How Golf and Game Theory Apply to Capital Rules

In our spare time my wife and I enjoy playing golf together. But it can be a frustrating sport. Recently, in exasperation, she announced a new approach to putting: aiming directly at the hole without any attempt to anticipate the break of the putt. I thought to myself that this was a defeatist and misguided strategy. I also thought of the leverage ratio.

In terms of measuring capital adequacy, the trend for decades has been to use measures that assign a risk weight to each type of asset, derived from models that predict the probability of loss on that asset. This process is not only the foundation of current Basel Committee capital standards but also how banks manage risk and assess capital adequacy for their own purposes.

It is beyond dispute that the financial crisis revealed weaknesses in this process. The most obvious example was the significant underestimation of the tail risks of mortgages and mortgage-backed securities. However, the Federal Reserve's post-crisis stress testing regime significantly reduces this problem by testing how banks withstand stresses not based on recent performance; so too does the Fed's rigorous review and approval of each bank's internal model that is used for stress testing.

Nonetheless, critics continue to push enhanced reliance on an alternative that would not rely at all on risk weights and instead abdicate all responsibility for evaluating relative risk: the leverage ratio (a simple ratio of capital to total assets).

Although the leverage ratio is seen as an alternative to risk-based measures of capital, the leverage ratio is in fact a risk-based measure of capital, albeit a very bad one. It assesses the risk of each asset to be exactly the same. The risk of a Treasury security is assessed as the same as the risk of a loan to a startup with uncertain cash flows. The risk of holding a market-making portfolio of liquid, highly rated bonds is assessed as equal to holding a portfolio of illiquid loans to untested companies.

In other words, the leverage ratio simply aims at the hole and ignores the shape, speed and cut of the green, and the weather conditions. It says, "I misread my putts last weekend, so I will never read a putt again."

Of course, there are a few times when aiming directly at the hole is not such a bad strategy: flat putts. And there are assets for which imposing the current U.S. leverage ratio of 5% for holding companies, and 6% at the bank level, may approximate the capital suggested by a reasoned assessment of probability of default and loss given default. An example would be relatively illiquid loans. But the strategy breaks down when one considers safe, high-quality assets where the probability of default and loss given default are dramatically lower. There, lining up the putt the same way ensures a miss.

Indeed, the inaccuracy of the leverage ratio – and the resulting misallocation of capital – has worsened dramatically in recent years as a result of other regulatory mandates. Liquidity rules now require large banks to hold over 20% of their balance sheets in high-quality liquid assets – predominantly cash, Treasury bonds and other government securities. Those rightly receive a zero risk weight in risk-based capital measures. Large banks now hold approximately four times as much of these assets as they did pre-crisis, but the leverage ratio completely ignores this dramatic reduction in risk. Again, the slope of the green has increased markedly, but the aim remains the same.

It is difficult to understand how such a system could arise, unless we switch to soccer, and goalkeeping.

Our youngest son is a goalkeeper, so we think more than most about the game theory of penalty kicks. In Think Like a Freak, Steven Levitt and Stephen Dubner – the authors of Freakonomics – analyze trends and tactics from the perspective of a player shooting on goal.

A goalkeeper does one of three things as the kicker strikes the ball: dive to the strong side (to the left for a right-footed kicker); dive to the weak side; or stay put. It turns out that goalkeepers dive to the strong side 57% of the time, to the weak side 41% of the time and stay put 2% of the time. So, the best strategy is clearly to aim straight at the goalkeeper.

But in recent World Cup competition, it is fair to say that players uniformly rejected this approach. Why? The authors concluded that the primary objective of the kicker is not actually to maximize the odds of scoring, but rather to avoid humiliation. And there is nothing more humiliating than aiming straight and watching the goalkeeper catch the ball without moving, perhaps with a smirk.

The same choice confronts bank regulators. Is the objective of capital regulation to allow banks to allocate capital more effectively, leading to sustainable economic growth? Or have the crisis and the resulting political environment made it important for regulators to avoid being accused of relying on models developed by highly unpopular banks and to avoid justifying such reliance on the fact that bank competition in risk management is actually healthy?

Recent evidence suggests the latter. Recent regulations have increased reliance not only on the leverage ratio but also on so-called "standardized" approaches to risk-weighting, where the regulators themselves design a single model for credit, market, interest rate and other risks and require every bank to use them.

And that could lead to more missed putts and happy goalkeepers in the future.

Greg Baer is the president of The Clearing House Association.