Risk Weights or Leverage Ratio? We Need Both

As a Jew I will be spending Dec. 25 at my local Chinese restaurant. This fact has made me think about, of all things, bank capital.

One can eat Chinese food elegantly with two chopsticks. With only one you are just stabbing at your food with a stick. It is the same with bank capital regulation.

The two chopsticks in this analogy, which I initially made at a recent conference, are the simple leverage ratio — core capital divided by assets — and the more complex risk-weighted capital calculations. Much of the discussion around capital requirements is which of the two measures is preferred, as well as which measure should be the standard for an adequate capital base.

But this shouldn't be an "either or" debate. The banking system ultimately needs a balanced approach to capital, which allows banks to efficiently function while also maintaining financial stability. Apply only one capital measure — it doesn't matter which one — and the capital system will eventually flop.

Advocates for either the leverage ratio or risk-weighted measures often fail to appreciate this need for balance. Hence the regulatory debate too often consists of people on both sides embracing a one-chopstick approach.

The simple leverage ratio and risk-weighted tests each have their benefits and drawbacks, their supporters and opponents, and their proper place in the regulatory system. Start with the leverage ratio. The beauty of the leverage ratio is its simplicity: It is hard to game and easy to compare across institutions. It provides clear data for markets to react to and to help regulators do their job. However, the leverage ratio also provides a perverse incentive to banks to seek risk. Because capital charges are identical regardless of the riskiness of assets, banks seeking to maximize returns on capital should load up on risky assets.

Risk-weighted capital tests are designed to avoid that problem. Tailoring capital levels to adjust for risk should make banks safer. It should also improve the distribution of capital for society by eliminating a governmental incentive for the financial sector to steer investment toward risky assets.

The problem with risk-weighting is in the name: getting the risk weights right. Humans and models are not perfect judges of risk and when they err, they can do so in a correlated and spectacular fashion. Historical experience and multiple regression analysis completely botched the risk of subprime mortgages and the corresponding securities built off of those supposedly "low-risk" assets. A secondary problem is who runs the models and sets the weights: the banks, the regulator, or both? Since banks underwrite their assets, the industry has long argued that individual institutions should be given some control or deference to set risk weights. However, letting the banks run the models has a fox-guarding-the-henhouse problem.

Given the debate, it should come as no surprise that certain regulators are predisposed to either approach. The Federal Deposit Insurance Corp. — in its role as insurer and resolver of failed institutions — has historically backed a strong leverage ratio as part of the capital regime. The Federal Reserve Board — in its role as economic modeler, monetary policy setter and quantitative regulator — has usually focused on risk-weighted tests.

In the years following the financial crisis and passage of the Dodd-Frank Act, both the leverage ratio and risk-weighted capital ratios were increased substantially. Both were also tailored to specifically increase capital on the largest banks — whether through leverage surcharges on globally significant banks or through additional conservative assumptions in big-bank stress tests through the Fed's Comprehensive Capital Analysis and Review.

Much of the focus of the current debate around capital requirements is on which measure — leverage or risk-weighted — should be the "binding constraint" for banks, that is, which standard should determine the adequate amount of capital a bank should hold. Some in the banking industry argue that increases in the simple leverage ratio have made that the binding constraint for banks. Meanwhile, the debate continues as to whether additional capital is needed or whether regulators have gone too far.

Policymakers, industry and the general public should become comfortable with a world in which the binding constraint alternates between the simple leverage ratio and the complex risk-weighted test. Each has its own advantages and disadvantages and the relative value of those tradeoffs changes throughout the business cycle. It is tempting for one to skew toward each of the two sides at various moments, as some did in favoring the Basel II advanced risk-weight approach right before the financial crisis (which in hindsight was a clear mistake.) But the reality is that if both are set properly, each will take turns as the ultimate decider of total capital.

In that world, the chopsticks work together, and you can actually eat a meal. Otherwise we will continue to face warring factions, poking each other with sticks.

Aaron Klein is a fellow at the Brookings Institution and policy director at the Center on Regulation and Markets.