Conventional wisdom has it that compensation in the financial industry is apparently responsible for much of the credit crisis. For instance, Paul Krugman states that “reforming bankers’ compensation is the single best thing we can do to prevent another financial crisis a few years down the road.” Unfortunately, the facts are stubborn and they do not fit this conventional wisdom.
Rüdiger Fahlenbrach and I study the incentives of bank CEOs before the start of the crisis and how the performance of banks is related to these incentives. We just posted the latest revision of our paper on SSRN ((Bank CEO incentives and the credit crisis, available at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1439859).
The paper shows that the value of the shares held by CEOs in the companies they manage was roughly ten times the value of their total annual compensation in 2006. Such large holdings dwarfed annual bonuses and a large share of these holdings was held voluntarily – i.e., not in the form of restricted shares. With such holdings, it would have made little sense for CEOs to take actions that somehow led to higher bonuses at the expense of shareholder wealth.
In the revision of the paper, we add important new results concerning the role of cash bonuses. Our sample includes 95 large banks for which we have detailed information on CEO compensation, option holdings, and equity holdings. The average base salary for the CEOs in our sample is $760,000 in 2006. On average, the base salary is less than 10% of a CEO’s compensation. Most of a CEO’s compensation is therefore incentive based. On average, a CEO’s cash bonus was 2.8 times the CEO’s cash salary. Bonuses tied to the achievement of goals in terms of accounting performance are paid both in cash and in shares. In 2006, on average, 40% of the bonus was paid in shares and 60% was paid in cash. Interestingly, the fraction of the bonus paid in equity was larger for better-performing banks. We investigate whether the performance of banks during the crisis is related to the size and composition of bonuses. When we do control for bank characteristics, we find no evidence that banks that paid a higher cash bonus relative to the salary in 2006 performed worse during the crisis. In other words, there is no evidence that greater cash bonuses led CEOs to take more risks that worked out poorly during the crisis.
The revision of the paper also has new data where we investigate the relation between the compensation of non-CEO top executives and bank performance during the crisis. We find no relation between the compensation of non-CEO top executives and bank performance during the crisis.
Some observers have argued that our reasoning is not correct because bank executives accumulated vast amounts of wealth that were not tied to the performance of their banks. In particular, Professors Bebchuck, Cohen and Spamann show that the top executives at Bear Stearns and Lehman received large cash payments and sold large amounts of shares in the years that preceded the crisis. However, irrespective of the size of their wealth not tied to the performance of their firm, it is still the case that CEOs had powerful incentives to increase the value of their firms’ shares.
Our analysis focuses on whether incentives created by compensation plans of CEOs help explain the performance of bank stocks during the crisis. The bottom line of our paper is that our evidence is inconsistent with the view that these incentives were detrimental to the interests of shareholders. We do not address the issue of whether these incentives were not appropriate from the perspective of the stability of the financial system or whether the level of compensation was appropriate. Setting compensation within financial firms is extremely complicated because assessing performance in relation to the risks taken is difficult and because compensation decisions have to be made in the midst of what is often intense competition. For sure, errors were made. However, more regulatory intervention in the pay process is unlikely to make the U.S. financial system safer and it would seem that better tools are available to make the financial system safer than regulating compensation. The last time regulators attempted to address executive pay, limits on the tax deductibility of cash salaries were established. Unintended consequences involved a dramatic growth in option compensation, which is now viewed as part of the problem. What will the unintended consequences be this time?