Lee Wakeman, one of the derivatives market's pioneers, is an expert on credit issues. A former associate professor of finance at the University of Rochester, he was recruited by Citicorp to join its London swaps team in 1984, where he worked on some of the earliest yield-curve and value-at-risk pricing models. He went on to senior positions at Chemical Bank, Continental Bank and Sakura Global Capital. In 1992, he founded TMG Financial Products, a derivatives subsidiary of the Mutual Insurance Group of Canada, and served as the chief executive officer until the end of 1996. He is currently working as a consultant for Republic National Bank, evaluating alternative strategies for its derivatives products group. He spoke with editor Joe Kolman in June.

Derivatives Strategy: You've been around the swap market since its beginning, and credit has been an issue that has always intrigued you. How did bank credit departments first deal with this new thing called the interest rate swap?

Lee Wakeman: In the beginning, they had a lot of trouble determining what the loan equivalent of a swap was. The underlining model was still, "How much can I lose? With what probability? And how much can I recover?” The first question, about the credit exposure for a loan, had a simple answer. You could say "I lent them $100, and that's what's at risk, give or take the interest that's due.”

The problem with derivatives was that there was no money out the door, yet something could go wrong. If you do a swap and interest rates move, one of you owes money to the other, and a credit exposure has been created.

At that point, at least two things happened that credit officers had never seen before. First, the exposure was uncertain, but it could go negative, which would blow their minds. Second was the concept of netting. If you have made a loan and you decide to make another loan, you're adding to the credit risk. But with swaps, if you're paying fixed and you want to do another deal where you pay floating, it may actually reduce your exposure by doing another transaction.

There was a general lack of understanding about the credit risk of swaps. It started out in the early 1980s with formulas being applied. If I remember rightly, several banks said, "We'll charge you 1 percent a year for interest rate swaps and 4 percent on currency swaps,” as a way of recognizing that cross-currency swaps were more risky. But 1 percent a year is a linear measure that doesn't match the time path of exposure for an interest rate swap. And it didn't matter that there was a significant difference in credit exposure between receiving the strong currency and paying the strong currency in a currency swap. It was still 4 percent.

In 1985 and 1986, a group of people at Citibank, Chase Manhattan, Bankers Trust and Chemical that I knew were working pretty much jointly on the problem. We actually had meetings and went down to the Federal Reserve and gave lectures, and we modeled the input of movements in interest rates on swap exposures.

When I was working with the credit officers at Chemical, I would say, "The good news is that this 1 percent formula is too high for an interest rate swap, although you need to check whether you're receiving or paying fixed. The bad news is that 4 percent is too low for a cross-currency swap, particularly for long-dated swaps in volatile currencies.”

I remember one meeting in particular with a Chemical Bank credit committee in which we were dealing with a long-dated swap on the Swiss franc versus the U.S. dollar, at a time when the Swiss franc/U.S. dollar exchange rate was quite volatile. We showed the maximum exposure if you received the strong currency—Swiss francs could exceed 100 percent of the notional. Once they realized that was the problem they were facing, we had a lot of acceptance of our exposure models.

DS: But the swap industry always used notional amounts to measure the size of the market, not the real credit exposures.

LW: Yes, in the early 1980s we shot ourselves in the collective foot. We were a small industry proud of what we were doing, and what we were doing was actually very simple. But in order to stress our own importance, we quoted notional numbers. Initially it sounded as though we were doing a big size—$50 million, $100 million. It was only later, when we started doing more and more, that numbers came up to the trillions.

That's when the credit officers started saying, "You've got a lot of exposure.” And we had to turn around and say, "No, it's not really that much exposure.” So the pendulum swung back to the idea that derivatives have much less exposure than the notional. Which is true for interest rate swaps, but not necessarily for cross-currency swaps.

Actually, it's an uncertain credit risk because it's difficult to measure. If you put $100 into equities, you can lose $100. But if you enter into a swap, how much can you lose? The answer is "I don't know. I can tell you with 97 percent to 99 percent probability, but not 100 percent.”

