When international banking regulators agreed on new capital rules for the worlds leading banks in 2004, they hoped that the so-called Basel II accord would strengthen the global banking system by tying bank capital levels more closely to the underlying risks on banks loan books. But the deepening credit crisis has exposed serious flaws in the accord that threaten to aggravate the economic and financial costs of the global credit squeeze.
The main innovation of Basel II, which was put into force in Europe over the course of last year and is supposed to be phased in beginning later this year in the U.S., was to endorse banks use of complex mathematical models to determine the likelihood that a loan or trading position would go bad, and thus the amount of capital a bank should hold in reserve for that exposure. Banks that didnt have the resources or expertise to develop their own models could simply use credit ratings of agencies like Moodys Investors Service or Standard & Poors to determine capital requirements. The approach was supposed to be a vast improvement over the original, 1988 Basel agreement, which grouped all loans into only a few broad categories and required banks to hold the same amount of capital for a loan to a triple-A-rated corporation as for a junk bond holding. But the credit crisis has undermined faith in models because they spectacularly failed to work for many banks and investors, as triple-A-rated subprime-mortgage-backed securities and other assets collapsed in value and the market for trading them dry up.
Theres far too much reliance on models, says Charles Goodhart, a finance professor at the London School of Economics and former member of the Bank of Englands Monetary Policy Committee. Theres a belief that if you have a clever physicist or a clever mathematician who can feed in a lot of algebra, that somehow makes the numbers that come out at the end more reliable. And theyre not.
The new Basel accord also fails to require that banks set aside capital or make contingency plans for the risk that these institutions could simply cease lending to one another. Such liquidity risk is, by definition, hard to predict or quantify, and bank regulators have debated the issue for years without coming up with a solution a fact that provides little comfort to shareholders in Northern Rock. The government had to rescue the U.K. mortgage lender with some £25 billion ($50 billion) in emergency funds and then nationalize the bank after its wholesale funding suddenly dried up in September and depositors staged a run on the bank the first in Britain in 140 years. Indeed, critics of Basel II point to the fact that Northern Rock increased its interim dividend by 30 percent on July 25, little more than a month before the bailout because, under the new rules, it had excess capital.
The U.K.s Financial Services Authority last month acknowledged multiple failings in its supervision of Northern Rock, saying that senior officials prevented supervisors from examining the bank thoroughly so that the regulator could be seen to be reducing red tape. The agency also disclosed that meetings scheduled with Northern Rock executives to discuss funding risk never happened and that senior regulators had not talked with Northern Rock or questioned its business model in the lead-up to the crisis.
Regulators and politicians around the world are scrambling to plug holes in the new capital rules in response to the crisis. The Basel Committee on Banking Supervision is even considering adjusting sections of the accord to boost capital requirements in some areas, notably for asset securitizations. And the Financial Stability Forum, a group of senior central bankers and regulators that was established following the collapse of Long-Term Capital Management in 1998, is expected to recommend this month that banking supervisors raise capital standards generally, in light of the risks posed by the credit crunch. Such a move would ensure that capital buffers are appropriately forward-looking and take account of uncertainty associated with valuations, the forum said in an interim report to the Group of Seven Finance ministers and central bank governors in February.
The U.K.s FSA issued a discussion paper in December that is likely to lead to stronger safeguards against liquidity risk, including more stress testing and contingency planning by banks and new requirements that they hold a bigger buffer of liquid short-term assets to ensure survival during strained market conditions. The Basel Committee on Banking Supervision published a similar report in February and plans to issue new guidelines this summer on managing banks liquidity risk. It also cited a need for better stress testing and contingency planning, noting that many banks had mistakenly assumed that the markets for mortgage-backed securities and asset-backed commercial paper the very areas where the credit crisis first erupted would be sources of liquidity in a crisis. In addition, the committee has proposed increasing capital requirements for banks trading book exposures to safeguard against losses.
Credit rating agencies, whose lucrative work in advising banks on how to structure complex products like collateralized debt obligations appeared to pose a conflict with their rating of those products, are moving to amend their practices to fend off pressure for regulation. They have endorsed proposals that they be banned from advising on structuring products and are considering recommendations to adopt a different ratings scale for such products than for corporate or government bonds. Both ideas were presented last month by President George W. Bushs Working Group on Financial Markets, which is seeking to restore market confidence.
