This content is from: Home

Basel under threat

When regulators set out to overhaul capital requirements for the world's major banks five years ago, their objective seemed simple enough: Maintain the overall level of capital in the global banking system while aligning it more closely with the underlying risks of banks' activities. Turning that goal into regulatory language has proved devilishly difficult, though. Regulators have had to weigh the interests of universal banking giants alongside those of smaller regional and local banks. They have bucked political pressure to make concessions for the midsize companies in Germany's Mittelstand and for sophisticated players in markets for asset-backed securities and credit derivatives. And to ensure consistency of treatment, they have proposed a mass of detailed rules estimating the risk of everything from auto loans and credit card balances to sovereign credits and borrowings by special-purpose vehicles.

After years of painstaking negotiations and untold compromises, most regulators believe that they have struck the right balance. The Basel Committee, made up of banking supervisors from Europe, North America and Japan, issued a revised draft of the proposed rules in late April and aims to reach a final agreement by the end of this year -- two years behind its original target. A recent test by 365 banks in 43 countries found that average regulatory capital requirements under the proposed rules would be close to today's levels. For committee members the test provided long-awaited evidence that their ambitious attempt to rewrite the rules of global banking would succeed.

"The results are quite convincing," says Danièle Nouy, secretary general of the Basel Committee.

Nouy's optimism is far from widely shared, however. Many bankers, analysts and even some supervisors still express deep misgivings about the proposed rules, dubbed Basel II, and they are expected to lodge serious complaints before the official comment period ends later this month. John Hawke Jr., the U.S. Comptroller of the Currency, told the Senate Banking Committee last month that supervisors were far from sure whether the draft accord would work as intended and that more testing and potentially significant modifications were needed. He called the year-end timetable "daunting" and vowed to block a final agreement unless he was satisfied with the accord.

Skeptics, including Hawke, warn that the rules are too complex and could unfairly affect competition in the industry. They point out that although average capital levels were relatively unchanged in the recent test, individual banks showed huge variations, with potential capital requirements rising as much as 55 percent for some big banks and falling as much as 36 percent for others. Some critics contend that the rules, which are based on credit ratings that banks and agencies like Moody's Investors Service give to borrowers, threaten to aggravate economic recessions by worsening credit squeezes at times when ratings are falling. Some bankers believe that Basel II could even increase systemic risk.

"We will not accept Basel II in this form," declares Hugo Banziger, chief credit officer at Deutsche Bank. "It is dangerous to the banking system."

Basel II is a dramatic departure from today's crude capital rules, which classify borrowers into a handful of categories such as sovereigns and corporates but don't require banks to hold any more capital when the borrower is a single-B fallen angel than when it is a triple-A-rated corporation. The proposed new accord would base capital requirements on the credit ratings of borrowers. Under a simplified version, banks would use ratings provided by agencies like Moody's and Standard & Poor's. Two other, more advanced versions allow banks to use their own internal ratings of borrowers' creditworthiness. The simple aim is to make regulatory capital more sensitive to the actual economic risks that banks take (see box, page 29).

The debate has exposed sharp differences between the two leading U.S. banking supervisors. Roger Ferguson Jr., the Federal Reserve Board vice chairman who has become a pivotal champion of a new capital accord, is determined to reach agreement on Basel II by the end of this year and see the new rules come into force at the end of 2006. He insists that the draft accord will strengthen the banking system and encourage banks to improve their risk management capabilities.

In stark contrast, Hawke has strong reservations. Although he supports the idea of risk-based capital requirements, he believes that the proposed rules are excessively complex and detailed, making them difficult for banks to implement and supervisors to enforce. The test results lauded by Nouy as "convincing" were actually "not reliable," Hawke says, because banks lacked sufficient data and misunderstood some of the complex rules. Hawke told the Senate Banking Committee he would not sign off on Basel II until he had carefully considered banks' objections to the rules, and said regulators should take "whatever time is necessary" to get a good deal rather stick to a timetable. Hawke told Institutional Investor that he also favored another test of the proposed accord by banks, a precaution that could delay a final agreement. "We really need another QIS," or quantitative impact survey, as the test is called, he said.

