Infectious Greed

by Frank Partnoy

  A spokesman for General Electric was smug about accusations of earnings manipulation, saying, “You have to have the earnings to be accused of smoothing them.”49 General Electric, in particular, received special treatment. Many of the SEC’s cases had been prompted by financial journalists’ exposés of wrongdoing. Ultimately, those reports led to the bankruptcy of Enron and WorldCom, and prosecutions of officials at both firms. But, notwithstanding numerous reports since 1994 of careful timing of capital gains, and creative restructuring charges and reserves, at General Electric, the SEC did not bring a case against that firm or its officers.

  In hindsight, it is obvious that stock options were a key factor in the major corporate fiascos of the mid-to-late 1990s. As noted in the previous chapter, the rise of stock options was due primarily to regulatory changes: a $1 million cap on tax deductions for executive salaries, with an exception for performance-based compensation (which included stock options), and an accounting rule that said stock options were essentially free.

  Professor Kevin J. Murphy, one of the leading scholars studying executive compensation, has estimated that option grants in industrial companies more than doubled from 1992 to 1996.50 Compared to the average worker’s pay, CEO compensation in 1996 was three times higher than it had been before 1990. Including stock options, the average CEO made 210 times more than the average worker.51

  There were many arguments in favor of options, most notably that they aligned the incentives of managers and shareholders. However, options only aligned incentives in one direction. Managers who held options had a limited downside, whereas shareholders were not so protected. That meant that both managers and shareholders benefited when shares went up, but managers were more willing to gamble because of their limited downside. This was especially true when companies, such as Cendant, repriced executives’ stock options when the stock went down, lowering the price at which they could buy stock, or when companies reloaded executives’ stock options by issuing new options when an executive cashed in some options before their maturity.

  In other words, executives who owned stock were aligned with shareholders—for good or for ill—but executives who owned options were less aligned with shareholders because their downside was limited. The difference between a CEO with stock and a CEO with options was like the difference between a captain who was prepared to go down with the ship, and a captain with the only lifeboat (or, better yet, a helicopter) set aside for times of danger. The best way to align the incentives of executives and shareholders was to put them in the same boat, which stock options did not do.

  Some economists worried that CEOs were too risk-averse, and therefore favored stock options precisely because they encouraged CEOs to take on more risk. This argument was more speculative. Before the mid- 1990s, CEOs had taken on plenty of risk without the incentives of huge, 10-year stock-option grants—Walter Forbes, Dean Buntrock, Al Dunlap, and Martin Grass were all examples. Moreover, when the stock price declined, options could lead CEOs to take on too much risk.

  Besides, if companies really wanted to “incentivize” their CEOs, they could do so in two ways that made much more economic sense. First, they could pay CEOs more in stock, based on their performance. For example, Jack Welch had received hundreds of millions of dollars of restricted stock—stock that could not be sold in the market for a period of several years. This stock carried many of the regulatory benefits of stock options, but forced executives to think more about downside risk.

  Second, they could pay CEOs in stock options that were based on how well their company did relative to its competitors or an index, so that CEOs were rewarded for outperforming the market and not merely for being in office as the tide rose.

  Many companies offered restricted stock, but the percentage of stock held by CEOs—ignoring stock options—actually declined throughout the mid-1990s.52 Only one in 1,000 firms offered indexed options—based on how well its stock did relative to other stocks. Why? The only plausible explanation was legal rules: these two alternatives were better economically, but they forced companies to record an expense. (Unlike standard stock options, outperformance options had to be disclosed as an expense, because the 1972 accounting rule had created exceptions only for options with a fixed-exercise price.)53

  Stock options had other drawbacks. They diluted the value of shares, because if the price of a stock went up and executives exercised their right to buy shares, there would be more shares outstanding and each shareholder would own a smaller fraction of the company. Companies were required to disclose this dilution in their financial statements, although many companies tried to minimize dilution by repurchasing from shareholders enough stock to set aside for executives when they exercised their options. The problem with repurchases was that they required companies either to spend cash that could have been used in more valuable projects, or—if they wanted to repurchase shares and do the projects—to borrow money, thereby increasing their debt.

  Options also were a very expensive form of compensation, because the cost of the stock options granted to executives, in terms of how they were valued in the market, was much greater than the value to the executives themselves. An economically rational CEO would much rather have $1 million in a diversified portfolio (or in cash) than $1 million of stock options. In order to pay a CEO $1 million of value (to the CEO), it would cost the company either $1 million of cash or, say, $2 million of options. The reason was that CEOs had so much invested in their companies—reputation, other compensation, pensions, and so forth—that a concentrated position in stock options was just about the least valuable form of compensation they could receive. One of the central tenets of modern finance is that a diversified portfolio is worth more than a concentrated one, because it is less volatile. CEOs valued diversification, like anyone else. Some studies showed that, in rough terms, a typical CEO valued stock options at half their market value.54 In other words, it was twice as expensive for companies to pay their CEOs in stock options as it was to pay them cash.

