The Risk Factor
Professor of Finance and
Chair, Finance Department
The dot-com crash that hit at the beginning of the century exposed a gap in the financial analysis of security prices. The dot-com start-ups had been running on credit (often publicly traded convertible bonds), but the prospect of default wasn’t adequately factored into these bond prices.
“There are three sources of risk for a hybrid security such as a convertible bond,” said Sanjiv R. Das, Professor of Finance at the Leavey School of Business at SCU. “Those are equity, interest rate and credit risk. There was no complete model in place to deal with all these, since credit risk wasn’t adequately factored in.”
The fallout was a number of distressed convertible bond funds whose shareholders claimed that portfolio managers hadn’t properly valued these securities. Das decided to create a value-analysis model that would give credit risk its full due.
Working with his long time collaborator Rangarajan K. Sundaram of New York University’s Stern School of Business, Das began to develop a “hybrid” model for valuing these complex bonds and quickly had an early version. They would spend the next five years improving this model and coming up with a formula that would capture the whole picture better than it had been painted to this point.
“As with any theory,” he said, “you posit a theoretically rigorous mathematical structure you believe represents the real world, it’s a physical embodiment at which you throw in data to see if it works. A model should give consistent results or it isn’t any good. Like a good car, it needs to perform well on the road conditions for which it is designed.”
The models that Das and Sundaram came up with evolved over the next five years (“It was a long journey,” he said). The model is a dense, multi-lined equation linked to a larger algorithm that they feel makes a significant inroad into addressing the credit -risk question. A useful feature of the class of models is that they rely only on information readily available in the public markets.
Wall Street was definitely interested and gave the researchers a helping hand by road-testing the models and developing variants of the same. Working papers were put out part way into the process to get comments and criticism, and a securities firm that is a household name invited Das to its trading room and showed him their implementation of the model.
“The model is pretty sound, and traders do like its intuitive nature,” Das said, “but we’re still very early in the life cycle of the idea. Our approach is much more a framework than a specific formula.”
Das and Sundaram published their findings in the September 2007 edition of Management Science. The formula applies to hybrid securities trading such as convertible arbitrage, and any form of defaultable debt. It is also useful for trading credit risk, an area that has been burgeoning in the last five years as investors try to hedge their credit exposure in a volatile and rapidly changing market.
“It’s not uncommon to see firms trading on small differences in price,” Das said. “It’s become a real high-tech business. The difference in value between the market price and what our formula shows may be as little as a tenth of a percent,” he said. “But when you’re talking about hundreds of millions of dollars in bond face value, that’s enough of a spread to drive a tank through.”