Basel II also risks running into transatlantic political difficulties. Differences aside, Ferguson and Hawke stunned European bankers and regulators in February when they announced that only ten major U.S. banks would be obliged to comply with the new accord, and another ten or so were expected to do so voluntarily. They also said that the U.S. would allow banks to use only the most advanced of the three proposed variations of Basel II being developed. The news went down badly in the European Union, where governments are drawing up legislation to require all banks to comply with the accord and where many small banks are expected to use the simplest version, which bases capital requirements on ratings from external agencies.

"The first impression was, typical United States," says one European member of the Basel Committee. "It's just like Iraq. They do what they want. If they don't like an international agreement, the hell with it."

Tempers have cooled since, helped by some diplomatic reassurances from Ferguson, but regulators admit that the incident has left suspicion in its wake. Although Hawke insists that small U.S. banks will remain strongly capitalized and well supervised outside of Basel II, he acknowledges that some members of the Basel Committee fear that the U.S. action "could undermine the credibility of Basel as a worldwide standard setter." As if to prove the point, Italian Finance Minister Giulio Tremonti surprised his EU colleagues in April by suddenly demanding that Italy's small banks, which are key lenders to the country's legion of small and medium-size companies and family-owned firms, be exempted from Basel II just like small U.S. banks. To make matters worse, the political rupture between Washington and many European capitals over Iraq hasn't created the best climate to work out any differences. The risk that Basel II could be infected by transatlantic geopolitical differences is one of the biggest fears bankers share, even if there is no sign of that happening so far. "In order to resolve this, we need good will on both sides," says the Basel Committee's Nouy.

Some of the big banks that are supposed to be the prime beneficiaries of Basel II have strong doubts about it. Sir George Mathewson, chairman of the Royal Bank of Scotland Group and president of the British Bankers Association, contends that the potential risks of the new accord are so great that regulators should rethink their approach. "Basel II risks being too much, too fast. Adopting a more evolutionary approach is in all our interests," he told a recent gathering of European bankers in Brussels. Even staunch backers of the Basel process have major complaints about some of the proposed rules, and with potentially huge differences in regulatory capital requirements hinging on the final outcome, they are getting ready for some bare-knuckled lobbying. "We and others are prepared to escalate this for key issues where we are being reasonable and the committee is not," says a senior executive at a major U.S. bank.

All these tensions leave regulators in a bind. Officials acknowledge that Basel II has its flaws and express sympathy with some bank arguments. The Federal Reserve Board, for example, published a paper last month that suggested easing capital requirements for loans hedged with guarantees, insurance or derivatives like credit-default swaps. Big banks like J.P. Morgan Chase & Co. are clamoring for such a change, arguing that the risk of a double default -- that is, both the borrower and the guarantor failing -- is much lower than Basel II assumes. But for all their sympathy, regulators also are desperate not to lose the momentum for a deal. The draft accord contains so many compromises and balances so many competing interests that any attempt to make significant changes risks unraveling it like a ball of string. A concession for U.S. retail banks, for example, is liable to provoke howls of complaint from European wholesale banks. As one senior European regulator puts it, "We understand the point, but we are so far down the line and this point is so central to many parts of the accord that you would unpick the whole thing if you looked at double default in a new way."

Regulators insist that any defects in Basel II can be dealt with as the accord is implemented between now and 2006, but they face a tight timetable. While the Basel Committee studies banks' comments this summer, U.S. regulatory agencies will issue an advance notice of proposed rulemaking in July on how to implement Basel into domestic banking regulations. American banks will have until October to comment on the advance notice, and the results could affect the U.S. negotiating position at Basel. The European Union is conducting a similar process with its own capital adequacy directive. Because of all this, a Basel Committee meeting previously scheduled for November to reach a final agreement on the accord now looks like it will be postponed until December, or possibly early 2004. Most regulators are determined to avoid a lengthier delay, which could prove fatal to the process and leave banks operating under an outdated capital regime. "At some point you have to get started," the Fed's Ferguson told II. "We will never reach perfection in the sense of the ultimate development of risk management technology. We can't let our inability to reach that stop us from moving ahead from Basel I to Basel II."

