Bank behavior and how it relates to bank fragility and systemic risk have been in the spotlight since the 2007-2009 financial crisis. Regulators claim that bankers’ compensation structures played a role in encouraging behavior which contributed to the financial crisis. Despite this, we know little about how finance industry executives are compensated. Executive compensation research typically does not distinguish finance industry executives from non-finance industry executives, and on many occasions it excludes executives in financial industries. Our new article, How Are Bankers Paid?, fills this void, studying how executive compensation in the finance industry differs from that of non-financial firms during and after the financial crisis.
The article addresses the following questions: How are bankers paid, how is the pay linked to performance, and how does its design compare with the design of non-financial executives’ compensation? To answer these questions, we use multiple datasets to fully understand and describe the compensation design in banks and compare it with a sample of size and performance-matched control firms in non-financial industries.[1] In addition to highlighting the distinct aspects of bank executive compensation, we speak to the extent to which differences in compensation are related to differences in leverage and risk.
Our results fall into two broad categories: (i) how are financial executives compensated compared with their non-financial counterparts, and (ii) how is the compensation adjusted for changes in risk? First, we find that executives in financial firms are paid significantly less than executives in the control sample. Specifically, the total compensation of a bank executive is more than 20 percent less than that of a non-financial executive at a similarly-sized firm. We find an even starker difference for executives at insurance firms; they make less than half as much as executives at matched non-financial firms. Second, we find that performance targets in financial-firm compensation contracts are not adjusted for changes in risk.
In terms of the split between equity and non-equity (cash) pay, a larger fraction of financial firm executive pay is non-equity. Specifically, the fraction of non-equity pay to total pay is approximately 10 percent higher for financial firm CEOs as compared with CEOs of control firms. Regarding pay-for-performance sensitivity (PPS), we find no difference between treated and control CEOs. However, we find that executives below the rank of CEO in treated firms have a higher PPS than those in control firms.
In terms of pay structure, we find that a higher portion of financial firm compensation is tied to accounting metrics like return on equity (ROE) and earnings per share (EPS) while a lower portion is tied to stock price. A larger portion of banker pay tied to accounting metrics, especially short-term accounting metrics, may encourage bank CEOs to increase tail risk in a way that increases short-term performance while also increasing the probability of an outcome with losses more than three standard deviations from the mean. There are many possible reasons why bankers have such performance clauses in their compensation contracts. First, banks are required to mark more of their assets to market, which means that accounting measures of performance may be more informative about true economic performance in banks than in non-financial firms. Second, bank shareholders may want to increase the value of the safety net provided by deposit insurance. Many years ago, Nobel Laureate Robert Merton recognized that deposit insurance is like a put option given by taxpayers to the bank, so the bank benefits from increasing the value of this put option. One way the bank can do this is by increasing risk. The banks’ shareholders thus may not be averse to greater tail risk. Third, bank managers could be exploiting poor governance.
To determine which of these reasons is responsible for the observed pay design, we differentiate between firms with high and low institutional ownership, as it has been shown that (higher) institutional ownership is related to better governance. We find that banks with high institutional ownership have a larger portion of their compensation tied to stock price performance. This suggests that governance affects compensation design. However, we do not find a significant difference in the portion of pay tied to short-term accounting metrics (i.e. ROE and EPS) at the treated versus the control firms.
We next relate short-term pay to bank risk and leverage for treated firm executives. We do this to determine if performance targets are adjusted for the level of risk taken. Interestingly, we find no relationship between the performance targets in bankers’ pay contracts and the level of bank risk. This implies that bankers may be tempted to meet their short-term performance goals by increasing bank risk, as there is no cost associated with doing so. Consistent with this theory, we find a positive relationship between the fraction of bank CEO pay tied to short-term accounting metrics and bank risk measured as either aggregate risk or tail risk.
Lastly, we evaluate the effects of the TARP program and its associated “pay czar.” After TARP, we find a reduction in the portion of banker pay linked to EPS. However, we find no reduction in the fraction of banker pay linked to ROE. It is important to note that ROE is a function of both leverage and EPS. An increase in leverage (risk) will increase ROE, even if EPS remains constant.
Executive compensation design, the performance goals used therein, capital structure, and the level of risk are all endogenously co-determined. As such, we only document relationships between them. While our analysis does not show a causal impact related to an exogenous variable on an endogenous choice, it does allow us to better understand performance goals used in bank executives compensation contracts.
ENDNOTE
[1] In our analysis, we also consider insurance firms as part of our treated sample. Throughout the article we perform separate analysis, specifically we: i) compare bank executives to non-financial firm executives and ii) compare insurance firm executives to non-financial firm executives.
This post comes to use from professors Benjamin Bennett at Tulane University’s A.B. Freeman School of Business and Radhakrishnan Gopalan and Anjan V. Thakor at Washington University in St. Louis’ John M. Olin Business School. It is based on their recent article, “How Are Bankers Paid?,” available here.