Credit Derivatives Conference 1999: |
New Alternatives In Credit Management
Excerpts from a roundtable discussion sponsored by Chase Securities Inc. and Moody's Investors Service held November 30 at the Essex House Hotel in New York.
|Joe Kolman, editor, Derivatives Strategy|
Edward Altman, professor of finance, Stern School of Business, New York University
Frank Ronan, managing director for structured credit products, Chase Securities Inc.
Richard Kennaugh, managing director, Chase Manhattan International Ltd.
Carl Schuman, managing director for credit derivatives, Westdeutsche Landesbank
Robert Coors, vice president, Capital Re
Sanjay Mithal, vice president of financial products and services, eCredit.com
Joe Kolman: What might happen to collateralized debt obligations and other credit derivatives in a tough credit environment? What scenarios do you see?
Edward Altman: It's a difficult thing to predict. Collateralized loan obligations and collateralized bond obligations are structured so that you get prepayments. You don't really see a lot of defaults, but people can nevertheless lose money. Defaults aren't the criteria for risk in this market. I'm not a doom-and-gloom scenario guy, but let's face it—we've been living in a very nice world for seven years and that world is coming to an end and defaults are increasing.
The repricing has started already, of course. Yield spreads have widened, but spreads reflect expectations as well as what's happened in the past. There may be higher default rates or rating downgrades. The question is whether you're being paid for it adequately based on what you expect to happen?
Frank Ronan: I think defaults would be good for the development of the credit derivatives market. There are lots of people who have been sitting on the sidelines, who haven't quite bought into the idea of credit derivatives. When there are a few defaults, they realize pretty quickly that this is one of the only ways they can hedge the long positions they naturally take.
We saw it in Southeast Asia a couple of years ago and last fall. A lot of people had looked at the product, but weren't quite ready. But when some of the credits started blowing out, it was pretty easy for them to get products approved, and many became big players in the credit derivatives market within a two-week period. If we have sustained volatility laced with some defaults in the next couple of years, I think you'll see this market grow exponentially.
Altman: I think interest rates are going to go higher before they go lower, but not a great deal. The economy is good, interest rates are relatively low and the stock market is booming, so we're still talking about a fairly benign environment.
|"I'm not a doom-and-gloom scenario guy, but we've been living in a very nice world for seven years and that world is coming to an end.”|
I also agree that defaults may be good for the CBO/CLO market. When those spreads change, they stimulate the appetite extremely quickly. I think we're going to see how robust that market is in a higher interest rate environment.
Richard Kennaugh: Investment-grade defaults don't occur often in developed markets. This is one of the problems for modeling and recovery assumptions. We just do not have the experience of what actually happens, so our models are based on an awful lot of assumptions.
Altman: But you do have downgrades.
Kennaugh: Defaults happen all the time in the high-yield market, and there's a great deal of data about the recovery. The high-yield credit derivatives market, however, is relatively undeveloped. I'm sure that will pick up. The high-yield market is driven by cash-bond investors and bridge-loan lending people. These groups are at the two ends of the spectrum. Now we need to fill in that middle ground, and I think that's the role the high-yield credit derivatives market will serve in the near future.
|"When there are a few defaults, people realize pretty quickly that this is one of the only ways they can hedge the long positions they naturally take.”|
Altman: It's a little shortsighted to say that an investment-grade bond will never default. More than 25 percent of all defaults were investment-grade bonds when they were born. It depends on the tenor, the maturity and so on.
There's never been an AAA or AA bond that defaulted in the course of one year. Whatever you're getting paid to take that risk is too much, in my opinion, because it simply doesn't happen. If you're writing a derivative for five years or seven years, however, you've got to look at what you're getting compared with the probability of default of a portfolio of AAs or As—and they do default, particularly if there's some shock to their industry sector.
So you're quite right. The real test in terms of defaults is now in the high-yield market.
Kolman: I'm confused. What's the difference between writing a seven-year credit default swap and owning a seven-year bond in that market?
Altman: There's no difference between writing insurance and owning a bond. But if you write insurance on 50 of them, you've got a portfolio. If you write one to three of them and you're wrong, those couple basis points are not going to mean too much to you. But there's another element, and that's the portfolio effect of writing a big book, as opposed to taking a bet on one to three securities. The pricing in the market is not on a portfolio basis; it's on a standalone basis.
Carl Schuman: Some people say a default swap is the same as a cash bond. But it's not quite the same thing, unless you structure an extremely unusual credit default swap, where you specify a particular bond as deliverable. But even in that case, it's not the same as the bond, because you may have a short squeeze. When you're looking at this market vs. the cash market as an investor, do you take into account the option of the buyer of protection in a conventional contract to deliver you a fairly wide range of securities?
Robert Coors: Our only concern there is on the credit event of restructuring. There's a lot covered in that definition. A perfect example is RiteAid. Did it default or didn't it? The bonds don't reflect the default. But if people wrote protection on that, under current definitions of restructuring you would have a credit event. There's some disconnect there when it comes down to what we call soft credit events.