The stakes could hardly be higher as Ferguson and his Basel colleagues try to keep the accord on track and convince banks, and their clients, to live with it. The final accord will affect the competitive balance between large, internationally active banks and smaller, regional players, and between U.S. banks and their counterparts in Europe and Japan. It could restrict the availability of corporate credit in an economic downturn, slow the continued growth of the burgeoning markets for credit derivatives and asset-backed securities and stimulate housing markets around the world by sharply lowering capital requirements for mortgage lenders.

"We're creating winners and losers out of the system," one senior Basel staffer puts it.

The big winners are likely to be retail-oriented banks, which will be able to take advantage of the proposed rules' sharply lower risk weightings for mortgages and small-business loans. HBOS, the U.K.'s largest mortgage lender, could see its minimum capital requirement decline by more than 25 percent because of its large mortgage and small-business loan portfolios, estimates William de Winton, a banking analyst at Morgan Stanley in London.

Banks with big corporate, sovereign and interbank exposures are likely to face higher capital requirements or enjoy smaller savings than pure retail plays, de Winton contends. And big capital markets players will probably face higher capital charges for securitization activities and be harder hit by operational risk charges, which are largely based on revenues, he adds. Deutsche Bank, for example, could see its capital requirement rise by more than 20 percent, de Winton believes.

Basel II also could spur a fresh round of consolidation, because the accord's most advanced approach, which only large banks are expected to use, offers much greater capital savings on retail and small-business lending than does the standardized version of the accord, which many small banks will employ. The accord "is overwhelmingly to the benefit of the larger banks," de Winton says.

The potential for big banks to free up excess capital by buying a smaller bank might foster mergers "in some places," acknowledges Sir Howard Davies, chairman of the U.K.'s Financial Services Authority. But he insists that economics, not regulation, drives most strategic decisions at banks, a view shared by the Fed's Ferguson. Davies notes that most U.K. banks have concentrated heavily on retail rather than wholesale lending over the past decade because the business offers higher margins and more-predictable earnings, not because of capital requirements. "I'm really not expecting a major behavioral shift in this market as a result of Basel II," he told II.

The original Basel accord had a profound effect on the banking and financial services industry. Adopted in 1988 in response to worries about the level of capital in a number of industrialized countries, Basel I focused the attention of investors and bank executives on the desirability of a strong capital position. The accord set a minimum capital requirement of 8 percent of risk-weighted assets, including a 4 percent minimum for equity-type capital known as tier 1. But many banks ran into difficulties when their ratios came close to the regulatory minimums after the Latin American debt crisis in the late 1980s and the early '90s recession, and they resolved to build a healthy capital buffer in excess of the Basel minimum. Most major banks today, particularly those active in the swaps, derivatives and securitization markets, maintain a capital ratio in excess of 10 percent and a tier-1 ratio of more than 6 percent.

The simplicity of the current accord has proved its undoing. Basel I sorts credit risks into a handful of crude buckets, such as corporates and sovereigns, that bear little relation to the underlying risk. Banks today have to set aside the same amount of capital -- 8 percent of the loan -- whether they are lending to General Electric Co., which is rated AAA by S&P, or Xerox Corp., which has a sub-investment-grade rating of BB­. Loans to Turkey require no capital because the country is a member of the Organization for Economic Cooperation and Development and gets a zero risk weighting, while loans to Singapore, a superior credit in bankers' eyes, require capital of 2 cents on the dollar. The gap between regulatory capital -- what Basel requires -- and economic capital -- what banks believe they need based on their internal credit models -- has grown so wide as to make Basel I almost meaningless. Lenders have exploited the accord's weaknesses with a string of innovations like asset-backed securities, which enable banks to reduce their capital requirement by off-loading assets onto the capital markets. In practice, many banks tend to securitize better credits and keep lower-quality borrowers on their books, something that Basel I's crude formula failed to reflect.