In the beginning I mentioned three elements of credit risk—exposure, default probabilities and recovery rates. Because we concentrated so much on the exposure, we didn't spend as much time on defaults and recovery rates and pretty much left them as the concerns of the credit officer and internal lawyers.

We're now finding that by going into credit enhancement techniques, we can in fact change all three of these elements. We can use credit enhancement techniques to reduce the credit exposure with collateral or recouponing techniques, and credit downgrade triggers to reduce the probability of default.

The second thing we're finding is that there's been a tendency throughout financial history to say, "You made the loan, you sit and wait.” Now we're finding you can move away from that. Now we're saying: "OK, we're prepared to do business with you on this swap or derivative even though you're not an AA credit. But if your credit deteriorates, we want you to start posting collateral at lower ratings.” So we're beginning to manage this process dynamically.

One way we do this in the credit default area is with mutual termination clauses. Every five years, we look at the agreement, and if either of us doesn't like the agreement, we'll wind up the swap. So you can then limit the credit exposure to the next five years.

DS: These types of collateral arrangements have become quite common in the last two or three years...

LW: They've been concentrated mainly on the longer end and on the larger deals, but they had almost no impact as far as I can see on the standard two- to five-year maturity corporate business. Where they come into their own is with aircraft leases and project financing—the long term stuff—at 10 years or more, and especially with cross-currency swaps, where as we've said the risk is larger.

DS: When you were at TMG, you concentrated on longer-dated transactions. What particularly interested you about those kinds of transactions?

LW: We had a good credit rating but we had no corporate client base and no other products to sell in terms of cash management, so we had to look for niches. One niche I call "the crumbs off of AIGFP's tail.” It's basically the high-credit, longer-term structured deal approach. By structuring it well, you can actually reduce the credit risk and your capital reserves and therefore give a better price.

"Howard Sosin didn't get in trouble doing longer-dated deals. He had a disagreement as to what the value of those deals should be. There was never any question that they were mishedged as far as I know.”

DS: Howard Sosin got in trouble doing longer-dated transactions like that when he was running AIG.

LW: No, he didn't get in trouble doing longer-dated deals. He had a disagreement as to what the value of those deals should be. There was never any question that they were mishedged as far as I know.

DS: I suppose he would argue that they were properly hedged.

LW: I don't think anyone disagreed with him that they were properly hedged. I think the disagreement was where the curve should be in 30 years. If you remember, at the time Mr. Sosin was active, there were very few trades at the longer end beyond 10 years. There was a 30-year government, but there were very few quotes off of that for the swaps market. There was also very little concept of what volatilities would look like beyond 10 years. Mr. Sosin took an innovative approach and did the long-term business and created yield and volatility curves.

DS: What do you think of those curves? Were they figments of his imagination or were they based on...

LW: From what I understand, you went out and looked at the shape of the treasury curve and then you looked at any corporate bonds you could find. Then you looked at the swap spreads and interpolated and extrapolated.

It became easier for us when the Resolution Trust bonds started trading. Because of the S&L disaster, they had these extra bonds coming out that were types of U.S. government securities that traded in the range of 25 to 35 years in strips.

DS: And you could use that as a reference point...

LW: It helped us fill in the underlying treasury curve. In fact, it was the first time I showed myself that the curve was downward sloping from 25 years to 35 years.

DS: That was quite an insight at the time.

LW: It was certainly quite a shock at the time.

DS: What did you think before?

LW: Before that, most people had done extrapolation. Depending on the mathematical model used, you tended to have a smooth extrapolation, so that if it was rising from 20 to 30 years, you tended to have it rising thereafter, but at a declining rate so it eventually becomes horizontal.

The other insight came from pricing long-dated bonds such as the 50-year Walt Disney bond. If you started moving your forward curve by 50 basis points beyond 30 years, you moved your price a lot. So it started, for the first time, to place some serious limits on where you can set your curve. Those bonds gave you some benchmarks and Resolution Trust bonds gave you some idea of the shape of the curve. It's now at the point where the 40-year swap market is quoted with a bid/offer spread of no more then four basis points, and the mid-market 40-year quote is consistently below the mid-market 30-year quote.

