Lucian Bebchuk is professor of law, economics and finance and director of the corporate governance program at Harvard Law School, and Holger Spamann is co-executive director and fellow of this program. This post builds on their joint paper “Regulating Bankers’ Pay.”
It is now widely accepted that compensation structures in financial firms should be devised to avoid excessive incentives for risk-taking and that doing so requires tying executive compensation to long-term results and preventing cashing out of large amounts of compensation on the basis of short-term results.
What long-term “results” are we talking about though? We propose that risk-taking incentives could be improved by tying executives’ pay not only to the long-term payoffs of shareholders but also to those of preferred shareholders, bondholders and taxpayers insuring depositors.
In examining how executive compensation can affect risk-taking in financial firms, attention has focused on distortions that can arise from the ability of executives to cash out large amounts of compensation before the long-term consequences of risk-taking are realized. The importance of eliminating such distortions, which was first highlighted in a book, “Pay without Performance,” that one of us published with Jesse Fried five years ago, has become widely accepted in the aftermath of the financial crisis.
But there is another type of potential distortions that should be recognized. Bank executives’ payoffs have been insulated from the consequences that losses could impose on parties other than shareholders. This source of distortions is separate and distinct from the short-termism problem; and it would remain even if executives’ payoffs were fully aligned with those of long-term shareholders.
Equity-based awards, coupled with the capital structure of banks, tie executives’ compensation to a highly levered bet on the value of banks’ assets. Bank executives expect to share in any gains that might flow to common shareholders, but they are insulated from losses that the realization of risks could impose on preferred shareholders, bondholders, depositors or the government as a guarantor of deposits. This gives executives incentives to give insufficient weight to the possibility of large losses and therefore provides them with incentives to take excessive risks.
How could pay arrangements be redesigned to address this distortion? To the extent that executive pay is tied to the value of specified securities, such pay could be tied to a broader basket of securities, not only common shares. Rather than tying executive pay to a specified percentage of the value of the common shares of the bank holding company, compensation could be tied to a specified percentage of the aggregate value of the common shares, the preferred shares and all the outstanding bonds issued by either the bank holding company or the bank. Because such a compensation structure would expose executives to a broader fraction of the negative consequences of risks taken, it will reduce their incentives to take excessive risks.
Indeed, even the above structure would not lead bank executives to internalize fully the adverse consequences that risk-taking might have for the interests of the government as guarantor of deposits. To do so, it would be necessary to broaden further the set of positions to whose aggregate value executive payoffs are tied. One could consider, for example, schemes in which executive payoffs are tied not to a given percentage of the aggregate value of the bank’s common shares, preferred shares and bonds at a specified point in time, but rather to this aggregate value minus any payments made by the government to the bank’s depositors, as well as other payments made by the government in support of the bank, during the period ending at the specified time.
Alternatively, one could consider tying executive payoffs to the aggregate value of the bank’s common shares, preferred shares, and bonds at the specified time minus the expected value of future government payments as proxied by the product of (i) the implied probability of default inferred from the price of credit default swaps at the specified time, and (ii) the value of the bank’s deposits at that time.
Even if such schemes are not used, however, tying executive pay to the aggregate value of common shares, preferred shares and bonds will already produce a significant improvement in incentives compared with existing arrangements.
Similarly, to the extent that executives receive bonus compensation that is tied to specified accounting measures, it could instead be tied to broader measures. For example, the bonus compensation of some bank executives has been based on accounting measures that are of interest primarily to common shareholders, such as return on equity or earning per common share. It would be worthwhile to consider basing bonus compensation instead on broader measures like earnings before any payments are made to bondholders.
Recognizing this problem highlights the limits of corporate governance reforms for fixing the design eliminating excessive risk-taking incentives in banks. Concerns about excessive risk-taking have led legislators and regulators, both in the United States and abroad, to adopt or propose various corporate governance measures, such as say-on-pay votes, aimed at improving pay-setting processes and better aligning pay arrangements with the interest of banks’ shareholders.
Although such measures can discourage some inefficient risk-taking that is undesirable from bank shareholders’ perspectives, they cannot be relied on to eliminate the incentives for excessive risk-taking that arise from moral hazard: The common shareholders in financial firms do not have an incentive to induce executives to take into account the losses that risks can impose on preferred shareholders, bondholders, depositors, taxpayers underwriting governmental guarantees of deposits and the economy. This moral hazard problem is at the heart of the extensive body of banking regulations that we have. Consequently, regulatory encouragement or even intervention may be needed to eliminate all excessive risk-taking incentives.
Be that as it may, any attempt to eliminate excessive incentives for risk-taking requires a full understanding of the sources of such incentives. Such understanding requires focusing not only on the length of executives’ horizons but also on the definition of long-term results to which executives’ interests should be tied.