Basel II aims to close those gaps by basing capital on the credit ratings of borrowers. The proposed accord better captures more off-balance-sheet activity, such as securitizations, that wasn't fully covered by the original Basel agreement. And it will introduce a controversial new capital charge for operational risk, such as the risk of loss because of a rogue trader or legal liability. Many banks have opposed the introduction of such charges as an ill-defined concept, and money management specialists like Boston-based State Street Corp. contend that the charge will put them at a disadvantage compared with nonbank fund managers, which aren't covered by Basel. But regulators are determined to stick with the charge, and most big banks today accept it as a fait accompli.

The whole package represents a huge shift in capital treatment that promises to differentiate between banks to an unprecedented degree. The recent quantitative impact survey of 365 banks conducted by the Basel Committee showed the potential for big variations in capital. Using the advanced internal-ratings-based method, capital requirements came out 2 percent lower than under current rules for big banks in the Group of Ten, which includes major European countries, Canada, Japan and the U.S. The capital charge for credit risks was 13 percent lower, but that was offset by an operational risk charge of 11 percent. The average masked big differences among individual banks, which showed capital requirements rising by as much as 46 percent or falling by as much as 36 percent. For smaller banks using the foundation IRB approach, which tend to be more retail-oriented, the reduction in capital was a more dramatic 19 percent, with the extremes ranging from a reduction of 58 percent to an increase of 41 percent.

Big banks in the U.S. and the European Union that used the advanced IRB approach showed an average drop in capital requirements of 6 percent in the test, with charges for credit risk falling 17 percent and charges for operational risk averaging 11 percent. For U.S. banks capital charges fell by 26 percent for corporate loans, 33 percent for small-business loans and 53 percent for mortgage lending. Those declines were partly offset by increases of 12 percent for sovereign lending, 16 percent for revolving credit, including credit cards, and a whopping 232 percent for equity holdings.

Senior Bundesbank officials say Germany's big banks showed a similarly wide variation of results among individual institutions and on average had a modest increase in capital requirements of about 5 percent, apparently reflecting the deterioration in creditworthiness of the country's major corporations. Far from being cause for concern, the variations in capital requirements show that Basel II does indeed reflect credit risk accurately, asserts Edgar Meister, a member of the executive board of the Bundesbank and chairman of the Banking Supervision Committee of the European System of Central Banks. "That was proof for us that the system works," he said in an interview with II. "We are happy with the results."

THE TEST PROVES THAT BASEL II WILL WORK as intended, regulators say. Capital will vary according to the underlying risk at each bank, while the overall capital in the system should remain little changed. What's more, the results showed a modest capital savings on average for banks that used advanced IRB, giving them an incentive to make the big investments needed to use that formula. "It will reflect state-of-the-art risk management technology," Ferguson says. Big banks "will see that it is in their interest to adopt it. The market will expect them to adopt it."

Adoption won't be cheap or easy, though. Banks wanting to use either of the internal-ratings-based formulas need to collect data going back many years on the performance of virtually every credit on their books to calculate and justify the default probabilities and loss estimates they provide to supervisors. Doing that in a format that meets Basel II's detailed criteria demands heavy resources in personnel, software and accounting. D. Wilson Ervin, managing director of strategic risk management at Credit Suisse First Boston, estimates that his bank will need to spend $100 million to comply with Basel II, and Crédit Agricole has projected a similar figure of E100 million ($115 million). Other major banks give numbers ranging from $50 million to $100 million, while financial services consulting firm Mercer Oliver Wyman estimates that compliance could cost as much as 1 percent of assets.

That's serious money, and it explains why banks are determined to make sure they stand to reap some benefits before they start investing big-time. "In this institution there's a sense of quid pro quo. If we're going to implement Basel, we want to have lower capital," says the senior risk officer at a major U.S. bank.