  Another odd consequence of stock options was that they advantaged CEOs who were already rich or who liked risk, because they placed a higher value on stock options than their middle-class, risk-averse counterparts. Again, the explanation was diversification. A rich CEO would have only a small part of her wealth in options, whereas a middle-class CEO would have a much more concentrated position. According to Professor Murphy, a CEO with 90 percent of her wealth in her company’s stock options would value those options only a fifth as much as a wealthy CEO with the same options, if those options represented only half of her wealth. In other words, companies might prefer to hire rich CEOs because they didn’t have to pay them as much.

  The list of problems with stock options goes on and on. Because options do not receive dividends, CEOs who receive large stock-option grants have an incentive to reduce dividends. In fact, the dividends paid by public companies decreased dramatically during the 1990s. Before that, dividends—not capital gains—accounted for three-fourths of the returns from stocks, and companies signaled that they were doing well, not by increasing reported earnings to meet expectations, but by paying high cash dividends.55 (Another reason not to pay dividends is that they are taxed, but this feature did not change during the same period.)

  But the biggest problem with options granted during the 1990s was that they were hard to value. Most options granted during the 1990s had long maturities, and restrictions that limited the use of option-pricing models, such as the Black-Scholes model. For example, 83 percent of options had 10-year terms, and most executives could not cash in the options for several years. That meant the maturities of most executive stock options were more than twice as long as the options Bankers Trust and Salomon Brothers had been unable to value. Options models did not work well for long-maturity options, primarily because the variables—especially the key variable, volatility—change so much over time. Moreover, the restrictions on cashing in options were very difficult to incorporate into an options-pric
ing model. By one estimate, the true value of a typical executive stock option was only about half of the value estimated by the Black-Scholes model.56

  In other words, the opponents to the FASB rule requiring companies to include stock options as an expense—particularly Senator Joseph Lieberman—had a valid point: these options were quite difficult to value. Consultants to corporate boards gave a wide range of valuations for options, and some directors used a rough rule of thumb that the value of a typical option was about one-third of the value of the stock. But given the uncertainty about value, was it really a responsible business decision to give executives options in the first place? Even with a rough rule of thumb, it was easy to see that the stock options were of very high value. By this measure, one million options on a stock with a price of $60 were worth about $20 million. Walter Forbes, Dean Buntrock, Al Dunlap, and Martin Grass had millions of stock options, each. And they were not alone: with the dramatic increase in stock-option grants to corporate executives in the mid-to-late 1990s came greater temptation for corporate executives to commit fraud.

  The corporate frauds of the mid-to-late 1990s likely would not have occurred without the dramatic increase in stock-option grants to corporate executives. Three conclusions from this period are indisputable: first, legal changes in the early 1990s led companies to give unprecedented amounts of stock options to their CEOs; second, those stock options became much more valuable if companies inflated profits and hid losses for a few years; third, many companies, in fact, engaged in accounting fraud.

  As stock options were on the rise, other new financial instruments were coming back to life after a brief post-1994 hibernation. The over-the-counter derivatives markets continued to grow, and with them the concern that corporate executives who wanted to manipulate earnings might move beyond simple accounting tricks to use derivatives in schemes of unimaginable complexity.

  One reason these new instruments hadn’t appeared in a major accounting scandal yet was that top-level corporate executives—whose involvement typically was required—didn’t understand them. Risk magazine, a leading financial-industry publication, harshly concluded, “Top management, whether corporate or otherwise, failed to understand the nature of the products being used in their treasuries or trading operations.”57 Whereas the derivatives scandals of the early 1990s had come from the bottom (salesmen and traders at Bankers Trust, First Boston, and Salomon; junior treasurers at Gibson Greetings and Procter & Gamble), where at least some of the people understood derivatives, the more recent accounting scandals had come from the top, where most CEOs did not.

  Even several years later, corporate boards still had a lot to learn about financial engineering. According to partners at several major accounting firms, most board members had only a cursory grasp of their risk exposures in 1996.58 In response to the fiascos of 1994, most boards had approved policy statements dealing with derivatives, but not much more. A traditional audit was of little value in assessing financial risk, and members of corporate-audit committees often didn’t even know the right questions to ask. Corporate directors attempted to insulate themselves from liability for any future problems by implementing corrective procedures and controls, but that didn’t mean they actually understood any of the details themselves.

  Corporate directors of many of the top companies in the world—from Aetna to Dow Chemical to General Electric—did not spend much time focusing on these new financial risks. Barbara Scott Preiskel, a member of the board of General Electric, told Derivatives Strategy magazine, in 1996, “At GE, we’ve never had a full board discussion about derivatives.”59 That was a shocking admission, given that General Electric had made (and lost) huge amounts of money on new financial instruments, and its financial subsidiary, GE Capital, had been involved in derivatives markets for many years. Clayton Yeutter, former chairman of the Chicago Mercantile Exchange and a member of half a dozen major boards, agreed that corporate directors rarely discussed or understood new financial techniques: “It’s a rare occurrence where the board is in a position to dispute the presentation of a senior financial officer on risk management matters.”60 No wonder so many CEOs were walking away with $100 million of stock options.