Sanjay Mithal: Liquidity also plays a big part. People try to do apples-to-apples comparisons between portfolios of high-yield bonds and credit default swaps or the mezzanine tranches of CDOs. All these securities are being priced as though they were single names, but this market is still quite illiquid. Most of the price spreads, if you decomposed them, would have a liquidity premium that you're paying for holding that security. If liquidity in this market improves, we would expect spreads to come in and the structure to be priced more accurately. Then you'll be able to see a distinction between single names and structured bonds.
|"The liquidity of credit derivatives in developed markets is no worse than it is in the cash market.”|
Kennaugh: The liquidity of credit derivatives in developed markets is no worse than it is in the cash market. It's also worth noting that a portfolio of credit derivatives or floating-rate notes is only the sum of the component parts. We're not really considering the portfolio effects here.
Schuman: From a liquidity standpoint, it's not entirely clear to me whether credit derivatives or cash bonds have more liquidity. We tend to be buy-and-hold investors, but liquidity ought to be a factor. We're not that involved in the high-investment-grade sector, so I can't speak to the liquidity of floating-rate notes vs. credit derivatives in the high-investment-grade sector. We've dabbled in BBB-type assets, but it's not clear to me there's a heck of a lot of liquidity in the bond issues out there.
Kennaugh: In Southeast Asia, when the market started to melt down, the dealers in general were short risk. There had been a lot of people around selling protection, so dealers were always long protection and short risk. So if you have a particular sector that goes to hell, you can generally find someone who's long protection. That gives you some naturally provided buy-side liquidity.
Coors: We've done four or five binary trades, which give us a percentage of the notional value back. The issue arises when there is no other instrument to settle on. You need some type of outcome. We couldn't pick a physical instrument for settlement, but we liked the credit and we became comfortable getting a fixed payoff. The last thing you want is to have a credit event and then argue over how you are going to settle it. We are long credit. We don't trade our book, we are extremely cautious about how we manage the risk of potential credit events.
Kennaugh: A couple of articles of faith are involved if the credit derivatives market is going to grow. The first article of faith is that the documentation will work. The documentation is not perfect, but what we have now is enormously better than the first International Swaps and Derivatives Association agreement. In another year or two, it will be redrafted again. There are some differences of opinion that will probably never be resolved, but eventually an acceptable compromise will be reached, and everyone in the market will use the same contract, which will reduce documentation basis risk.
The second article of faith is the equivalence of recovery. If you're going to lose sleep about being delivered something that may have a recovery different from assumptions, you shouldn't get too involved in the credit derivatives market as it now stands. You'll have to accept that the contract specifications allow for delivery of certain instruments that will give you a broadly similar recovery. If you start looking at issues such as post-default liquidity, you're asking for trouble. Some people say, "I don't want to have loans delivered, because they're going to be less liquid than bonds that default.” Well, who knows? It's all probably going to be pretty damn illiquid, and what you should be concentrating on are the details of the workout.
|WestLB Tackles Capital Efficiency
|Carl Schuman, managing director for credit derivatives at Westdeutsche Landesbank, talks about his bank's use of credit derivatives as a synthetic investment tool.
What motivates us to use credit derivatives as a synthetic investment tool? While we have a large portfolio of asset-backed securities in New York, we have little in the way of corporate bonds. We've also reduced our exposure to high-grade U.S. corporate loans over the years as a result of a focus on the return on regulatory capital. As a result, of the 80 or 100 names that trade actively in the U.S. credit derivatives market, we have essentially no exposure to half of them. So we've looked to the credit derivatives market as an alternate way to take risk in those names.
Since we're constrained by regulatory capital, we need to find ways to make those investments capital-efficient. Consequently, we've been involved in first-to-default basket-type structures, in which we'll take four or five names the bank has little exposure to—or some exposure to but not approaching any kind of cap—and we'll put them in a basket and take a first-to-default position, and in that way leverage the returns. You've got to be aware of what your own regulators think about those structures. A common regulatory approach to that structure is to say your maximum loss is to only one of the names and therefore the capital charge is equal to 8 percent times the risk to one of the names.
In our experience, if you basket together five names with a 30-basis-point default swap quote, you can earn 100 basis points to120 basis points from a dealer paying for protection on a first-to-default basis. The less the correlation in the names, the more of the available spread you're going to get. While many dealers are unwilling to pay such a high percentage of the total spread for first-to-default protection, we have consistently been able to earn 60 percent to 75 percent of the sum of the spreads.
We're involved in single-name default swaps in which the spreads meet certain return objectives, which basically takes us out of the high-investment-grade market. We cannot do deals for 30 basis points. On the single-name level, we work for triple digits. Given our conservative credit culture, it's a needle in a haystack to locate the names that work for us at that 100 basis-point to150 basis-point pricing level, but there are plenty of names in the BBB- sector we don't have exposure to and the odd one works. I don't think we've ever done a deal on a name that didn't have at least one investment-grade rating.