Most big banks remain firm supporters of the Basel process even if they have complaints about parts of the draft accord. "This isn't nirvana, but it's not a three out of ten," says Eric Aboaf, head of regulatory capital at Citigroup in New York. "We think it's closer to a seven or eight. We're trying to get it to a nine." Tom de Swaan, the Amsterdam-based chief financial officer of ABN Amro who headed the Basel Committee in the 1990s, says the proposed accord "is clearly an enormous step forward. The development of thinking among regulators about ways in which banks can measure and manage risk is a remarkable and admirable achievement."

Still, the big differences in how Basel II will affect individual banks are expected to fuel a furious round of lobbying in coming months as bankers seek to win favorable treatment in the final version of the accord. Balancing these competing interests will not be easy. Indeed, in their efforts to cater to everyone, the proposed rules drafted by the Basel Committee please almost no one. The most common complaint is the sheer complexity of the proposals. The latest draft, a weighty 244 pages of rules, mathematical formulas and detailed explanatory footnotes, is "impenetrable," says none other than U.S. Comptroller of the Currency Hawke, whose agency supervises the largest number of U.S. banks. That complexity will make it difficult and costly for banks to comply with Basel II and will increase the chances that supervisors in different countries will apply the rules in very different ways. Indeed, supervisors and bankers acknowledge that confusion about detailed aspects of the proposed rules and different assumptions made by bankers explained a good part of the wide variations in results in the qualitative impact survey. Hence Hawke's desire for another round of testing.

A fair degree of complexity is inevitable given the state of modern risk management techniques. As the Fed's Ferguson puts it: "Risk management has gone a more mathematical way. I'm not going to apologize for that." He insists that the latest draft of the rules, which is barely one third as long as the previous, 723-page version, has been simplified substantially. Critics like Hawke, however, contend that the proposed rules still go overboard. He sees a "strong predisposition on the part of a lot of Basel countries to be very prescriptive, to dictate every jot and tittle of the rules." He blames much of the problem on European members of the Basel Committee, who favor more-detailed rules in an effort to ensure more consistent treatment by different national supervisors. "We are much more comfortable with supervisory discretion than they are," Hawke notes.

Many bankers deride Basel II as the supervisory equivalent of U.S. generally accepted accounting principles. They worry that attempts to lay down detailed rules on the regulatory treatment of each type of credit will merely encourage bankers to find clever ways of flouting the spirit of the rules while respecting the letter, just as corporate accountants exploited loopholes in U.S. GAAP to flatter earnings. Royal Bank of Scotland's Mathewson notes that under the proposed accord banks will be able to reduce their capital requirement by more than 10 percent simply by reducing their estimate of loss given default by 10 percent, a seemingly modest shift that supervisors could find hard to detect or question. "Basel II will provide unscrupulous banks a real temptation to tamper with their internal estimates, especially if their commercial ambitions risk being constrained by tight capital ratios," he says. "Regulators may not have the quantity and quality of staff to effectively police the system on an ongoing basis."

Hawke promises to try to simplify the rules when U.S. officials draw up the domestic regulations to implement Basel II later this year. "What I would like and what's realistically possible are two different things," he says. "I don't think it's going to be possible to achieve any significant simplification" in Basel itself. Similarly, European banks are focusing their attention on the European Commission, which is drafting a capital adequacy directive to implement Basel in its 15 member countries. Italian officials say Tremonti's comments were really aimed at winning concessions for small banks in the EU directive rather than in Basel II. Alexander Radwan, a German member of the European Parliament and a key player in the debate on the directive, says Parliament may seek more freedom in the directive for small banks as a response to the U.S. decision to exempt its small institutions.

All these strategies may be politically expedient, but if they produce a plethora of different national interpretations, exemptions and derogations, they could upset the level playing field that Basel was designed to create.

