Steven N. Kaplan is the Neubauer Family Professor of Entrepreneurship and Finance at the University of Chicago Booth School of Business.
IN THE AFTERMATH of last year’s banking crisis, world leaders are looking for ways to see that this never happens again. G-20 leaders, the Obama administration, and, apparently, the Federal Reserve have focused on pay practices at financial firms as being a key cause of the recent disaster and have proposed restricting bankers’ pay. These restrictions largely amount to reducing short-term cash bonus payouts, increasing the use of restricted stock and options, and requiring the executives to hold the restricted stock and options for a period longer than the usual four-year vesting period. The big question is: “Will it work?”
IS PAY THE ROOT OF ALL EVIL?
Before instituting what is likely to be a quite invasive regulatory system, perhaps we should take a closer look to see if past pay structures really were at the heart of our recent problems. The poor pay practice explanation for the crisis implies the following:
1. Top bank executives were rewarded for short-term results with large amounts of up-front cash pay.
2. Bank executives did not hold sufficiently large amounts of stock to align their interests with those of shareholders.
3. Executives with more short-term pay and less stock ownership should have had the greatest incentive to take bad and excessive risks and, so, should have performed worst in the crisis.
In recent work, Rudiger Fahlenbrach and Rene Stulz (2009) test these implications by studying the CEOs of almost one hundred large financial institutions from 2006 to 2008. They start in 2006 because that is a good candidate for the point at which financial firms took on the risky positions that led to the crisis.
In 2006 the mean CEO took home $3.6 million in cash compensation, which represented less than half of total compensation. The larger share of pay was in restricted stock and options. At the same time, the mean CEO held $88 million worth of the firm’s equity and options. In other words, the CEO took home $3.6 million in cash, on average, while leaving more than twenty-four times as much in his or her firm. It seems unlikely that the up-front cash pay provided much of an incentive for the average CEO to knowingly take bad or excessive risks that would jeopardize his or her much larger equity stakes.
Consistent with this conclusion, CEOs lost a lot in the crisis. From 2006 to 2008, the average CEO lost $31 million in his or her holdings of the firm’s stock, again, dwarfing any gains from cash compensation. The CEOs lost large amounts on their options as well.
Finally, and again in contrast to the excessive risk assumption, Fahlenbrach and Stulz do not find that, for banks, CEOs with less equity generated worse stock performance. If anything, bank CEOs with more equity had worse performance.
These results led Fahlenbrach and Stulz (2009) to conclude that bank “CEO incentives cannot be blamed for the credit crisis or for the performance of banks during that crisis.”
David Yermack (2009) points out that many prominent financial executives lost “small fortunes” in 2008. He adds that “farther down the ladder, most mid-and upper-level managers of financial companies also lost a significant amount of their net worth in 2008.”
Yermack concludes that “the recent scrutiny of executive pay seems to stem from an odd mix of envy and vengeance, unsupported by facts or theories.” Yermack, a noted researcher on CEO pay, has written several articles highly critical of specific CEO compensation practices—such as corporate jet usage. Nevertheless, his conclusion on the relation of CEO pay to the financial crisis is diametrically opposed to those of advocates for tight pay regulation.
Ing-Haw Cheng, Harrison Hong, and Jose Scheinkman (2009) find some evidence consistent with the story of short-termism leading to excessive risk. They find that financial firms that paid higher total compensation relative to their size had modestly higher stock volatility and significantly lower stock returns from 2001 to 2008.
Their results, however, are largely driven by insurance firms. With insurance firms excluded, their results are only marginally economically and statistically significant. They do not point to compensation as a major cause of the crisis.
Furthermore, the specific pay restrictions under consideration—dramatically increasing the period between the time stock compensation is issued and the time it can be sold—would not have stopped the longer-serving executives like Cayne of Bear Stearns, Fuld of Lehman, and Lewis of Bank of America from selling their shares before the crisis. These executives received large amounts of stock and options in the 1990s, essentially all of which would have been eligible for sale under most of the proposed plans.
If front-loaded incentives were not the reason that bank CEOs got in this mess, what did cause the crisis? John Taylor (2009) points to highly expansionary monetary policy in the years leading up to the crisis. Douglas Diamond and Raghuram Rajan (2009) cite the so-called capital glut—large inflows of external capital from China and much of the developing world. Charles Calomiris (2009) highlights the role of the political system in inflating the banking sector and real estate prices, particularly the subprime sector, through low-income housing mandates implemented by Fannie Mae, Freddie Mac, and others. Ruling out compensation practices hardly leaves us at a loss for culprits in the recent debacle.
Perhaps the following episode can also provide some insight. A well-known partnership, led by a charismatic CEO, raised outside capital to fund a new investment vehicle. The partnership put up 10 percent of the money in the investment vehicle. That 10 percent represented 50 percent of the partnership’s capital. Most of the partners’ net worth was invested in the partnership, so the partners had substantial “skin in the game.” And the partners did not take large fees or any other meaningful short-term compensation out of the partnership, completely in line with the current recommendations for pay reform.
Unfortunately, the partners leveraged the first investment vehicle by investing in a second and third vehicle. Then the stock market declined substantially, almost bankrupting the partnership.
The partnership was Goldman Sachs in the late 1920s and early 1930s. The investment vehicle was the Goldman Sachs Trading Corporation. It took Goldman more than twenty years to recover its capital. When asked to explain why Goldman had made such a risky bet, Walter Sachs did not mention compensation. Instead, he simply responded, “to conquer the world” (Endlich 1999).
The point of the Goldman Sachs example is that restrictions on pay are not necessarily effective in the face of a bull market and the subsequent crisis. One suspects they also are ineffective in stopping CEOs who want “to conquer the world.”
