Q&A: Derivatives Hard to Evaluate But Worth Effort, Expert Says

Gifford Fong is much in demand these days.

In a recent public statement, a newly formed Derivatives Policy Group consisting of six major dealers in the controversial instruments said its members would provide information to help regulators assess the risk each firm had in its portfolio.

The announcement, like those issued by the Group of 30 and others, stressed the need for independent valuation of derivatives and their effect on portfolios as a part of ongoing risk management. The purpose of such independent assessments is to help regulators and firms alike gauge the riskiness of their derivatives portfolios.

Financial technicians such as Mr. Fong have stepped forward to fill this need.

Since 1991, Gifford Fong and Associates of Walnut Creek, Calif., has valued derivative positions for Wall Street firms, insurance companies, mutual funds and other institutions. Mr. Fong, founder and president of the 21-year-old portfolio management firm, said these services help clients determine the quality of their risk management systems.

"An audit process offers the opportunity for independent confirmation that what's reported internally is accurate," he said. "It provides comfort to management and the line people of the accuracy of their derivatives book."

Mr. Fong was interviewed recently at American Banker's New York office.

Q.: Why is finding values on derivative products more difficult than for other types of securities?

FONG: When you value derivatives, you typically can't directly observe what the price is. Most over-the-counter derivatives are very customized and developed for specific end users. Since you can't observe what the price is at any given time, you need to develop a framework with a sound theoretical basis.

Q.: What are some of the assumptions this framework uses?

FONG: If you want to develop a framework that produces a price in interest rate derivatives, then you must define how interest rates will behave over time.

We also assume the price will move in the way you would expect these products to behave in efficient markets, meaning certain relationships would be expected to hold. The most important condition is that there be no opportunity for riskless arbitrage, meaning that the instruments are fairly priced in all markets. Though market efficiency doesn't hold at all times, you need to use this kind of assumption as a base case anyway.

Q.: Is there a general valuation model that companies like yours can use to value these products?

FONG: For short-dated products, as an example, the Black-type model is okay. The reason is that the volatilities required by a Black model can be implied by the volatility of options on futures contracts. In other words, there is an available methodology for estimating the volatility for this kind of model.

Q.: And what is the Black model?

FONG: The Black model was created by the same person who helped develop the Black-Scholes option model. It is used to project the behavior of the derivative you're trying to price because it allows you to observe the instrument's volatilities in a straightforward way. But our approach is to use more than one model so that we get the best possible valuation.

Q.: Does this model apply to longer-term deals?

FONG: For deals longer than five years, the Black model is not adequate. If you use that kind of model in these cases, the projected data become unrealistic.

Q.: Is there a model that bridges short-term and long-term horizons?

FONG: The Heath-Jarrold-Morton model. It seems to be the model more and more street firms are adapting because of the relative ease in estimating volatilities. There is a lot of research being devoted to a model they can use for this purpose.

Q.: So are models a problem in valuation?

FONG: When we do an independent valuation, we want to differentiate between the integrity of the model and the integrity of the input data. We want to know if they have quantified risk accurately and, then, does it result in a reasonable estimate of value.

Q.: And do they?

FONG: Unfortunately, especially in CMOs, it doesn't always happen.

Q.: Why are collateralized mortgage options such a problem?

FONG: Because the cash flows are very uncertain, making the value more uncertain. They are really tough to get your arms around because there is a lot of variability in possible values. That's one of the reasons I think you will see a dramatic decline in their use and availability. People learned the hard way they were difficult to analyze.

Q.: What other types of instruments are difficult to value?

FONG: Long-dated swaptions are the most difficult to value. From a theoretical standpoint, these instruments cause a lot of problems. That is because you're trying to predict the behavior of something that doesn't even exist right now.

Q.: What do you do with these values once you have developed them?

FONG: Once you do the valuation and analysis of risks associated with the product, then you have to extend it to the context of the portfolio to see what the total risk of the portfolio is. Then you are in the position to relate the portfolio characteristics to the stated investment objectives of the investor.

Q.: How can banks use these instruments to their advantage?

FONG: I think asset/liability management will be one of the most important areas. Derivatives provide a flexible, cost-effective way of doing this, because they really deal with the manipulation of cash flows. And that's what you want to do with asset/liability management.

For reprint and licensing requests for this article, click here.
MORE FROM AMERICAN BANKER