The second main complaint about Basel II is that it may exacerbate the business cycle by forcing banks to clamp down hard on credit during slumps and allowing them to extend it freely during good times. Corporate credit ratings, which will have a big bearing on capital requirements under the new accord, have clearly deteriorated since the collapse of the technology boom and the resulting recession. A model credit portfolio of some 5,000 listed European companies created by Moody's KMV Co., which provides credit data to banks, indicates that the probability of default for those companies has risen as much as tenfold over the past five years. Deutsche Bank estimates that such deterioration would require a bank with exposure to that portfolio to boost its capital by 300 percent. "If with a model European portfolio you need 300 percent more capital in five years, that is not a good framework," says Deutsche chief credit officer Banziger. "Anybody who's banking with European corporates is going out of business."

Clearly, corporate Germany has been hit hard by the downturn, and German banks have been actively moving to shed bad loans. But Basel's impact would extend much wider. Running its test of the new rules, Credit Suisse First Boston found that it would have had to raise capital by 20 to 25 percent last year to maintain the capital ratio on its corporate loan book. The implication was clear. "Our bank would have cut back new lending significantly more last year if we were in a Basel II world," says CSFB's Ervin. "Maybe we would have been alone, but I have to believe at least some people would have reacted with us."

Many banks simply don't have sufficient data about the performance of their borrowers during recessions to predict the effect that a sharp slump would have on lending, contends Daniel Bouton, chairman and chief executive of Société Générale and head of a Basel working group at the Institute of International Finance. "All simulations that have been made have been made with data from a period of economic growth," he says. "What happens when we are in the third or fourth year of a very sluggish situation? No one has a good assessment." Bouton's call for more study of cyclical effects is regarded by many bankers as an implicit appeal to delay the Basel timetable.

Almost any capital requirement is bound to have procyclical effects. A credit crunch by risk-averse and capital-shy banks helped to deepen and prolong the recession of the early 1990s, particularly in commercial real estate. But regulators say the quantitative impact survey did not suggest any significant problems. For one thing, many banks' internal credit ratings are not as volatile as those of public agencies like Moody's. Bankers, after all, are paid to take a long view. If banks had problems with the Basel II test, it most likely reflected weakness in their loan portfolios. "The most procyclical behavior probably comes from those banks that are badly capitalized or do not have good risk management systems," asserts Jaime Caruana, the Banco de España governor who has just taken over as head of the Basel Committee from William McDonough, the former head of the Federal Reserve Bank of New York who now heads the new Public Company Accounting Oversight Board.

Basel II's so-called three-pillar approach should help counter procyclical effects, Caruana contends. The capital requirements make up the first pillar. The second pillar is closer cooperation between banks and their supervisors, and the third requires banks to disclose much more detail to the markets about their risk profiles and capital backing. Under the second pillar banks will have to run stress tests to show how their portfolios would perform if interest rates spiked up or growth slowed down. That should enable banks and their regulators to set capital at levels that are appropriate for a business cycle rather than a precise point in time. Most banks hold a significant capital cushion over and above the current minimum required by regulators. Once markets get used to the greater disclosure provided by banks, they should become more comfortable with banks, allowing that capital cushion to compress in hard times and expand in good times, or so the theory goes. "Managing the buffer between the regulatory minimum and actual capital you hold is going to fix those problems" of procyclicality, says Thomas Garside, a consultant at Mercer Oliver Wyman.

Beyond the broader complaints, banks cite myriad problems with the details of Basel II. The method of determining capital requirements credit by credit in a bank's corporate and sovereign loan portfolio is inherently conservative and fails to reflect the real benefits of credit diversification, bankers say. The German government made the same point about lending to small and medium-size enterprises, asserting that although individual firms may be relatively risky, the chance that many firms would default at the same time is much more remote. The Basel Committee accepted the argument and introduced a discounted risk weight for SMEs last year after Chancellor Gerhard Schröder threatened to withdraw German support for an accord. So far, however, the committee has turned a deaf ear to the demands of big banks that a portfolio-type approach be extended to their corporate and sovereign loan books.

