Last summer, in an effort to correct what have been broadly viewed as distorted pay incentives in the financial services sector, AIG instituted a new compensation policy for senior management that bases pay on a mix of 80 percent debt and 20 percent stock.
The debt-based pay innovation aims to solve incentive problems that uniquely apply to financial institutions because of the high proportion of leverage that the firms take on — 90 to 95 percent of bank and financial institution balance sheets are debt, on average. This makes financial institutions much riskier than nonfinancial institutions, where only about 40 percent of the average balance sheet is made up of debt.
But deposit insurance provides an explicit guarantee in the event of a failure, and the widely accepted (if increasingly questioned) notion that banks are too big to fail provides an implicit guarantee against failure, effectively subsidizing debt financing and shifting the consequences of risk taking from banks and their shareholders onto insurers, the government, and taxpayers.
This idiosyncratic nature of banks means that equity-based pay and other off-the-shelf solutions used at nonfinancial firms shouldn’t be applied to financial firms. “At more highly leveraged institutions with risky debt, maximizing equity value creates incentives for excess risk taking,” says Professor Patrick Bolton.
AIG’s new pay structure is one of a number of proposals that aim to reign in CEO pay by using debt to correct misaligned incentives. Doing so could work, Bolton says, so long as such proposals consider the firm’s overall value — or whole enterprise value — rather than only equity value.
At nonfinancial firms, actions taken to maximize shareholder (equity) value also maximize the whole enterprise value, as long as corporate debt is mostly safe from default. “But even if debt is risky, firms have to take into account how actions that increase the value of equity might decrease the value of debt,” Bolton says, pointing out that in a highly-leveraged institution with risky debt, equity is like a call option. “Increasing risk or volatility always raises the value of a call option. But it also lowers the value of the debt, because debt holders are the first to be exposed and will pay a disproportionate amount of the loss if the firm goes under.” Therefore, when debt is risky, shareholders would like to be able to commit to not take excessive risk. For financial firms the problem is amplified by the fact that the value of debt may not go down in proportion to the higher risk because of the explicit and implicit guarantees provided by the government.
Yet there’s a simple solution to control excess risk taking, Bolton says. “We have market measures of risk. It’s just a question of exposing CEOs to those measures.”
Bolton worked with Hamid Mehran of the Federal Reserve Bank of New York and Joel Shapiro of the University of Oxford to test the notion that some forms of debt can be used to pay CEOs at financial institutions while better aligning incentives to avoid excessive risk taking. In conjunction with a bank’s stock price, the researchers propose using a bank’s credit default swap (CDS) spread, or the deviation of the CDS spread relative to the market average. The CDS spread measures how the market perceives the risk to have gone up on the debt of the institution. Much like a thermostat regulates temperature, a bank’s CDS spread would regulate CEO compensation: as the spread widens, CEO compensation would decrease; as the spread contracts, compensation would increase.
The researchers first conducted a theoretical analysis of CEO pay and risk taking, modeling the interplay between shareholders, debt holders, depositors, and executives, showing that a combined equity- and CDS spread–based compensation structure could curtail excessive risk taking. They complemented their theoretical analysis with an empirical study of bank CEO compensation for 27 banks in the lead up to the financial crisis, comparing CDS spreads over the same time period. The research confirmed not only that CDS spreads captured risk accurately, but also that lower CDS spreads were associated with higher fractions of total CEO pay in the form of deferred compensation and more debt-like CEO compensation in general.
Bolton is less enthusiastic about using other forms of debt to correct CEO incentives. “Paying CEOs by asking them to hold part of the bank’s debt is problematic because you expose them to a lot of risk that has nothing to do with their actions, like aggregate risk with respect to interest rates,” he says. “If the Fed raises rates, then debt goes down — and so does CEO pay — which has nothing to do with what the CEO is doing. We agree that enterprise value, not just equity value, is what matters. But we think a CDS spread is a better way of aligning incentives because it more directly identifies the risks that the CEO and the firm are taking.”
Shareholders could be reluctant to embrace the use of CDS spreads or other forms of debt as a payment vehicle since they get more value if the bank takes more risk. “But if you remove some of the implicit and explicit guarantees that encourage risk taking — and we are moving in that direction — then this potential conflict between regulators and shareholders will shrink,” Bolton says.
Correcting CEO incentives toward risk taking would correct trader incentives for excess risk taking. “If CEOs have skin in the game,” he says, “they will put more pressure on their risk managers to report risk accurately and to monitor traders more.”
Banks in turn wouldn’t have to work as hard up front trying to measure risk the way they currently do. “Nor would banks be forced to set aside capital in proportion to those measures, which are noisy and unreliable. This could streamline the way risk is measured and correct the weaker aspects of Basel III,” says Bolton, referring to the most recent round of accords that set international standards on bank laws and regulation.
“The weakest link in financial firms is typically between the risk management unit and the traders, but CEOs paid based on a CDS spread would have incentives to implement new internal risk management mechanisms,” Bolton says. “All it takes is a few banks to take the lead.”
Patrick Bolton is the Barbara and David Zalaznick Professor of Business in the Finance and Economics Division and a senior scholar at the Jerome A. Chazen Institute of International Business at Columbia Business School.