Wednesday, June 15, 2011

Slippery bankers slither through the constraints...

A couple months ago I wrote this short article in New Scientist magazine (it seems freely available without subscription). It pointed out a troubling fact about recent measures proposed to stop bankers (and CEOs and hedge fund managers, etc.) from taking excessive risks which bring them big personal profits in the short run while saddling their firms (and taxpayers) with losses in the long run. The problem, as economists Peyton Young and Dean Foster have pointed out, is that none of these schemes will actually work. If executives keep tight control on details about how they are running their firms and how they are investing, they can always game the system by making it look outwardly to others that they're not taking excessive risks. There's no way to control it without much greater transparency so shareholders or investors can see clearly what strategies executives are using.

Here's the gist of the article:
You might think that smarter rules could get around the problem. Delay the bonuses for five years, perhaps, or put in clauses allowing the investor to claw back pay if it proves undeserved. But it won't work. Economists Dean Foster of the University of Pennsylvania in Philadelphia and Peyton Young of the University of Oxford have now extended Lo's argument, showing that the problem is simply unsolvable as long as investors cannot see the strategies used by managers (Quarterly Journal of Economics, vol 125, p 1435).

As they point out, delaying bonuses for a few years may deter some people as the risky strategies could explode before the period is up. But a manager willing to live with the uncertainty can still profit enormously if they don't. On average, managers playing the game still profit from excess risk while sticking the firm with the losses.

So-called claw back provisions, in the news as politicians ponder how to fix the system, will also fail. While sufficiently strong provisions would make it unprofitable for managers to game the system, Young and Foster show that such provisions would also deter honest managers who actually do possess superior management skills. Claw back provisions simply end up driving out the better managers that incentives were meant to attract in the first place.
This work by Young and Foster is among the more profound things I've read on the matter of pay for performance and the problems it has created. You would think it would be front and center in the political process of coming up with new rules, but in fact it seems to get almost no attention whatsoever. From this article in today's Financial Times, it seems that bankers have quite predictably made some quick adjustments in the way they get paid, conforming to the letter of new rules, without actually changing anything:
Bank chiefs' average pay in the US and Europe leapt 36 per cent last year to $9.7m, according to data compiled for the Financial Times, despite variable performance across the sector....

Regulators have declined to impose caps on bank pay, instead introducing changes they believe will limit incentives to take excessive risks. That has led many banks to increase fixed salaries, reduce employees’ reliance on annual bonuses and defer cash and stock awards over several years.

“The real story around pay is the progress on ensuring bonuses are deferred, paid in shares and subject to clawback and performance targets, rather than the headline figure,” said Angela Knight, British Bankers’ Association chief executive.
 It's just too bad that we already know it won't work.