"We'll certainly try and convince regulators that if you have a diversified portfolio, you should have a lower risk weighting than if you have an undiversified portfolio," says ABN Amro's de Swaan. Spain's BBVA Group believes that the absence of any diversification effect adversely affects its sovereign lending, particularly in Latin America. Sovereign credits make up 20 percent of the bank's E272 billion of assets and generated a 90 percent increase in required capital in the recent QIS test. "It is a key for us," says Manuel Méndez, the bank's chief risk officer.

Big banks active in the derivatives market criticize Basel II's approach to credit default swaps, guarantees and other risk-mitigation instruments. The draft rules allow a bank to substitute the credit rating of the insurer for the rating of the borrower, but the International Swaps and Derivatives Association considers that approach "so conservative as to discourage the use" of such instruments. A bank with a hedged loan would have to experience a double default -- both the borrower and the insurer going under -- to lose money. "The risk of that happening is dramatically lower" than the risk of the insurer alone going bust, contends Adam Gilbert, head of the credit portfolio group at J.P. Morgan Chase in New York. He believes that the new accord should assign much lower risk weightings, and hence capital requirements, to hedged loans. Basel's conservative treatment "sends the wrong signals about the value of credit derivatives," he says. The Basel Committee has rebuffed these arguments so far, partly reflecting a conservative bias among supervisors, who worry that derivatives may be concentrating a great deal of risk in a handful of big financial institutions like J.P. Morgan Chase. But some supervisors also worry that a concession on double default could unravel other parts of the accord because most of the benefits would go to a few big banks.

Banks also have significant complaints about the treatment of certain retail credits, despite the fact that the proposed accord provides its most substantial reductions in capital requirements in the retail arena. Take residential mortgages. ABN Amro's data over the past 15 years show the bank has experienced an average loss given default of just 3 basis points on its Dutch mortgage book, which today totals E50 billion. That would imply a huge reduction in capital requirements if fully reflected, far beyond the 50 percent drop that some banks came up with in the most recent QIS test. Basel Committee members are suspicious of the banks' mortgage data, however. They worry that the boom in residential housing markets in recent years might lead banks to understate the long-term risks, and they don't want to see capital reduced drastically. So they set a minimum floor for loss given default of 10 percent of mortgage loans outstanding in the latest draft rules.

U.S. banks cite historical loss figures nearly as low as the Dutch ones for domestic guaranteed mortgages. They persuaded the committee to allow lower loss estimates for government-guaranteed mortgages but not for those backed up with private insurance. A risk manager at one big U.S. bank says permitting lower loss estimates for the latter category could save his bank a cool $1 billion in capital. Regulators are reluctant to offer any more capital relief on mortgages, though. One possible reason: U.S. banks would be the prime beneficiaries, because private mortgage insurance is uncommon in Europe. So a shift here might upset the balanced compromise regulators are seeking.

Even the biggest mortgage lenders recognize the political and prudential reasons why regulators can't accept their ultralow loss estimates, which would reduce capital to a fraction of a penny per dollar of loans compared with 4 cents today. But by setting an arbitrary floor for mortgage capital requirements rather than following a clear economic rationale, regulators appear to be striving to maintain capital close to today's level rather than letting it reflect real risk. "They're doing their damnedest to make sure we are roughly capital-neutral," says the senior risk manager at one major U.S. bank. "Whenever levels of capital go much below what they expect, it gives them pause. I think they're being unduly conservative."

Bankers and regulators will have to sort out their differences quickly if they want to reach agreement on Basel II by the end of this year. Banks have until the end of this month to make comments on the draft rules contained in the latest consultation paper, and many of them are demanding significant revisions. "We expect they will keep a sufficiently open mind to make changes until the end of this year, and after if necessary," says J.P. Morgan Chase's Gilbert. "If they're only in a tinkering mind-set, then the consultation process won't be very valuable. Support for the accord could drift away."

