is essentially immune from securities litigation. The filing of a petition in bankruptcy stays all pending litigation against the company.

So, Enron won't be sued. Even if it were sued and found liable, the tort judgment would come last in bankruptcy 末 it would come below every other claim.

MM: What about Arthur Andersen?

Coffee: Arthur Andersen and the directors of Enron can be sued. They are not quite as inviting a target, because they don't have quite as deep a pocket and they have more defenses.

Still, they have deep enough pockets. So we will see major litigation against Arthur Andersen. From the evidence we know, this is the kind of case that could produce a very large settlement.

MM: Even with the weakened state of the tort law against auditors that you mentioned earlier?

Coffee: In cases that really do look egregious, auditors have been held liable or at least have settled prior to going to trial. Ernst & Young settled for something like $300 million in the Cendant case. That was a case where some of the officials of Cendant and one Ernst & Young employee were indicted. Once you have cases that are that egregious, you can get over the pleading hurdle.

Also, the proportionate liability provisions say that you can get the full judgment, not just the proportionate share, if you can prove that the auditors had actual knowledge of the material omission or the material misrepresentation.

And certainly there is that possibility in this case. And that is what the plaintiffs are going to seek to prove. But they will have a much easier time pursuing management, which will have some insurance, and will each be potentially able to contribute $50 million or so to a settlement. But it still is difficult to sue the auditors, unless you can prove at the outset of the case enough information to raise a strong inference of fraud.

And whether or not they can meet that standard with regard to the auditors, is still an open question.

MM: Enron created partnerships that moved debt off the books. What's your understanding of what happened?

Coffee: Many companies try to use off-balance sheet financing so that they don't have to show liabilities on their balance sheet 末 so that they look less leveraged and thus are able to sell their bonds or get bank debt at a lower interest rate. The more leveraged you are, the higher your risk and the higher the interest rates you must pay because the interest rates will reflect the risk of the enterprise. Companies do try to hide liabilities.

They often do it through mechanisms such as lease financing. And accounting rules, to some extent, combat this by sometimes requiring long-term charges to be capitalized.

Thus, if you have a long-term financing lease, you will have to show it on your balance sheet, not simply as in this year's payment, but rather as the capitalized value of that lease over the 40 years that it exists. That will put a lump sum on the balance sheet.

Enron was ingenious about trying to find ways to exploit the ambiguities and the limitations of those rules. They in particular tried to treat the partnerships as independent entities, even though they might have liability for it.

In some cases, they were looking to an accounting rule that says 末 if you own 95 percent of an entity, but some other independent organization or company owns the other 5 percent 末 then you can treat that entity as independent of you. And its liabilities do not have to be capitalized on your balance sheet.

Now that is a rule that may be excessive as it is. But Enron exploited that by finding third parties that would buy a 5 percent equity interest in some of its affiliates, thereby allowing the liabilities of those affiliates to be kept off the balance sheets.

The one instance in which Arthur Andersen has admitted an error was their failure to detect one of these entities that allegedly owned a 5 percent share of an Enron affiliate and was actually being guaranteed by Enron so that it didn't have its own equity at risk. That would require capitalization of that liability on the Enron balance sheet if it had been detected.

We are in this context about to witness the familiar dispute that often occurs in these cases in which the auditors say about the client: "They lied to us, they deceived us." I'm not in a position to know the facts here. That's for the courts and the SEC to determine. Nonetheless, this was a rule that was probably too weak to begin with.

I don't think it is in the interest of investors to allow a company to set up a partnership which issues large obligations and claims that it doesn't have to be capitalized on the company's balance sheet simply because 5.1 percent of the affiliate is owned by someone else. It would be much more sensible to capitalize 95 percent of that obligation on the parent's balance sheet.

But that's a problem with the accounting rules themselves. And there is also the problem of whether Arthur Andersen was overlooking some fairly clear warning signals.

After all, Enron had something like 3,500 subsidiaries. In my 30 years in corporate law, I have never heard of another company that had 3,500 subsidiaries. That was a structure that was so extraordinarily complicated as to be suspicious in and of itself.

MM: When the investigators and regulators are sitting down to determine whether to bring criminal charges in this kind of case, in terms of bringing justice to the victims, does a criminal charge help or hinder?

Coffee: Bringing a criminal case does change the balance of advantage in the private litigation.

In Cendant, criminal charges were brought and the private plaintiffs ultimately obtained a $3.1 billion recovery in the securities class action. That stands as the all-time record.

In general, the private plaintiffs feel that if the government indicts management, that puts management in the position where they cannot really defend the case.

The managers are less interested in the private case than whether or not they will go to prison.

In terms of whether or not this will pay off all of the employees who have lost their life savings, I don't think there is any way that the management of Enron, even if their entire assets were somehow located and seized, would be able to cover even 40 percent of the total losses.

MM: What about the auditors?

Coffee: Remember, the auditors are a personal services firm. We are talking about theoretical market losses of over $30 billion. There is no auditing firm that could pay a $3 billion judgment.

© Multinational Monitor January/February 2002


list of articles