OPINION

SEPTEMBER 30, 2010.

Two Cheers for the New Bank Capital Standards

Why do we still rely on the rating agencies, and why are we still allowing Lehman Brothers levels of leverage?.

By ALAN S. BLINDER

On Sept. 12 the heads of the world's major central banks and bank-supervisory agencies met to bless what is called "Basel III," the latest international agreement on bank capital requirements. Should we be applauding or frowning upon this agreement? A little of each.


The first big achievement, and it is a big achievement, is that 27 countries, each with its own disparate views and parochial interests, were able to agree at all—just 18 months after many of them were still fighting the last acute phase of the financial crisis.


But what about the substance of the agreement? What was it supposed to fix, and did it?


Remember, the essence of the Basel accords is establishing a minimum ratio—of capital to risk-weighted assets—and ratios have both numerators and denominators. It turns out that defining the numerator, a bank's capital, is fraught with difficulties: What counts and what doesn't? Most of the changes from Basel II to Basel III are about the numerator: raising the amount of capital required and stiffening the definition of what counts. Measuring assets is more straightforward, but risk-weighting them is not, which is the essence of the denominator problem.


Before the crisis, at least three major shortcomings of Basel II were apparent:


• Once you cut through the complexities, Basel II actually reduced capital requirements relative to Basel I. Even before the financial wreckage of 2007-2009, that looked like a mistake. After the crisis, it looked absurd.


• In determining risk weights for the denominator, Basel II assigned a major role to risk assessments by credit rating agencies like Moody's and Standard & Poor's. Once again, that looked dubious before the crisis and ludicrous thereafter.


• Basel II allowed the largest—did someone say, the "most sophisticated"?—banks to use their own internal models to measure risk. Let me repeat that: The biggest foxes were allowed to assess the safety of the chicken coops—another serious risk-weighting (denominator) problem.


Then along came the crisis, revealing two more glaring weaknesses:


• One was the startling extent to which some banks had used structured investment vehicles (SIVs) and similar arrangements to avoid capital requirements by shifting assets off balance sheet. This loophole cried out for plugging.


• The Basel Accords have always focused on minimum capital requirements. But the crisis demonstrated that, in a crunch, shortages of liquidity can be just as hazardous as shortages of capital. Indeed, it was often hard to tell one from the other. That made the need for minimum liquidity requirements apparent.


Those five issues should have formed the core of the Basel III agenda. What was actually accomplished? Let's go down the list.


First the good news: Capital requirements will be raised substantially. Right now so-called Tier 1 capital must be at least 4% of risk-weighted assets and Tier 2 capital must be at least 8%. The Basel II definition of Tier 1 capital includes some things that are not common equity, such as some types of preferred stock; and Tier 2 includes many more things, such as certain types of reserves and subordinated debt. Basel III places the focus squarely where it belongs: on common equity, which is undoubtedly real capital. And, after a long phase-in period, it will raise the minimum common-equity requirement to 7%. Hooray for both. But, folks, couldn't we have asked the world's bankers to comply with the higher standard before 2019? Maybe if we said, "pretty please"?


Because of demonstrable problems in assigning appropriate risk weights, Basel III also resurrects, as a kind of backstop, the old-fashioned leverage ratio: Tier 1 capital divided by total assets, with no risk weighting. Good idea. But, once again, why must we wait until 2018 for full implementation? Furthermore, the chosen capital requirement is only 3%—which you may know by its other name: 33-to-1 leverage. Isn't that about what Lehman Brothers had?


Second, while the Dodd-Frank Act wisely removed most provisions in U.S. law that gave the rating agencies special exalted status, Basel III did not. So the agencies that did so poorly in rating mortgage-backed securities and collateralized debt obligations will continue to play major roles in the risk-weighting process.


It gets worse. Didn't the Basel Committee notice that the internal risk models of most of the world's leading financial institutions led to disaster? Whether it was gross-but-honest errors in assessing risk or self-serving behavior is an important moral question, though not an important operational one. Either way, letting banks grade themselves worked out about as well as letting students grade themselves. Yet this grotesque shortcoming of Basel II remains in place.


Fourth on the list is the off-balance-sheet entities that caused the world so much grief. Here, some technical improvements were made, thank goodness. For example, SIVs and the like will be put back on banks' balance sheets for purposes of computing the leverage ratio. But unfortunately not for the main risk-weighted capital requirements.


Last, but not least, genuine progress was made toward new minimum liquidity requirements. The technical problems and novelty in defining liquidity proved to be formidable, as did the opposition from the banking industry. So this job is not finished. But the Basel Committee did at least institute a new liquidity requirement that will become effective in 2015.


Beyond that, the committee kicked most of the novel ideas down the road. For example, imposing higher capital requirements on systemically-important institutions is left for the future.


So let's applaud Basel III, though one-handedly. More capital, better capital, a leverage ratio, and a liquidity requirement are all important steps forward. But the unwarranted reliance on rating agencies, the disgraceful internal risk models of banks, and the disastrous SIVs should have been easy marks for reformers.


Should the U.S. adopt the Basel III changes? Absolutely, with no hesitation. But work on Basel IV should begin immediately.


Mr. Blinder, a professor of economics and public affairs at Princeton University and vice chairman of the Promontory Interfinancial Network, is a former vice chairman of the Federal Reserve Board.


Copyright 2009 Dow Jones & Company, Inc. All Rights Reserved