SO FAR REGULATORS ARE GIVING MIXED SIGNALS about their flexibility. The Fed's Ferguson told the Senate hearing last month that he would support a fourth test of Basel II, something he had previously rejected as unnecessary. He also insists he is in listening mode. "This is a serious comment period. This thing is not cast in stone," he told II. But Caruana of the Basel Committee appears reluctant to consider major changes. "We are very close to the final product," he said in an interview.

Regulators have succeeded to a large extent in defusing the tensions that arose earlier this year over the U.S. decision to apply Basel II to only a handful of big banks. Ferguson stresses that the 20-odd banks expected to adopt Basel II hold 99 percent of the international assets of U.S. banks and two thirds of the U.S. banking industry's total assets. Thus the decision should have no real impact on international competition, a point most European regulators now accept. Ferguson also promises to cooperate with his overseas counterparts to supervise U.S. regional banks owned by the likes of ABN Amro, BNP Paribas and Royal Bank of Scotland, a stance that has been welcomed by the Financial Services Authority's Davies.

Ferguson exudes a firm determination to keep Basel II on track and out of political difficulties in the months ahead. He's not worried about the range of bank complaints -- "this is not a popularity contest," he says. Indeed, he asserts that he would be concerned if they were enthusiastic about the accord, which might suggest they were getting too good a deal. That's not a worry he expects to have anytime soon, though.

"When you get to what appears to be the endgame, that's when you find out what the real issues are," Ferguson says. "'The next year will be interesting."

Basel's three-pronged approach

The setting of international capital standards was a revolution for the banking industry when the Basel Accord was first agreed to in 1988, but today's bank capital rules bear little relation to the actual risk in an institution's loan portfolio. For instance, banks must hold capital equal to 8 percent of a loan's amount whether the borrower is rated triple-A or junk.

The proposed Basel II accord aims to correct this defect by tying capital requirements to the creditworthiness of borrowers. It gives banks three options for assessing credit risk. The standardized approach, tailored for smaller, less sophisticated banks, uses ratings provided by agencies like Moody's Investors Service and Standard & Poor's to determine capital requirements. Companies rated triple-B or double-B would have a risk weighting of 100 percent, requiring a capital charge of 8 cents for each dollar loaned, while a company rated double-A would have a risk weighting of 20 percent, requiring a capital charge of just 1.6 cents on the dollar.

The accord also provides two internal-ratings-based, or IRB, approaches, which allow banks to use their own estimates of creditworthiness to determine capital requirements. The IRB approach uses four key variables to determine capital requirements: probability of default; loss given default, the percentage of an exposure that a bank will lose if a borrower defaults; exposure at default, the amount of a lending facility that is likely to be drawn if default occurs; and maturity.

Under the foundation IRB method, banks estimate the probability of default for their borrowers and then use figures provided by banking supervisors for loss given default, exposure at default and maturity to calculate capital requirements. Under the advanced IRB approach, banks use their own estimates for all four variables. To qualify for the advanced approach, banks need to collect several years of data on the performance of their borrowers to demonstrate to supervisors the soundness and reliability of their rating systems. That requires a big investment in time and money -- Credit Suisse First Boston estimates that it will spend $100 million. Only a relatively small number of big banks are expected to use the advanced IRB approach from the outset. Of the 365 banks that participated last year in the third Basel quantitative impact survey, the test known as QIS 3, only 74 banks were able to calculate capital requirements using the advanced approach.

The accord also gives banks three alternatives for operational risk, such as the possibility of losses because of a rogue trader or lawsuits. The basic-indicator and standardized approaches link the operational risk charge to a bank's gross income. The advanced measurement approach, or AMA, allows banks to determine their own operational risk charge if they can meet strict requirements that include having well-tested risk controls and detailed historical data on any loss exceeding E10,000 ($11,500). Only one of the 365 banks was able to use the AMA approach for operational risk in QIS 3, a fact cited by those, including U.S. Comptroller of the Currency John Hawke Jr., who want more testing before any final agreement on the accord. -- T.B.