But even as they move to address Basel IIs weaknesses, regulators are deeply divided over the perceived extent of those defects. Donald Kohn, the vice chairman of the U.S. Federal Reserve Board, expressed the dominant view last month when he told the Senate Banking Committee that the accord was a huge step in the right direction and that any problems could be repaired with relatively simple adjustments. In fact, he insisted, the accord could have averted or softened the credit crisis if it had been in effect earlier because it imposes capital requirements for the first time on banks off-balance-sheet exposures, such as conduits and structured-investment vehicles. Basel II actually addresses some of the issues that have come to light, he said.
Striking a much more skeptical note, however, the chairwoman of the Federal Deposit Insurance Corp., Sheila Bair, told the same committee hearing that the models at the heart of Basel II were flawed, especially in assessing the risk of new products where historical records of default are thin or nonexistent. Models are only as good as the data that you put into them, she said. And in a speech to the Global Association of Risk Professionals in late February, Bair noted that half of the U.S. banks that participated in the most recent qualitative impact survey of Basel II found that the new rules would reduce their tier-1, or equity capital, requirements by 31 percent or more from current levels. Although supervisors can impose higher capital requirements during the initial years of implementing Basel II, she worried that banks would use international competitiveness issues to lobby for lower capital levels.
A race to the bottom in bank capital standards would be a profoundly negative development for the future stability and health of the global financial system, Bair said.
The split among U.S. regulators, combined with the distress of many major financial institutions, could further delay the introduction of Basel II at big U.S. banks. Those banks were supposed to begin implementing the new rules this month, but John Dugan, the comptroller of the currency, said recently that none would do so before July, and even that date might slip because the banks have a three-year window within which to comply. European banks fully adopted Basel II in January, and Japanese banks did so in 2006.
Efforts to prop up Basel II, meanwhile, risk intensifying the current economic slowdown because of the accords inherent tendency to amplify cyclical swings in the economy.
When economies are booming and loan losses are at low levels, bank risk models tend to set lower capital requirements, which in turn allow banks to lend more. But now that the cycle has turned down, critics fear that Basel II will force banks to raise capital at a time when it is scarce and expensive, and to cut back on lending at a time when economies are starved for credit. The LSEs Goodhart says the combination of Basel II with mark-to-market accounting rules, which force banks to write down assets that are falling in value, amplifies the procyclical effect.
The original Basel accord was revolutionary in setting for the first time a uniform requirement that banks around the world hold capital equivalent to 8 percent of their risk-weighted assets. That agreement failed to make much of a distinction between different types of risk, though, and it contained numerous gaps that banks could exploit. The securitization market took off in large part because banks wanted to get loans off their books and make fresh lending commitments without raising additional capital.
Basel II seeks to address those shortcomings with a three-pillar approach. The first pillar sets capital requirements for loans and allows banks to use their own risk models to determine how much capital to set aside. The second pillar imposes a new capital charge for operational risk, which covers everything from computer failures to fraud. The third pillar calls for banks to provide greater disclosure of their loan books and capital backing so that investors can enforce discipline from the outside.
The end result is a highly complex framework that contains numerous political compromises. Proposed capital requirements were shaved for loans to small and medium-size companies in response to lobbying by Germanys Mittelstand; major universal banks succeeded in winning more lenient treatment for securities trading operations and bonds backed by credit guarantees; and supervisory bodies sought to offset some of the capital reductions won by the industry with modest increases elsewhere. The result is far from perfect, most bankers and regulators agree, but after a decade spent painstakingly refining the rules, conducting multiple rounds of testing and having big banks spend $50 million or more apiece to develop the complex computer models and gather the historical loan data required by the accord, Basel II has gained too much momentum to be abandoned.
The new framework has its defenders, particularly among risk professionals who consider new rules overdue in light of the accelerating pace of financial innovation. They note that although damage from the collapse of subprime lending has extended beyond the U.S., in Japan, where Basel II has been in place for nearly two years, risk management practices have improved and systemic risk has been reduced.
I think Basel II is definitely an improvement, says Hugo Bänziger, chief risk officer at Deutsche Bank. In principle, Basel II forces everyone to upgrade risk management. Thats good. It forces bank management under the second pillar to have a risk-reward strategy and define the banks risk appetite; thats good. And it forces people to disclose their risk.
The accord does need improvement, though, Bänziger says. He believes that regulators should reexamine ways to minimize the procyclical effect of Basel II on lending.
European regulators insist they are enforcing Basel II in a way that minimizes its procyclical impact. Daniele Nouy, the secretary general of the French Banking Commission and former head of the Basel Committee who oversaw much of the drafting of the accord, says supervisors are pushing banks to build prudential margins into their risk models, a kind of discretionary buffer designed to raise capital requirements at the top of the credit cycle.
