To the extent that microeconomic proposals have been put forward, many could make the current problems worse, not better. These proposals come in three flavors. The first is limiting the size of bonuses (as Germany’s financial stabilization fund, SoFFin, has done for the managements of banks that took state aid), or taxing them when they are above an established threshold (as in the U.K. chancellor’s and the IMF’s plans). This raises the problem of knowing at what size bonuses become immorally large, requiring regulators and legislators who have judgment far superior to that of ordinary mortals. Furthermore, those who promote the taxation of bonuses seem to base their claims to legitimacy on intellectually questionable premises: Because these gains are “ill gotten,” taxing them will not distort an economically useful activity.
We know that the second proposal, deferring bonuses to discourage short-term speculative risk taking, won’t work because it did not work before. In 2008, almost all the big banks’ management teams had their compensation tied to long-term stock performance. Moreover, systemic crises appear every eight to 10 years, and no deferral proposal extends to that horizon. So as long as a trader is lucky with timing, he or she will be long gone before the full systemic effects of the relevant trades come to fruition. Finally, we are already seeing the deferrals being factored in to signing bonuses for star traders — the deferral migrates, but the risk stays with the original institution.
The third proposal, clawback provisions, seems attractive, especially to the punitive-minded. Many banks have already adopted variants of these, but enforcement is likely to be fraught with legal and ethical issues. Skeptics suspect that this is precisely why they have been adopted so quickly.
A better approach would be for banks and regulators to link compensation, risk, and capital at both the institutional and micro levels. The place to start is at the source: the trading desk. Most of the enormous bonuses have gone to a small subset of bankers — traders, whose activities were the most significant driver of the financial crisis. Although much public anger is directed at banks’ executives, they are not generally the recipients of the largest packages. In simplified terms, traders are paid as much as 50 percent of the net present value of their position each year, even though the results are played out over time. When they win big, both they and the bank prosper. But when traders lose, they still get the reward, and the bank — and in some cases, the taxpayer — takes the punishment, particularly when the bank has reserved insufficient capital to protect against this eventuality. There is thus a mismatch between the traders’ interests and those of the bank’s shareholders and the taxpayers who are the underwriters of the state’s implicit guarantee of these institutions.
The solution lies not in aggregate, rules-based regulations but in a reassessment, within each bank, of how the “triangle” principle should be applied; that is, how to interweave the ways risk is taken, people are paid, and capital is allocated, and hence the share of profits that goes to insurance, to compensation, and to shareholders’ accounts. Instead of shifting the burden of judgment to Solomonic regulators, this approach would better harmonize individual and institutional incentives. When bankers have reason to pay attention to the true economics of their trades, they will make better trades. By aligning incentives for traders with the long-term stability of the institution, the interests of long-term investors and the system at large are also likely to be better looked after.
The first set of issues a microeconomic approach must address is how to structure traders’ compensation in a way that better ensures institutional, and therefore systemic, interests. The most direct path would be to have the traders’ interests mimic those of the institution. In the current compensation model (which remains largely untouched), traders are paid a share of their net asset position at some interval (usually annually) based on the mark-to-market valuation of that position in the context of a value-at-risk model. Two aspects of this model are striking. First, traders are paid on accounting profits, particularly troublesome in the light of long-tail contracts. Second, the assets are treated the same regardless of their underlying riskiness.