Good Company

In the early 1970s, Wells Fargo's Management Science Department began organizing annual conferences that brought select academicians together with half a dozen bankers to discuss various cutting-edge topics in finance. The 1970 affair featured two young turks named Fischer Black and Myron Scholes, who had just begun thinking about the problem of valuing equity options. Their effect on Vasicek was immediate. “It was like being in heaven, being exposed to all of these new ideas and these people,” he recalls. Later conferences brought Merton Miller, Franco Modigliani, and Robert Merton, who introduced his continuous time equation before it was published.

The conferences awakened in Vasicek a passion for finance he never knew he had. By the early 1970s, he was working with Black, Scholes, William Sharpe, and others to develop for Wells Fargo a radical new investment vehicle known as an index fund. “You cannot imagine how revolutionary an idea it was at the time,” Vasicek recalls. “The basic plan was to form a capitalization-weighted stock market fund. The whole bank went up in arms. The security analyst division was aghast. They said, ‘You mean you want to buy all the dogs along with the good stocks? You're not going to do fundamental analysis of stocks to see which ones are good and which are bad?’ We said, ‘No, the market's already doing that.’” After two years of in-fighting, the team was encouraged to resign. “The bank just couldn't comprehend the idea, because nobody else was doing this,” Vasicek says.

Vasicek headed east to take a teaching job at the University of Rochester. After two upstate New York winters, he went back to California, this time as a visiting professor at the University of California at Berkeley. In 1977, he wrote a paper that would change the face of finance. In “An Equilibrium Characterization of the Term Structure,” published in the Journal of Financial Economics, Vasicek first traced the relationship between the term structure of interest rates and the pricing of bonds. The paper examined how interest rates affect prices of riskless bonds, such as Treasuries. It asked how bonds of different maturities are related to each other, what kind of stochastic processes derived them, and what kind of conditions have to hold across the whole bond market, for all maturities, so that it stays in equilibrium.

“That was, at the time, kind of a novel thing,” Vasicek says modestly. “When I was working on it, the theory available for stocks was the Capital Asset Pricing Model. Somehow, it wasn't perceived as being applicable to bonds—they didn't seem to have any beta. The bond question was hard to get hold of. Then the idea that illuminated my thinking was the consideration that the pricing of longer bonds must in some way be related to what the short rate would do over the tenure of the bond. Now it seems fairly obvious, but it surely wasn't at the time.”

The basic assumption of the theory was that the pricing of, say, a five-year bond is a function of the short rate—or, more accurately, today's probability assessment of the behavior of the short rate—over the next five years. “This was something to build on,” he says. “You simply need to specify what type of stochastic processes you're dealing with. The mathematical implementation was easy; the idea was the hardest part—that the short rate over that span is what should determine the price of the long bond.”

The idea caught on instantly. Within a few years, dozens of articles were popping up in journals that took the concept further. Now, in one form or another, anybody who buys, sells, prices, or structures an interest rate derivative is using some version of Vasicek's model.

Credit Is Due

Despite such success—or maybe because of it—Vasicek realized his heart wasn't in teaching. He needed to be on the front lines all the time, focusing his considerable mathematical powers on the furthest reaches of finance theory. In 1978, Vasicek left Berkeley to become a consultant, eventually ending up at Gifford Fong Associates, where he became a senior research associate in 1980, specializing in mathematical approaches to the newest exotic products in the rapidly growing derivatives markets.

After several years in that role, he left to become a partner in a new company, Diversified Corporate Finance, together with John McQuown, who had first hired Vasicek at Wells Fargo in 1969. The company was built with a groundbreaking mission: to pool bank loans to improve asset diversification. The concept was, by today's standards, quite simple: Banks would contribute their loans to a massive pool of diverse loans in exchange for a share of the entire pool. “They were entirely off-balance-sheet transactions,” Vasicek says. “In effect, they were the first credit derivatives.” But banks were afraid to take the plunge.

Meanwhile, Vasicek was working on some theoretical questions relating to credit. While at DCF, he created a proprietary credit valuation model to help banks evaluate the loans they were contributing to the pool and evaluate the pool itself. The model, later published under the title “Credit Valuation” in several publications, is based on the assumption that credit valuation should not be a subjective judgment of an individual credit analyst, but rather should be inferred objectively from the characteristics of the firm and the firm's share price.

“It sounds extremely natural now, valuing the debt of a company from knowing its equity and derivative asset pricing,” says Vasicek. “We take it for granted. But in the mid-1980s they were absolutely laughing at us.” The theoretical underpinning for the model went back to Black and Scholes, who argued in the 1970s that the stock of a firm is simply a call on the firm's assets. When Vasicek tried to apply that thinking to credit, he was met with tremendous resistance.

“Prospective clients said you value credit by knowing the corporate customer, working with him, analyzing, going to visit, having him visit you, studying the financial statements. It was a completely nonanalytical approach, based on the relationship with the client and on experience. We proposed that the stock market in effect does all that—that it's the aggregate judgment of hundreds of thousands of investors, with the bottom line of their evaluation expressed as the price at which they're willing to buy and sell the stock. If we could succeed in extracting the information from the stock and converting it to the valuation of the credit, we'd capitalize on all this information. In fact, unless the bank credit officer knows something that the market does not know, it should be a superior gauge. You'd have to know more than the aggregate of market participants to arrive at a superior valuation.”

Vasicek was so sure of his revelation that DCF hired a retired former bank credit officer to soothe potential clients. To no avail: The company folded in 1989.

But Vasicek moved forward. After the failure of DCF, he cofounded KMV Corp. with McQuown and Steve Kealhofer, a former Berkeley professor who had also worked at DCF. The new firm beefed up its credit capabilities. In addition to portfolio risk management systems, KMV offers explicit default probabilities from one year to five years for 20,000 companies worldwide. Whereas most portfolio managers used agency ratings to gauge expected losses, KMV offered a quantitative measure—which, among other things, also helped banks price loans and make lending decisions. While Vasicek's pioneering use of quantitative methods in credit analysis would prove instrumental to the credit derivatives boom of the 1990s, demand for KMV's services was nonexistent in the beginning. It took two years for the company to sign up its first client, Bank of America.

Once Bank of America signed on, however, business snowballed. One client led to two, then five, then ten, and before long KMV could count 35 of the world's 50 biggest banks as clients. KMV now models portfolios with combined assets of several trillion dollars.

Nowadays, Vasicek is spending much of his time developing pricing mechanisms for credit derivatives. “You need to know the complete probability distribution of the potential losses in the underlying portfolio before you can fully structure or write, say, a collateralized bond obligation,” he says.

Thanks to Oldrich Vasicek, this is no longer such a difficult proposition.