David Brickman, who until last month was executive director of European credit strategy at Lehman Brothers in London, says that such official reassurances ring hollow because they ignore the new dynamics that produced the credit crunch. This cycle is not like any other cycle because of fundamental changes in the way that banks operate, he says. Weve never had a functioning structured-credit market with such liquidity in it and the banks lending so significantly on the back of it.
Computer models also suffer from problems with insufficient data. Cubillas Ding, a senior analyst at international financial research consulting firm Celent, says many bank models are based on limited historical data, in many cases reflecting only the boom conditions of the past five years or so. The rapid pace of product innovation leaves large gaps in what is known about the way many new instruments behave in times of stress, he adds.
Models also tend to encourage herd behavior by banks and investors that increases risks to financial stability, says Avinash Persaud, chairman of Intelligence Capital, a London-based consulting firm that advises investors and governments on investment projects. When everyone is using similar risk-sensitive models, they end up owning similar stuff and end up being forced to sell at the same time, he says. Liquidity is driven by diversity of behavior. Basel II takes naturally heterogeneous market participants a range of market participants of very different investment strategies and it makes them very homogenous.
A sudden drying up of wholesale market liquidity triggered the run at Northern Rock, but the U.K.s FSA insists that Basel II wasnt to blame for the collapse. Liquidity is a fundamentally different type of risk than insolvency, which is the main purpose for banks to hold capital, says Thomas Huertas, who oversees U.K. banking supervision as the FSAs director for wholesale firms. You can have an institution such as Northern Rock, which meets its capital requirements but doesnt have adequate funding liquidity. You wouldnt expect capital charges to be the mitigant for liquidity risk.
But some experts fault the FSA for failing to apply Basel rules adequately at Northern Rock. Tony Blunden, director of Chase Cooper, a London-based risk management consulting firm, says the sudden loss of access to wholesale funding was an operational risk that was almost certainly taken into account, but was deemed so extreme that it was implausible.
Belatedly, regulators are beginning to address the liquidity issue. In its December discussion paper, the FSA identified a need for more-qualitative stress testing to account for a broader range of scenarios. The paper also called for greater quantitative liquidity requirements under the FSAs mismatch regime, whereby a banks assets may be called on to meet liabilities with different maturities.
The Basel Committee plans to propose new liquidity risk measures by July, according to Nout Wellink, the Dutch central bank president who chairs the committee. The proposals will seek to mitigate risks posed by off-balance-sheet exposures and aim to ensure that banks have contingency plans in place in the event that major funding sources dry up for extended periods.
Basel II also fails to adequately deal with the risks involved in major banks securities trading operations, says Andrew Cross, Basel II program director at Credit Suisse in London. In many ways, Basel II is solving the wrong problem, he says. The first pillars capital requirements focus on traditional bank lending, where assets are held to maturity. It doesnt really get to grips with the broad category of structured-credit assets which include everything from CDOs to residential-mortgage-backed securities, he says. Moreover, the accords reliance on quantitative models and its massively detailed rule book have tended to override what he terms an old-style reliance on judgment. He adds, In some ways its become more of a cost of doing business rather than an aid to doing business.
Many of the losses incurred by large U.S. and European investment banks in recent months stem from their trading books, and regulators are now urgently trying to amend Basels treatment of those risks. The Basel Committee in October proposed tightening supervision of off-balance-sheet activities, such as SIVs, that borrowed short term to invest heavily in mortgage-backed securities and other long-term assets. Its proposals call for banks to develop internal models for measuring default risks in their trading books, where assets are held only for sale, and then calculate a so-called incremental default risk charge against those books.
The proposed charge would be in addition to the existing value-at-risk methodology in place for calculating risks on trading positions. The committee aims to make a decision on the proposal this year and impose the new capital charge on trading positions starting in 2010.
Banks and trade groups, including the International Swaps and Derivatives Association, have already begun lobbying to soften the proposal, contending that the new capital requirement is potentially onerous and wont necessarily align capital with the underlying default risks that banks face in their trading books. An ISDA study last year estimated that the new charge could, on average, triple capital requirements on large banks trading operations and produce increases of as much as seven times for some banks, says Ed Duncan, head of risk and reporting for ISDA.
In February, ISDA and several banking industry trade bodies called for more-flexible rules that would permit varying models for estimating trading book risks. But privately, risk managers acknowledge that the enormous write-downs that several major investment banks have taken on trading book assets will probably result in the committees adopting a more restrictive stance.