THE DOWNSIDE TO PAY REGULATION
While the proposed pay restrictions are unlikely to stop the next financial crisis, they are likely to damage the financial sector. To see why, it is important to understand why bankers are paid so much.
Over the last two to three decades, technological change and increased scale have led to much greater productivity and much higher incomes for those at the top of the income distribution. Corporate executives manage larger companies. Investors manage much larger sums of money. Entertainers and athletes access larger audiences. Lawyers oversee larger transactions and larger cases. Joshua Rauh and I (see Kaplan and Rauh 2009) have found that pay for all of these groups has increased much more than pay for the average worker.
Given this reality, the best bankers, traders, deal makers, and so on will work for the companies that are able to provide the most attractive compensation packages. Greater pay regulation will drive more of the most talented away from regulated banks and toward hedge funds, private equity funds, boutique investment banks, and other unregulated investment firms. We have already seen top talent leave AIG and other Troubled Asset Relief Program (TARP)-constrained institutions.
The potential talent drain is likely to be exacerbated as pay regulations impose a one-size-fits-all regime on all bank employees. Why impose restrictions on a deal maker who earns a large fee for putting a merger together that does not impose any risk on the bank after the transaction has closed? The same is true for a trader who makes money on spreads rather than from taking on risk. These types of employees will be penalized for no good end under the proposed pay regimes.
Perhaps more troubling is that pay restrictions open up a Pandora’s box of other restrictions. As Professor Bebchuk (2009) has admitted elsewhere, politicians are often more interested in setting limits to total compensation than in designing the optimal form of compensation. Demagogic politicians believe such restrictions make them look good in the eyes of angry voters.
We saw this earlier in 2009 when restrictions on total pay at TARP recipients were inserted in an unrelated bill as part of the Obama stimulus package. And we see this in the pay packages recently imposed by the Obama administration pay czar at AIG, Citigroup, and Bank of America. Gabriel Sherman (2009) reports that Feinberg’s decisions on pay levels were explicitly motivated by political considerations. It is naïve for Bebchuk to argue that only pay structures will be regulated, not pay levels.
One might also worry that the financial sector will move increasingly to a Fannie-Freddie style arrangement in which banks use their capital to advance the pet projects and political fortunes of lawmakers in exchange for lawmakers guaranteeing sympathetic regulatory treatment.
A BETTER SOLUTION
Banks have a special function in the economy that does warrant a special role for the government: not in setting pay but in imposing effective capital requirements. A better solution would impose higher and pro-cyclical equity capital requirements on banks combined with a requirement to raise contingent long-term debt—debt that converts into equity in a crisis.
Banks would be required to have some minimum amount of equity capital, say 10 percent of total capital—much like current requirements. Investment banks like Bear Stearns and Lehman got into trouble because they had too little equity capital—far less than 10 percent. Regulators might consider imposing pro-cyclical equity requirements—increasing the equity percentage in boom times in order to offset losses in the inevitable bust times.
In addition, regulators would require banks to issue an additional amount of capital—say, 10 percent—in the form of long-term debt that is forced to convert into equity if the bank and the overall banking system get into financial difficulty.
Lloyds recently issued such contingent capital bonds. The bonds automatically convert into a predetermined amount of equity if Lloyd’s tier-1 equity capital falls below 5 percent of assets. Although Lloyd’s did not do so, one might add a systemic trigger, that is, the bonds do not convert unless the banking system as a whole also has seen a decline in equity capital.
Banks will have to pay a higher interest rate on contingent capital bonds than on straight long-term debt, since debt investors will face the true cost of capital, not the government subsidized one. Over time, however, the extra interest rate is likely to be small, as the imposition of higher capital requirements and contingent capital will significantly reduce the likelihood of ever reaching the triggers.
Bankruptcy is a terribly messy process for large financial firms with stakes in a variety of complex transactions, operating in multiple jurisdictions. If this contingent capital structure had been in place before the crisis, troubled banks would have been recapitalized with the capital of the contingent debt holders while avoiding the delays, complications, and legal posturing that are inevitable in the formal bankruptcy process. These debt investors, not the government, would have bailed out the banks and investment banks. The financial crisis would have been substantially smaller, if it had occurred at all. Contingent capital and pro-cyclical equity requirements are also effective in reducing the potential damage done by financial sector firms that want to conquer the world.
REFERENCES AND FURTHER READING
Calomiris, Charles. 2009. “Banking Crises and the Rules of the Game.” NBER Working Paper No. 15403. Available at www.nber.org/papers/w15403.
Cheng, Ing-Haw, Harrison Hong, and Jose Scheinkman. 2009. “Yesterday’s Heroes: Compensation and Creative Risk-Taking.” Working Paper, Prince ton University. Available at www.princeton.edu/~joses/wp/yesterday.pdf.
Diamond, Douglas W., and Raghuram Rajan. 2009. “The Credit Crisis: Conjectures About Causes and Remedies.” NBER Working Paper No. 14739. Available at www.nber.org/papers/w14739.
Endlich, Lisa. 1999. Goldman Sachs: The Culture of Success. New York: Alfred A. Knopf.
Fahlenbrach, R., and R. Stulz. 2009. “Bank CEO Incentives and the Credit Crisis.” Working Paper, Ohio State University. Available at www.nber.org/papers/w15212.
Kaplan, Steven, and J. Rauh. 2009. “Wall Street and Main Street: What Contributes to the Rise in the Highest Incomes?” Review of Financial Studies 23, no. 3.
Taylor, John. 2009. “The Financial Crisis and the Policy Responses: An Empirical Analysis of What Went Wrong.” NBER Working Paper No. 14631. Available at www.nber.org/papers/w14631.