Comment on Proposed BCBS Capital Standards

Editor's note: Douglas Elliott sent the letter below to the Basel Committee on Banking Supervision — a supranational organization headquartered in Europe that coordinates international financial cooperation.

I am writing in broad support of your proposed revisions to international capital standards, although I do want to emphasize the great importance of calibrating the new levels so that they achieve the necessary degree of stability without excessive damage to lending and other financial activity. While my research has indicated there is considerable room for higher capital requirements, there will be some level of adverse effects on credit availability and pricing. It is critical that the total set of regulatory changes, including all the various revisions proposed by the Basel Committee, avoid creating a cumulative effect that is excessive. This is the reason that I underline the importance of a thorough and comprehensive quantitative analysis. In addition, I urge the committee to make clear that financial institutions will have to respond to the full range of policy actions, including higher taxes and any limitations on business operations, and that the cumulative impact could be substantially harsher than the effects of implementing the committee’s recommendations on a stand-alone basis.

By way of background, I am a Fellow at The Brookings Institution, a public policy institute based in Washington . I came to this position after two decades as an investment banker whose clients were financial institutions. The bulk of my career was with J.P. Morgan, where I aided financial institutions with strategic advice and help in raising capital. I also worked as an equities analyst and a credit analyst focused on financial firms. Given this background, I have spent the great bulk of my time at Brookings analyzing financial institutions and markets and their regulation. Approximately 30 papers of mine on these topics are available here.

I agree strongly with the Basel Committee that financial institutions must be subject to higher minimum capital requirements. Capital is the basic protection for financial institutions against the effects of mistakes and bad luck; the recent crisis has demonstrated once again how harmful those effects can be and therefore just how important capital is. Similarly, liquidity requirements must be raised significantly, since a liquidity crisis can turn into a solvency crisis, even with “lender of last resort” facilities available. I further agree that the systemically important institutions, which represent the greatest risk for our societies and taxpayers, should face even higher capital and liquidity requirements than other financial institutions.

However, the greater safety will come at a cost. All else equal, interest rates on loans will go up, availability of loans will decline, lending terms will tighten, interest rates on deposits will drop, and compensation levels in the industry will fall. (My comment letter focuses on lending activity, but there will be analogous issues for trading and other non-lending activities.) I commend the committee’s decision to do a quantitative analysis of the effects on credit and other activity of the higher capital and liquidity requirements. I urge that it be thorough and transparent and that all parties have the necessary time to absorb and comment on the analyses.

In two of my own papers, I have explored at significant length the likely quantitative effects of higher capital requirements on U.S. lending.

In the two papers, I use a simplified loan pricing model to demonstrate that significantly higher capital requirements would indeed have an adverse effect on loan pricing and availability, but that the likely effects are within a range that should be quite manageable, if implemented sensibly and with a sufficiently long transition period. I examined an increase in tangible equity ratios of two and four percentage points, but will only describe the results here for the larger increase, in the interests of simplicity.

My model showed that a four percentage point increase in equity would require a 74 basis point (bp) increase in loan pricing, if that variable bore the entire burden of adjustment, with no change in anything else. However, loan pricing would only need to rise by 20 bps on average, if all of the other variables were to adjust in reasonable ways. Importantly, the required return on investment for equity and debtholders should fall, to reflect the greater safety. (If the assumptions of Modigliani and Miller’s classic cost of capital analyses held here, there would be a complete offsetting adjustment in those required returns, eliminating the need for higher loan prices. However, as my paper discusses, several key assumptions do not hold in the real world of U.S. banking, considerably reducing the benefit of this adjustment mechanism.) In addition, banks can reduce compensation levels and other administrative costs, lower deposit rates, tighten loan covenants, and increase the ancillary business that they require from loan customers. In theory, loan availability could also be reduced, but my belief is that a 20 bp adjustment in loan pricing is too low to spur much of a change in lending standards.

The pricing model was used to estimate changes to the banking system as a whole, based on FDIC data and certain estimates of my own. Fortunately, sensitivity analysis showed that the conclusions are quite robust; the required changes in loan pricing are not affected significantly by the choice of initial assumptions for the key variables. The second paper went on to apply the loan pricing model to segments of the banking industry and of the types of customers and loans. I found that the overall conclusions remained the same in almost all cases.

I supplemented the loan pricing model analysis by looking in depth at the potential sources of competition for credit and deposit business in the U.S. The main conclusion is that banks are currently in a position where they have significant room to respond as necessary to higher capital requirements without being hemmed in by strong competition from other credit providers. The business that banks could easily lose to capital markets and other competing financial institutions has largely already been lost. The remaining types of business, such as contingent credit commitments for large corporations, are relatively insulated.

My analysis looked at these changes on a stand-alone basis. In reality, banks will be responding to these changes and many others. The cumulative effect could be large enough to cause credit pricing to rise excessively or availability to be reduced to a harmful extent. Coordination of regulatory changes will be critical to achieve the needed increase in stability without needlessly harming the real economy.

Finally, I would also like to indicate my support for the serious consideration of contingent capital, which could be a relatively inexpensive way of providing an extra measure of protection for the system. There are major theoretical advantages to contingent capital, which have been discussed extensively, but it will take significantly more time to work out how to structure its use most effectively.  There is even a chance that implementation difficulties would turn out to make it too expensive, although I do not think this will be the case. Of course, even without implementation issues, it would add costs to the financial system and should only be included to the extent that we would genuinely need that measure of protection. That is, we should not determine an ultra-safe level of capital requirements and then throw contingent capital on top for good measure. Rather, it should serve as the final layer of protection within a set of layers that would be necessary in any case.