I was surprised at the number of commenters on yesterday’s post who thought that executive compensation is a red herring or a political talking point or “populist pablum.” I agree that some of the outrage over compensation by TARP recipients is a bit overblown. But I also think that the incentives created by current compensation structures were a serious contributor to the financial crisis – which was, after all, largely about banks taking one-sided risks because of asymmetric payouts (lots of upside, limited downside) – and that fixing those incentives is an important task for regulatory reform.
So, I decided to call on some reinforcements. Lucian Bebchuk, a leading researcher of executive compensation (book; important paper discussed here), and Holger Spamann have a new paper called “Regulating Bankers’ Pay” that discusses precisely this issue. They conclude not only that regulation of banks’ executive compensation would be a good thing, but that it may actually be better than the traditional regulation of banks’ activities.
Section II (PDF pages 11-28) lays out, with simple examples worthy of a Beginners post, what I thought was already generally accepted (but apparently isn’t): leverage, combined with the bank holding company structure, combined with compensation in the form of stock options, combined with deposit insurance, combined with the implicit guarantee on uninsured liabilities, creates large incentives to take excessive risks – defined as actions that have a negative expected value for the bank’s assets, but a positive expected value for bank executives. First of all, in a highly leveraged financial institution, shareholders already have the incentive to take excessive risks, because their downside is limited. This is amplified for executives holding stock options, whose downside is even more severely limited. Finally, explicit or implicit guarantees on liabilities reduce the incentive for creditors to adequately monitor banks’ activities.
As a result, Bebchuk and Spamann argue that executives’ incentives should be tied not to the value of shareholder’s equity, but to the value of all of the bank’s assets. Wait a second, though – isn’t the whole point of corporations that managers’ incentives should be aligned with those of shareholders? Yes, that is one consideration. But there are two other considerations that matter.
First, banks are unusual in that a large portion of their liabilities is guaranteed by FDIC deposit insurance, which already helps distort executives’ incentives as described above. As the insurer, the government needs to protect itself from moral hazard – and one way of doing that is reducing managers’ incentives to take actions that are good for shareholders but bad for the insurer.
Second, as we should now know, the incentive to take excessive risks, when shared across all the largest banks, is a major contributor to systemic risk. A systemic crisis leads to both government bailouts to protect non-guaranteed creditors, and to severe collateral damage for the economy at large. Therefore, the government should attempt to reduce those incentives, both to protect taxpayer money and to protect the economy.
Traditional regulation attempts to deter excessive risk-taking by limiting the set of activities that banks are allowed to engage in. Currently, however, bank executives have strong incentives to try to get around those regulations. In addition, Bebchuk and Spamann argue that regulators should at least monitor executive pay structures in determining whether safety and soundness risks exist. Ideally, they would link executive compensation not to the value of common shares, but to the aggregate value of common shares, preferred shares, and bonds. This would take away the incentive to take actions that have positive expected value for shareholders but negative expected value for the assets in aggregate.
The idea is that, in principle, it’s better to give executives the incentive to do the right thing than to give them the incentive to do the wrong thing and then try to hem them in with regulations.
Indeed, if pay arrangements are designed to discourage excessive risk-taking, direct regulation of activities could be less tight than it should otherwise be. Conversely, as long as banks’ executive pay arrangements are unconstrained, regulators should be more strict in their monitoring and direct regulation of banks’ activities.
Would the banks go for incentive compensation tied to the entire balance sheet rather than just common shares? It’s an interesting question. Under Bebchuk and Spamann’s proposal, they could still have enormous bonuses; from this perspective, it’s the structure that matters, not the size. But if the banks insist on tying their bonus packages to common shares, then we know that they are perfectly happy taking excessive risks. In that case, then this passage becomes particularly relevant:
In principle, well-designed incentive pay can improve the management of firms. . . . That being said, our analysis above identified major problems with the current incentive structure for bank executives. From this perspective, scaling down financial incentives may be a good thing. No financial incentives may be better than bad ones. Thus, if incentive compensation remains structured in ways that provide perverse incentives, limits on incentive pay can actually improve matters.
(On a related note, Marketplace and ProPublica ran a story yesterday on TARP recipients that figured out ways to give their executives golden parachutes.)