A recent op-ed on this blog by Paul Kupiec misstates the Clearing House’s criticism of the supplementary leverage ratio. Kupiec’s article indicates that the Clearing House’s position is contained in a recent article by Darrell Duffie, a professor at Stanford University. Although Professor Duffie’s article appeared in the Clearing House’s quarterly journal Banking Perspectives, his views were his own, and he received no compensation from the Clearing House.
We will let Mr. Kupiec debate Dr. Duffie’s views with Dr. Duffie. As for us, the Clearing House’s arguments against the leverage ratio are explained in a Clearing House Research note, “Shortcomings of Leverage Ratio Requirements,” which has been available on our website for over a year. The leverage ratio’s core intractable problem is that it treats all bank assets as having the same level of risk, from deposits at the central bank and Treasury securities to leveraged loans, loans to small businesses and credit card loans. If the leverage ratio were simply a backstop that didn’t influence bank behavior, its poor design would be irrelevant for banks’ capital planning decisions. But as I explain below, it clearly is influencing banks’ behavior, and its flaws therefore have dangerous implications.
In particular, the leverage ratio encourages banks to shift toward riskier assets, reduces the liquidity of financial markets, makes it more costly for banks to comply with liquidity requirements, drives intermediation into the shadow banking system, and interferes with monetary policy.
In all modesty, though, it is incorrect to characterize these concerns as “The Clearing House’s argument against the leverage ratio.” It seems like this is virtually everyone’s argument against the leverage ratio these days. The same concerns were expressed by governors on the Federal Reserve Board when the SLR was adopted, by the Bank of England when it excluded central bank reserves from the SLR denominator, by multiple Fed governors (including Chair Janet Yellen) in recent weeks in their indications that adjustments to the SLR are being considered, by the Bank for International Settlements in its report on the impact of Basel III on monetary policy and its report on repo markets, and by many academic economists.
Bank capital requirements come in two basic types. Risk-based capital requirements require banks to hold more capital for riskier assets and less capital for low-risk assets. Leverage ratio requirements require banks to hold the same amount of capital for any type of asset, regardless of its risk.
Commercial banks in the United States have been required to satisfy a leverage ratio requirement since 1981, but until recently, banks in other jurisdictions have not. That changed when the international Basel III regulatory reforms included a leverage ratio requirement, the SLR, which includes both on-balance-sheet exposures as well as off-balance-sheet exposures such as undrawn amounts on credit lines and derivative transactions. The SLR requirement adopted by the Basel Committee is calibrated roughly at the same level as the existing Tier 1 leverage requirement in the U.S., which includes only on-balance-sheet exposures.
The U.S. leverage ratio requirement had always been calibrated as a backstop to risk-based requirements, and the SLR as proposed under Basel III would likely have behaved similarly. However, the U.S. banking agencies implemented the SLR in the United States at twice the level agreed upon in Basel III. At those high levels, the leverage ratio is binding for some banks and for many others close enough to binding that the banks take the leverage ratio into account in their capital planning decisions. It is important to note that the dramatically higher SLR requirement in the United States is set by a U.S. regulation. The requirement can be lowered without a change in the Basel III agreement or a change in U.S. law.
The Clearing House research note cited above illustrates the flaws of the leverage ratio as a measure of bank risk by testing the ratio’s ability to predict bank failures during the crisis. Specifically, we conduct a horse race between the Tier 1 risk-based capital ratio and the Tier 1 leverage ratio of the 8,000 banks that existed at the end of 2006 to predict the 400 bank failures that occurred between 2007 and 2011. We find that banks with higher risk-based ratios are less likely to fail, but banks with higher leverage ratios are more likely to fail. This is because a bank with a high risk-based capital ratio and a low leverage ratio must have a relatively large share of low-risk assets on its balance sheet and therefore is likely to be both less risky and more liquid. In fact, roughly one-third of the banks that failed in the crisis — 125 banks — went into the crisis with leverage ratios above the 10% level often cited by proponents as the measure ensuring a bank’s safety and soundness.
But are leverage ratios influencing bank behavior? Absolutely. As discussed in the Federal Reserve Bank of New York’s annual report on open market operations, and a number of speeches and blog posts, the consequences in financial markets are dramatic. The influence is easy to see across multiple jurisdictions.
The consequences of the SLR are easiest to observe for foreign banks since their leverage ratio requirement is effectively turned on and off as a result of how they report data. When the requirement is turned off, U.S. branches of foreign banking organizations (FBOs) borrow in the federal funds market and deposit the proceeds at the Fed where they earn interest, and foreign banks also borrow from other broker-dealers at the low triparty repo rate and lend to other financial entities at the higher General Collateral Financing repo rate.
But when the requirement turns on, foreign banks take steps to reduce their balance sheets, raising their leverage ratios. Activity in the federal funds market drops as does the federal funds rate. Activity in the tri-party repo market falls off sharply. At the same time, Fed borrowing through its reverse repo facility jumps, presumably because institutions that had been lending to FBOs temporarily shift their funds.
Similar consequences logically must also be there for U.S. banks, which face a leverage ratio requirement that is twice as high as foreign banks. And in the U.S., the leverage ratio requirement is effectively turned on every day, not just in certain reporting periods. For example, as discussed in the TCH research note cited above, the respondents to the Federal Reserve’s June 2015 Senior Financial Officers survey indicated that regulatory changes had decreased the willingness of broker-dealers to act as market-makers, impairing the liquidity and functioning of Treasury markets.
Economists at the New York Fed found that the SLR was the principal cause of the dislocation that has occurred in corporate bond markets since mid-2015. And major U.S. broker-dealers have exited entire lines of business that are no longer profitable because of the SLR requirement. The responses of banks to a binding SLR have likely resulted in a reduction of credit intermediation and higher lending rates being charged to bank borrowers to account for the higher costs.
The evidence is irrefutable that the leverage ratio is a poor measure of bank risk and that it is having a material impact on bank behavior. In addition, the SLR makes it more expensive for banks to comply with liquidity requirements by requiring banks to fund their holdings of riskless and liquid assets with expensive capital. As explained in a recent TCH blog post and research notes, those added costs drive credit intermediation outside of the regulated sector and into the shadow banking system, increasing financial stability risks. As discussed above, the SLR is also leading to unnecessary volatility and unproductive balance sheet shifts in money markets, complicating the implementation of monetary policy.
The Clearing House’s “argument against the leverage ratio” is specifically about the SLR requirement as it is implemented by the U.S. banking agencies, at a level twice the international requirement. Having a leverage ratio requirement as a backstop to risk-based capital requirements is a sound precaution against banks' investing heavily in assets with incorrectly low capital requirements along with the attendant financial stability risks. But the deep imprint left by the SLR on U.S. markets illustrates that it is not calibrated as a backstop and needs to be adjusted.