Hearing: A Review of the New Basel Capital Accord
Wednesday, June 18, 2003
Thank you for inviting me to the Senate hearings on the B.I.S. recommended regulations on Capital Allocations for Bank Credit and Other Assets -- the so-called "Basel II" accord. I have followed Basel II's consultative papers since the first one was issued in June 1999. We have submitted two formal commentaries to the Basel Commission on Bank Supervision, primarily related to the first of the so-called pillars of the new recommendations -- capital adequacy based on the specific risk characteristics of bank counterparties. Our major comments were that the capital requirements related to expected and unexpected losses from corporate and other loans should be based on the actual historical experience of Loss Given Default (LGD) from the corporate bond and bank loan markets. The original 1999 suggestions bore little resemblance to actual performance and we pointed this out fairly precisely. I am pleased to note that the latest version of Basel II's capital requirements based on the riskiness of bank portfolios does a much better job of relating the requirements to default experience, although too little capital is still being required for the most risky categories.
A problem with the suggested regulations, however, is the complexity in determining capital requirements and the somewhat arbitrary choice of modifications to the standardized scale due to such items as the size of the counterparty and the existence, or not, of collateral on the loan/bond. For example, small and medium sized enterprises (SMEs) are given lower capital requirements for comparable risk levels of as much as 25-50% less capital. The argument that the correlation of default rates amongst these small counterparties is lower than for larger corporations may be valid, but I have seen little evidence that the "haircut" for these loans should be as much as 50%. In my opinion, this was a concession to those national banking systems of the world whereby SMEs are the vast majority of borrowers -- hence lower capital requirements for banks in those countries. It is also true that SMEs make up the vast majority of loan assets of the smaller banks in the U.S. and the same lower capital requirements would hold for U.S. SMEs and the banks that make these loans -- close to all but 100 of our nation's 8,000 banks. But, as I will now discuss, almost all of U.S. banks will not be required to follow the recommendations of Basel II, so the reduced capital requirements on SMEs will not be relevant and the old Basel I's 8% rate will probably still be in effect for all except the very largest U.S. banks.
U.S. vs. Rest of the World and Basel II
As I indicated above, and as you are probably all aware of, Basel II's recommendations on credit risk and operating risk dimensions of bank activity are just that -- recommendations. The Central Banks of the world, and other bank regulatory bodies, who set national bank regulatory policy, may or may not choose to conform to all or any parts of Basel II's recommendations. Indeed, it came as an enormous surprise to some observers, including this writer, that only the largest 10 U.S. banks, and perhaps the next 10-20 banks in terms of asset size, would be required (top 10) or will have the option (next 10-20) to follow the advanced Internal Rate Based (IRB) version of Basel II's accord with respect to specifying the LGD dimensions of their portfolios and hence, set capital requirements based on portfolios risk characteristics. While it is true that as much as two-thirds of all bank assets are held by the top 30 U.S. banks and more than 95% of foreign bank assets operating in the U.S. will be covered by Basel II's most sophisticated guidelines, it is likely that all the rest of our banks (almost 8,000 smaller banks) will not be asked to conform and will probably not do so for the following reasons:
(1) Basel II is too complex and costly to introduce and conform with.
(2) The U.S. banking system is presently more than adequately capitalized and the recent decade's experience of very low numbers of bank failures makes change unnecessary.
(3) The added Basel II capital required for operating risk is based on highly arbitrary and extremely difficult to measure variables.
(4) The Federal Reserve System's, and other Bank Regulatory agencies, policy of "prompt corrective action," and maximum leveraged ratios, when bank capital falls below a certain specified level has worked very well in the U.S. and is not specified as part of Basel II -- even in pillar 2's regulatory oversight.
In other words, "if it ain't broke, don't fix it!"
I believe that the choice of only the largest ten (10) commercial banks to conform to Basel II, and the IRB approaches, is unfortunate and should be reconsidered. Not withstanding the recent consolidation movement of many of our largest and most sophisticated banks, the possible exemption of #11 to #30 (including HSBC Bank USA, Citibank [West], Bank of New York, Key Bank and State Street, #11-15) and the very likely exemption of #31 to #50 up to 8,000 (including such seemingly large banks as Charter One, Am South, Union Bank of California, Mellon Bank and Northern Trust, to name just a few, seems arbitrary and belittles the possible sophistication and motivation of these banks which would be substantial institutions in most other countries of the world. For example, the 50th largest bank in the U.S. in terms of assets (Compass Bank with $24.3 billion) or in terms of deposits (Mellon Bank with $15.2 billion) would be huge institutions in most countries.
