Doing the Math on the Leverage Ratio

On Tuesday, I had the pleasure of testifying before the House Financial Services Committee on a wide range of bank regulatory topics, including the regulation of bank capital.  Much of the discussion focused on the use of a leverage ratio to measure a bank’s capital adequacy.  As I explained in my written testimony and opening statement, the leverage ratio is highly problematic.

First, as a general matter, leverage ratios are poor measures of a bank’s capital strength.  The supplementary leverage ratio measures the capital adequacy of a bank by dividing its capital by its total assets and off-balance sheet exposures.1   Although sometimes viewed as an alternative to risk-based capital, the leverage ratio is, in fact, also a risk-based measure of capital – albeit a very inaccurate one.  It assesses the risk of holding every asset to be exactly the same – akin to setting the same speed limit for every road in the world, whether it’s a highway or a school zone.  Although the risk-weights used in risk-based measures can sometimes be wrong about the risk of an asset, a leverage ratio is almost always wrong.

This inaccuracy is especially pronounced for banks engaged in capital markets or custody activities, or those holding large amounts of liquidity – all involve large quantities of cash, Treasuries or other low-risk assets, which a leverage ratio penalizes harshly, requiring much more capital than economics and risk would otherwise suggest.  While leverage ratios can be useful as a simple backstop to other primary measures – namely, risk-based capital ratios and robust stress-testing – their one-size-fits-all view of risk is so inconsistent with the actual economics and risks of banking that, if it is set at a level that binds, a leverage ratio would inevitably frustrate and distort the allocation of credit to the economy.

To put a finer point on just how pronounced an effect a binding supplementary leverage could have on lending and the real economy, it’s useful to consider some simple numbers.  For example, if the entire U.S. banking were to meet a 10% supplementary leverage ratio, the current capital of the U.S. banking system would support at least $4.8 trillion less in loans and other productive activities than it currently does today.2   The economic ramification of such an enormous reduction in lending capacity would be extraordinarily negative.

The supplementary leverage ratio is a poor measure of capital, both in theory and in practice.  We would be much better served by limiting and discouraging its current use, rather than expanding and encouraging it.

1A traditional leverage ratio only divides a bank’s capital by its total balance sheet assets; it does not take into account off-balance sheet exposures.  The supplementary leverage ratio, which was introduced in Basel III and currently applies only to larger U.S. banks (14, to be exact), takes into account both.

2 Because that number may be surprisingly large to some, here’s the math:  Currently, the entire U.S. banking system holds $1.7 trillion in tier 1 capital, and $ 21.2 trillion in aggregate total assets and off-balance sheet exposures.  At a 10% supplementary leverage ratio, that $ 1.7 trillion in capital could support only $16.4 trillion in assets and off-balance sheet exposures – $4.8 trillion less than today.  This actually understates the likely impact – the vast majority of U.S. banks (i.e., all but 14) are not currently required to calculate or hold leverage capital against their off-balance sheet exposures.  If they did, the reduction in lending capacity would be even higher.