DS: What's the liquidity like way out there?

LW: I would take your time and have your cup of coffee. You can do size— we have done as much as $200 million in one deal, but it is negotiated. I'm somewhat surprised with how much liquidty has gone into that market in the last year. When we started a year and a half ago, we couldn't find anyone to deal with at 35 years. Now we have our choice of counterparties at 40 years.

DS: Where do you think the credit derivatives business is heading?

LW: It's beginning to break down into two areas. The first is the simple vanilla product—the credit options written on a single underlying stock or bond. They have a very simple exercise condition—default—and a simple payoff function—the difference between par and the market value.

This kind of product does not require a derivatives group. There's no rule that says "Thou shalt have a derivatives group.” You may have to have credit modeling, but it could be within and probably will become part of the credit departments as they start attracting more quants and looking ahead.

DS: It's sort of like the way some banks are moving the formerly separate derivatives group into the cash groups.

LW: Yes. I would say, in fact, that the simple option we talked about will have widespread use among people who never even dreamt of derivatives.

For instance, Barclays gave a talk last month in London about how they intend to take a simple Markowitz portfolio approach to their loan portfolio. They're admitting they have some concentrations and that those concentrations create a somewhat inefficient portfolio. They're not at the efficient frontier. You can apply the Markowitz concept and buy credit derivatives to reduce that risk concentration, and sell credit risk on companies that are not highly correlated with your existing portfolio to cover your costs.

There's another smaller market for more exotic credit derivatives. For instance, the credit option could be written on a basket of stocks or a basket of bonds. It could be triggered by a credit downgrade, and the payoff could be the right to exercise another option.

We had a deal in which the counterparty bought an option on a swap that could be exercised only if a certain airline defaulted on aircraft lease payments. So there was a mixture of underlying credit exposure on the airline with a derivatives evaluation of the swaption. That is where you're going to need both derivatives expertise and credit expertise.

DS: And that's what derivatives groups will do?

LW: I honestly have to say that's very interesting, but it's going to be a relatively minor part of the business compared with the simple vanilla credit derivatives.

DS: The modeling for credit derivatives is based on the simulation technology associated with value-at-risk (VAR). The two seem to be coming together.

LW: Looking ahead I see two developments. We're now at the point in the market where most people understand what VAR is. The next stage is not just reporting the VAR, but going back and explaining how you can manage it.

If, for example, for a 10-day period, you did 10,000 simulations and you're wondering about the 5 percent tail, you're in fact identifying which 500 of the 10,000 simulations had the worst values. What we now do is look back at those 500 worst simulations and ask what values of the underlining risk parameters caused that to occur. What deals were most involved? For instance, if in the 500 you find that one deal turns up 100 times, you might want to consider just getting rid of it.

"It became easier for us [to construct the long-term Treasury curve] when the Resolution Trust bonds started trading. In fact, it was the first time I showed myself that the curve was downward sloping from 25 years to 35 years.It was certainly quite a shock at the time.”

If you need to manage a risk, what we like to say is not: "Here is your limit—you're overlimit or your underlimit,” but, "By the way, you're coming close to the limit and the best way to stay under the limit is to manage that tail of the distribution.”

The second development is that we've been able to apply that same simulation technology to credit risk to negotiate with clients for different credit enhancements and explain to them how much we will change our price if they accept the enhancements. We recently did a 15-year cross-currency swap with a BB-rated counterparty. Normally we wouldn't do a cross-currency swap with that kind of counterparty, but by the time we had created collateral limits and mutual terminations and credit downgrade triggers, we were able to reduce our capital charges to the point at which we could do the deal.

This will be the way it will develop—not just analyzing the risks, but using the credit models to price these risks into deals to give clients an alternative set of opportunities. That is going to be really fun for the people in this area.