BankAmerica Disaster Arose From a Lack of Standards

Martin Mayer

The revelation that BankAmerica lost $372 million in a complicated joint venture with the New York hedge fund operator D.E. Shaw—and in the process acquired a portfolio of bonds with a face value of $20 billion, which could generate further losses—has raised troubling commercial and legal questions about the recent merger of NationsBank into the old BankAmerica. But with any luck, the story should help settle the debate over sorely needed new standards for bank accounting and bank regulation that have roiled the waters of finance for the past two years.

BankAmerica and David Shaw, a former Columbia University professor of computer science, launched their partnership with a fanfare of press releases in March 1997. Their plan was to use the bank’s money and credit and Mr. Shaw’s software to exploit the huge range of opportunities for arbitrage opened by the expansion of trading in roughly comparable debt instruments and derivatives. (A derivative is a financial instrument that makes or loses money according to changes in the price of some other financial instrument.) Mr. Shaw’s software was one of a number of “market neutral” trading systems developed through the analysis of immense quantities of data proving that differences between the prices of financial instruments almost inevitably regressed to historically established ratios.

But the real results of such trading were not even predictable, let alone inevitable, if market volatilities exceeded certain almost impossible parameters. That, of course, is just what happened: The Asian tigers rolled over, the Japanese banks ran out of let’s-pretend room on their bad assets, the Russian state imploded—and so did the computer models. Bids for the instruments the computers had modeled simply disappeared. In August, all the funds following the principles that informed the Shaw-BankAmerica partnership began to give their investors bad news. So far as the outside world knew, however, the BankAmerica venture was the Chevy truck of trading: like a rock. Stockholders in NationsBank and BankAmerica who voted in September to approve the merger had no notion that one of the partners was bleeding hundreds of millions of dollars.

Asked in October why the bank continued to carry its Shaw investment at cost in August, when the losses in the deal became overwhelmingly apparent, a spokesman said the bank had believed that the values would come back with the passage of time—and, anyway, the comptroller of the currency, the bank’s federal supervisor, had known all about it. If Shaw as an investment company had tried to value its holdings in the joint venture at cost, somebody could have gone to jail. But the bank could do so, because banking regulators have long permitted banks to state their assets at “historic cost,” without reference to market value.

Bonds held in a bank’s trading portfolio, as distinguished from bonds in its investment portfolio, must be carried at market price, but the bank itself decides which portfolio is which. In an activity called cherry-picking, bonds that show a profit can be moved into the trading account and sold to improve reported earnings while bonds that show a loss are slotted into the investment account and carried at cost. To keep bonds in the investment account, a bank need merely declare an “intention” to hold the paper to maturity. The chief accountant of the Securities and Exchange Commission in the late 1980s expressed his distaste for this brand of “psychiatric accounting,” but banks are supervised by banking regulators, not by the SEC.

The psychiatric element in bank accounting was then multiplied in the 1990s with the surge in derivatives activity. Because a derivative is neither an asset nor a liability, banks carried this portfolio off the balance sheet altogether. And they claimed the right to postpone the recognition of losses (or gains) in these instruments on the grounds that derivatives were often hedges against changes in the value of the investment portfolio or—taking “intent” to the nth degree—hedges against changes in the value of mortgages or securities the bank planned to acquire in the future.

About a decade ago, the Financial Accounting Standards Board decided that valuations of financial instruments should be standardized across industry lines, and commercial banks—like investment banks, brokers and mutual funds—should have to carry their securities at market price. Federal Reserve Chairman Alan Greenspan objected strongly, warning that if they had to carry government bonds at market price, banks might not buy such stuff, making it more difficult for the government to fund its deficit; and in 1991 the banks were in such trouble that Richard Breeden, the SEC chairman who had led the charge for reform, was pushed aside.

In 1996, however, the FASB went ahead anyway, proposing a set of rules for valuing all bank investments, including derivatives. Again, Mr. Greenspan and Comptroller Eugene Ludwig objected; valuing the assets, after all, is what their bank examiners do for a living. Mr. Greenspan went so far as to write letters to the chairmen of the congressional committees that supervise the SEC opposing such a plan.

But the Shaw-BankAmerica saga makes the FASB case. It demonstrates the need for imposing market-value accounting on banks, especially larger banks, which increasingly rely on the market rather than on deposits for their funding. And it argues decisively that when banks are given expanded powers, which will probably happen in the next session of Congress, they should exercise them only through holding companies that have to meet SEC reporting standards and not—as the comptroller and the Treasury Department want—through “operating subsidiaries” of banks that can cook their books and keep their shareholders conveniently in the dark.