Friday, 22 January 2010

The Regulator's Dilemma

If there’s one idea that has achieved consensus over the past few months, it is that regulators were asleep at the switch as the credit bubble inflated. A better regulatory system would have preempted the bubble, precluded the need for bank bailouts, and saved the world much misery.

If only it were that easy!

Imagine that you are the regulator in charge of the banking industry. What are your aims?

Well, on the one hand you want to prevent ‘unwarranted’ bank runs. An unwarranted bank run is one in which the bank did nothing wrong, and is actually well-capitalized, but due to ‘irrational’ investor panic faces a potentially life-threatening short-term funding gap. Preventing unwarranted bank runs is what lies behind well-known CB catchphrases such as “contagion”, “systemic risks”, “lender of last resort”, and “too big to fail”.

On the other hand, you as the regulator are (moderately) in favor of ‘warranted’ bank runs. If a bank does something stupid, it should pay. Depositors should withdraw their money from badly-run banks, and you don't want to stand in their way. You don't want to bail out the incompetent; that’s deeply unfair to the competent, and it messes up incentives all through the system. (To quote one famous investor, “Bailouts are bad morality as well as bad economics”).

Unfortunately, these two aims are fundamentally incompatible. Because smart bankers will simply pile into precisely those trades which pose systemic risks!

Why should a banker take the trouble to build a unique portfolio, thus exposing himself to all sorts of idiosyncratic risk factors? If these idiosyncratic factors go against him, he will appear (uniquely) stupid, and will probably not be bailed out. It’s much better for him to pile into the same trade as everyone else1. Then if things go sour, it will be a systemic crisis and so everyone will be bailed out, including the banker in question.

(This insight is nothing new; it is merely the compensation dynamic for 1 trader on a desk of 10 traders, applied to 1 bank in an economy of 10 banks, with bailouts substituting for bonuses.)

In fact the situation is even more perverse than it appears. A standard measure of trade quality is the risk-reward ratio: the lower this ratio, the better the trade. But if systemic crises and consequently bailouts are in play, then the reasoning becomes inverted. Losses from low-risk trades are, by definition, small; hence low-risk traders are unlikely to be bailed out. Conversely, losses from high-risk trades are, by definition, large and potentially life-threatening; hence high-risk traders will often be backstopped by the government. This is moral hazard at its most pernicious.

It gets worse. The more enthusiastically people herd into one (systemically risky) trade, the higher the odds of a bailout; the higher the odds of a bailout for a particular trade, the more people will want to enter that trade. Yes, it’s our old friend, positive feedback!

So what’s a well-meaning regulator to do? There are only two coherent choices, really: put an end to bailouts, or put an end to bank proprietary trading.

Sadly, I don’t see either of these happening.


# 1 Throughout this post I use ‘trades’ as a convenient short-hand for ‘institutional strategic decisions’.

No comments:

Post a Comment