The optimal level of risk is not zero

Geoffrey Manne —  19 August 2009

I have said it before and I’ll say it again: All of this hand wringing over executive compensation seems to exist in a parallel world where corporate executives have no risk aversion, where there is no real competition for managerial talent, and where firms can only take on too much–never too little–risk.  And this in a day and age (the age of never-ending financial reform regulation, Lehman/Bear, enormous public scrutiny of financial and banking industries, etc.) when the downside from excessive risk-taking is now either a) extremely large or b) non-existent (but only because of guaranteed government bail-outs).  In either case, fiddling with compensation schemes will not help matters.

And yet, as Marc Hodak reports, the German banking regulator is adopting strict compensation controls–including clawbackswith no actual evidence that compensation played a role in the crisis nor that controlling it will improve matters. And those clawbacks? For deals that “go sour.”  That’s right:  Ex post punishment for any downside, no matter the ex ante expected value of the bet.  Limited upside and negative downside.  The perfect recipe for optimal corporate governance.

Here’s Marc:

There is no distinction between whether the bets that led to those losses were good ones or bad ones at the time they were made, only whether or not they turned out bad.  Consider the following scenario:  A banker sees an opportunity to bet $100 on a project that has even odds of either doubling his money or losing half of it.  He would be a moron banker to pass up this bet.  The bank wants to encourage him to find these bets and make them.  They have two choices on how to reward him.  They can either reward him based on the expected value of the bets, i.e., $25 in this case, or they can reward him based on whether the bet actually succeeds of fails, i.e., plus $100 or negative $50.  A reward based on the latter has a much higher cost to the bank since it must compensate the banker for the added uncertainty.

According to the new rules, the bank must adopt the latter, costlier scheme.  They will have no ability to pay people bonuses for their expected value contributions if they must claw them back if good bets sour, as they often do in the business world.  And that latter scheme has additional problems in the real world besides cost.  In some cases it may be easier to estimate the quality of a particular bet than to know its actual result if the results of that bet get tied up into the results of other bets from the same book.  In some cases, the results of particular bets, even if they can be tracked, may not be known for several years, possibly after the banker has moved onto another position.  Delaying bonuses also significantly increases compensation costs since one must be compensated for deferring compensation.  If you don’t defer the compensation, and you have to take it back later, then you have the logistical issue of recouping compensation already paid–in essence  reaching into someone’s personal savings to get back the cash.

What did the regulator say to all these problems?

For the first time, Bafin has established provisions for clawing back money from individual employees if the deals they do turn sour.  In so doing, Lautenschlaeger acknowledged that she had overridden concerns from the banks that such provisions are unworkable.

We have a term for that over here.  It begins with an “f” and end[s] in “you.”

Ironically, the banks’ reactions to these provision are almost certain to both increase the costs to the banks, and also reduce the alignment of their bankers.  That’s what happens when you base prescriptions on the wrong diagnosis.

In the US we go even further. We have a “pay czar” (gag!) who claims unfettered power, including the power to clawback compensation for, well, any reason he feels like.  His exact words:  “Anything is possible under the law.”  At least his jurisdiction is (for now?) limited to firms that received TARP money.  I wonder if he’s ever heard of agency theory or thinks that compensation performs any function other than unduly lining greedy pockets.

Meanwhile, every week brings an new op-ed from Lucian Bebchuk or a shrill commentary from Simon Johnson and/or James Kwak pinning the responsibility for the crisis on excessive pay, with seemingly no regard for the risk (heh) of excessive risk aversion and the natural risk aversion of managers relative to shareholders.  It may be that things have gotten out of whack in some firms (although Falenbrach and Stultz in the article linked above find no evidence of this for banks), but the solution is not to regulate performance-based compensation schemes out of existence.  (Nor will nominally independent boards help any (see here, for example)).

It would be nice if the “solutions” to our financial market woes bore more relationship to the problems.

Geoffrey Manne