How New Global Banking Rules Could Deepen the U.S. Crisis

(page 2 of 4)

What happens, though, if loans and securities held by banks start to go bad all at once, as they are doing today? Neither of the two options that Basel II offers is satisfactory. The first one—shrinking the loan book—would be economically destructive, since cutting off credit in today's recession could send businesses and households into bankruptcy. Indeed, Federal Reserve Chairman Ben Bernanke has found in his academic research that a drying up of bank lending was a major factor in the Great Depression of the 1930s.

The second possibility—raising more money from shareholders—is better for the economy but extremely difficult in a downturn because no one wants to buy. So some banks will simply be swallowed up by other banks or opportunistic investors. The Washington-based consulting firm Federal Financial Analytics wrote positively about Basel II in a study released last December, but said: "The new rules kick in at a time of major credit-market problems, which will mean a sharp spike in U.S. bank regulatory capital." Said the firm: "Significant amounts of risk-based capital will need to be raised in a hurry, driving a new wave of industry mergers and acquisitions."

"IT'S PRETTY SCARY"

It's still too soon to see any constrictive effect of Basel II. Most European countries did not put the rule into effect until Jan. 1. And housing and mortgage markets are still holding up pretty well in most of Europe, notes Jon Peace, a banking analyst for Lehman Brothers (LEH) in London. As for the U.S., where the housing slump and the credit crunch are further along, Basel II won't start to be phased in until 2009.

If the credit crunch gets bad enough, regulators are likely to ease up on enforcement of Basel II's capital rules. But that might not happen until some real economic damage is done. "This cycle is going to turn out to be much more severe than the banks ever expected," says Christopher Whalen, co-founder of Institutional Risk Analytics, a Torrance (Calif.) firm that analyzes bank balance sheets. "When you start scoring the risk of this stuff using the Basel II framework, it's pretty scary."

How did we get to the point where an accord that's supposed to avoid trouble could potentially make it worse? To understand that, you have to go back to the predecessor accord, Basel I, which financial regulators devised in 1988 to get banks around the world to beef up their capital. It more or less did its job: Unsafe banks got safer. "It was a rough-and-ready thing," recalls Paul Volcker, who as chairman of the Federal Reserve until 1987 was instrumental in banging the compromise together.

But banks soon learned how to game the system. To avoid having to tie up capital supporting the mortgage loans they made, the banks got those loans off their books by securitizing them. In fact, Basel I was a prime mover in the staggering growth of the mortgage-backed securities market. BaselI didn't require capital backing for lines of credit as long as they lasted less than a year, so banks responded by issuing short-term lines of credit that they rolled over every 364 days.

Alarmed by the "Asian contagion" financial crisis of 1997-98 and tired of being manipulated by the banks, the Basel Committee on Banking Supervision announced in 1999 that it was taking another stab at the problem. The idea was to align the banks' capital more closely with their actual risks, in the process taking away some of the loopholes that let them hold less capital than they really needed. The biggest banks would be required to use their own computerized models to estimate the probability of default on each loan on their books, in keeping with the notion that no one knows a bank's vulnerabilities better than its own managers.

Up until the past year, loan default rates were exceptionally low, so the backward-looking Basel II rules indicated that banks had more than enough capital. That worried U.S. bank examiners, who didn't want banks to shrink their capital cushions in case conditions got worse.