  Directors of companies felt—perhaps quite rationally—that their role was to put procedures in place, and then rely on others to do their jobs. But, as a result, the responsibility for ensuring that shareholders were not exposed to undue or undisclosed risks fell on senior executives and their accountants—precisely the same people who had just completed a round of unprecedented accounting fraud.

  There were some signs, even in the early accounting scandals, of new financial instruments lurking in the background. To give one example, one of Cendant’s first acts after the merger of CUC and HFS was to issue $1 billion of new financial instruments bizarrely called FELINE PRIDES. A few months later, owners of these FELINE PRIDES would become embroiled in litigation surrounding Cendant, as yet another victim of the company’s fraud. But the Byzantine structure of these instruments—the very nature of the FELINE PRIDES—raised troubling questions about the state of U.S. financial markets.

  The name FELINE PRIDES was a mouthful of acronym. “FELINE” stood for Flexible Equity-Linked Exchangeable Security; “PRIDES” stood for Preferred Redeemable Increased Dividend Equity Securities. The bizarre acronym had an even stranger history, one that illustrates how much financial markets had changed during recent years. The lineage of FELINE PRIDES reads like an Old Testament family tree, with a stray cat thrown into the mix. If this discussion seems esoteric, remember this: by 2002, most public companies used these types of instruments—undoubtedly companies whose stock most investors own.

  In the beginning, there was equity and debt. Equity consisted of shareholders who owned a company. Debt consisted of bondholders whom the company owed. Together, equity and debt supplied all of the capital available for a company to invest. Debt was at the bottom of a firm’s capital structure; equity was at the top. Bondholders received interest and were repaid their principal at maturity. Shareholders received dividends plus any other increase in the value of the company.

  Then, preferred stock was created, a hybrid of equity and debt. Preferred stock had an infinite life, like equity, but was paid a fixed rate, like debt. Some preferred stocks could be converted into regular stock (called common stock) at a future date, at the option of the preferred stock-holder. Other preferred stocks had a cumulative dividend—an obligation that accumulated when it wasn’t paid (unlike a common-stock dividend, which was at the company’s discretion). Thus, preferred stock sat in the middle of a company’s capital structure, above debt but below equity.

  The key questions about preferred stock involved legal rules. When the rules for debt and equity were different, how would preferred be treated? For example, interest on debt was tax deductible, but dividends on equity were not. Could a company deduct payments to its preferred shareholders for tax purposes? What about the rating agencies, which were always critical to the financial markets, as the first chapters demonstrated? In assigning ratings, they looked at the ratio of debt to equity. How would they assess preferred stock? And what about the all-important accounting rules? Analysts compared companies by looking at the relative amounts of equity and debt on their balance sheets. Where would preferred stock fit?

  Financial engineers at investment banks were very good at answering these questions. They designed novel types of preferred stock to take advantage of tax deductions and favorable credit ratings, and to minimize the amount of debt disclosed in financial statements.

  First, creative bankers noticed that common stock was really composed of two pieces: dividends, plus any change in stock price. They separated these two pieces—just as they split apart the interest and principal payments on mortgages—by putting them into a common-stock trust (one well-known 1980s version was called Americus Trust), which issued two securities: one conservative security, that received dividends plus some of any increase in the stock pri
ce; and one riskier security, that received any additional increase in stock price. Such instruments were used in the United States, Europe, and Japan.

  In 1988, Morgan Stanley created a security called PERCS, based on the more conservative piece of the Americus Trust deals. PERCS stood for “Preferred Equity-Redemption Cumulative Stock,” and resembled a preferred stock, with cumulative dividends that were higher than the dividends paid on common stock (a “perk” for the investor, in case that reference wasn’t obvious). The key twist was that, in three years, PERCS automatically converted into common stock, according to a specified schedule. For example, PERCS would convert into one share of common stock if the common stock was at $50 or lower, but convert into fewer shares if the price was above $50, so as to limit the upside of PERCS. Essentially, an investor buying PERCS committed to buy a company’s stock in three years, and also sold some of the upside potential of that stock by selling a three-year call option (similar to those Andy Krieger was trading at Bankers Trust). The company bought the three-year call option from the investor, and paid the investor a “premium” in the form of a cumulative dividend for three years.

  PERCS were quite complex, and one might imagine that only the most sophisticated companies and investors would use them. But the first PERCS deal, in July 1988, was done by Avon Products, Inc.61 Avon’s common stock had fallen to around $24, but was still paying a $2 dividend. Avon wanted to cut the dividend to $1, but doing so would infuriate investors and drive down its share price even more. The brilliant solution was to offer to exchange common shares for PERCS, which would still have a $2 dividend but would convert into common shares in three years. The price cutoff for converting PERCS into common stock declined during the three years, wiping out any gain from the additional dividend—but that was much more subtle than simply slashing the dividend by $1. Investors didn’t complain. If Avon’s common stock were below $32 in three years, the PERCS would each receive one share of common stock; otherwise, they would get less.62 The company received some favorable regulatory treatment: the rating agencies treated PERCS as equity, as did Avon’s accountants, although the dividends on PERCS were not tax deductible.