The choice of a round number, like ten, would seem to be insensitive to world opinion as well as to risk management motivation. Speaking from an economic standpoint, rather than a political one, I would prefer to see either no banks be required to conform or some exemption level whereby the costs/benefits to our banking system would be more rationally presented and defended. Certainly, a number like the top 50-100 banks would be much more in line with the number of banks conforming in other countries. This would help ensure a "level playing field" amongst banks.
Our largest banks are probably relatively happy to conform to Basel II even with its complexity and added costs to develop information and credit scoring systems to conform to the requirements of the advanced Internal Rating Based (IRB) systems mandated under Basel II. The reason is that they expect that the total capital required for credit assets will be less than what is required under the current regime (which will continue until 2007). So, we may have a new regulatory regime where everyone -- large and small banks, as well as out bank regulators - are relatively pleased with the changes recommended under Basel II.
A Related Disappointing Result of U.S. Policy
I have always felt that despite its problems with (1) complexity, (2) too low capital requirements on the more risky counterparty assets and (3) the difficulty of managing against operations risks, Basel II had one extremely important by-product -- the motivation for banks to develop or improve upon their existing credit scoring models and systems to reduce total losses from non-performing and eventually charged-off loans. These systems can be used to rate and set capital for all bank customers rather than setting a "one size fits all" (8%) requirement on them. I have observed the enormous strides achieved by banks throughout the world, including ones of all size and location, as to developing risk management systems and training of personnel to prepare for Basel II. Indeed, from what I can surmise, banks in most countries, especially in the European Union, will all adhere to Basel II's Standardized, Foundation or Advance IRB approaches. Granted that regulators in these countries will need to sanction far fewer banks than U.S. regulators would have to do if all banks are mandated to conform, it must have come as a surprise, perhaps even a resentful shock, that the vast majority of U.S. banks will not adhere to Basel II. This is especially true since the United States and its representatives to the B.I.S. were early champions of the need to change the way banks allocate capital for credit risk of their clients.
What is disappointing to me is that the FED's decision to exempt all but the largest banks from building and implementing an IRB approach of some level of increased sophistication will reduce the motivation for most banks to move to a more risk sensitive lending policy. I recommend that our Federal regulators require some level of added due diligence on the part of banks with respect to economic capital decision making and the use of credit scoring or rating systems even if they are not absolutely required under the old (and continuing) Basel Accords. One way to accomplish this on a cost effective basis is for smaller banks to combine resources (data and money) to accomplish these goals. Our decision to exempt smaller banks from Basel II may backfire if many of the world's smaller, or even larger, banks from developing and developed countries, also opt out of the process, leading to greater instability in these banking systems and perhaps to ours through contagion.
A Note on Procyclicality
One of the likely by-products of Basel II and its reliance on systems that require a careful assessment of credit ratings and loss given default reserves is the possible procyclical impact. That is, any system which requires more capital when defaults increase and banks' portfolios become more risky, such as what is likely to occur during periods of economic stress, will motivate banks to provide even less credit, i.e., ration credit or a credit-crunch, thereby exacerbating economic downturns. The opposite will likely occur during periods of above average economic growth, thereby causing too much easy credit and subsequent higher levels of defaults and charge offs than would have been the case under the old system. Now I am aware that this problem called procyclicality already exists due to banks and other lending and capital providers having "short memories" of the last period of economic stress. Indeed, the procyclical problem resulting from the benign credit cycle of the mid-1990's (1993-1998) helped to cause the enormous level of defaults in 2000-2002. And, our research shows that when bond and loan defaults increase, we can expect a coincident reduction in recovery rates. Hence, the LGD result will be even greater due to the negative correlation between probabilities of default and recoveries given default. I would be surprised if bank regulators, and the bank themselves, have considered this double negative effect in times of economic stress. Fortunately, our banking system was very well capitalized prior to these problems and seems to have weathered the avalanche of large firm bankruptcies (77 in 2001/2002 with liabilities greater than $1 billion) without too much stress.
Despite out seeming capital adequate condition and the fact that a great deal of procyclical behavior (e.g., herding, over compensation for short-term loan losses) can be expected from current bank regulatory guidelines, I suggest that the FED consider a more smoothed capital allocation system to even out the normal fluctuations in bank reserves, capital allocations and lending behavior. This would require more capital set aside in good times and less during periods of stress.
Thank you for inviting me to attend today's hearings and express my views. On balance, Basel II has many positive recommendations but still may prove to be inadequate to overcome strong systemic problems that normally could be mitigated by a well-capitalized and prudent regulatory oversight policy. I look forward to observing the results of Basel II on a worldwide level as well as special concern for the U.S. banking system. For your information, I have provided my bio-sketch as an attachment to this document.