The new impetus for us as investors is to be balance-sheet suppliers. We've detected a rise in the pricing of total-return swaps in which we provide a balance-sheet rental. The trade is simple: We buy an asset and then transfer all of the credit risk of the asset back to the bank that sold us the asset. For the counterparty, it's a way for it to take the risk of an asset but not have to fund it. Funding is increasingly a primary constraint affecting the big commercial banks—and certainly has always been a big constraint for the large Wall Street firms. We've seen the pricing come up to a level where we like the asset class, and we're happy to take large exposures to the biggest New York banks, European banks and investment banks at spreads that are more attractive now than they have been at any time in the past. We're also seeing people wanting to do total-return swaps further out the curve and that works for us as well. So those are our investment parameters and rationale.
|Chase on Credit Derivatives
Bob Strong, chief credit officer and executive vice president at Chase, highlights the bank's approach to managing credit risk.
|I will sound a few key themes in our philosophy and highlight some of the limitations we're dealing with in terms of credit risk management tools. We have a loan portfolio of approximately $223 billion, about 54 percent wholesale and 46 percent retail. Most of my comments here relate to the wholesale side of the business.
For starters, we don't attempt to micromanage the bank's credit portfolio. We charge the business units with that responsibility. The entire central credit function at Chase is 72 people, and half of them work in a special loan (work-out) group. There's only a small group in the middle to manage the book. Our fundamental approach is to achieve alignment across the organization so we can have the kind of portfolio we'd like to have—and we use capital as the means to align people.
A basic premise of our wholesale business is that we originate for distribution. Everything we do, we look to sell. We believe the market is a terrific judge of credit—the underlying, the structure and the price we're putting out. We see the market as a second set of eyes that helps us manage our balance sheet and our credit risk. Consequently, our portfolio has been flat for the last two years. Although we originate more than $400 billion a year in volume on our syndicated finance side, we haven't grown the portfolio in the last couple years because we distribute more than 90 percent of what we originate.
One of our aims is to manage risk by maintaining low volatility and stability. We acknowledge we're in a risk business and we have an expected-loss component in our business—which we believe is 40 basis points to 60 basis points per annum on the wholesale loan portfolio, and is lowered a little once we add in risk management instruments. We look at risk as our exposure to unexpected loss—and that's what we carry capital for.
In general, we take a market-based approach to managing risk. We don't look at regulatory capital. Economic rather than regulatory capital is the key link for us in terms of managing the business, and we have high capital ratios. The credit function has sole responsibility for determining the risk grade of everything we put on, so we have a risk grade for every facility, structure and client—and that's a credit function. The business units have no authority to set that.
Our rating system has 25 grades. We feel it's important to have that level of granularity and accuracy, particularly in terms of pricing into the public market. Once we've determined the credit grade and tenor of a transaction, we determine the capital allocation and the provision we charge the line unit—the provision is equal to expected loss. The line is hit on Day 1 with that, so in good times they're not incented to build up the balance sheet; in difficult times, they get a bit of a carry, but it works out over time. So we continue to operate on the basis of expected loss.
We believe the market determines the price of credit transactions. People often say that Chase sets the price, but we don't agree. Our view is that the credit function makes a decision on what the risk is, and that determines the capital, the market determines the price, the price over the capital determines the yield—and it's up to our business units to determine whether or not they want to do that business. We hold every business accountable for achieving a 13 percent minimum return on risk-adjusted capital. That's what fundamentally drives our decisions in terms of buy-and-hold portfolios, and it's also how we price for the market.
Beyond pure economics, we make adjustments to capital in order to motivate behavior on the part of the line, which they might not otherwise do on their own accord. For instance, if we look at the actual balance sheet usage for our extremely high-quality borrowers, on a risk-adjusted basis we wouldn't need to hold much capital against those transactions. We don't wish to balloon our balance sheet, however, so we add a capital charge for that. As a result, we have to get a higher return on even the best-quality business in order to put it on our balance sheet, because we care about nominal leverage.
We also have, according to the line at least, a draconian approach to excessive concentrations. We have a catch phrase called threshold levels. It doesn't roll off the tongue the way hog shares does, so it has become known at Chase as the hog share tax. What happens is that when there's a transaction that exceeds the threshold levels we've established, the line can go ahead and do the deal as long as the credit function approves it, but we charge a 50 percent premium on the capital in excess of the threshold. That incents people to get as granular as possible, and gives us a screen when transactions exceed that so we can make sure we're comfortable that the transaction is actually worth it for us. Concentrations, after all, are clearly a killer for a bank. Finally, we have individual concentrations built into our capital allocation. If there are individual units we want to encourage or discourage in terms of their share of the portfolio, we can moderate their capital allocations—and that will flow through our shareholder-value-added return calculations and over time, hopefully, we'll get a correction.
As we see it at Chase, concentration management is the legitimate central function of credit. It's a key overlay to what is essentially an independent business model here. We let the businesses manage their credit risk, while we look at overall concentrations within the businesses. We manage by individual obligor, we manage by the obligor risk grades and what concentrations we want in that regard, we manage by industry, and we manage by country. The latter has been a particularly interesting area in the last few years, since we've dramatically reduced our cross-border exposures.
We establish threshold limits on a notional basis and on a capital basis. They are different notionally by risk grade, but they are identical in terms of capital. We have a maximum capital limit of $3 million for a given obligor. Since we allocate about $5 billion of capital to the wholesale credit products, we're obviously trying to get quite granular in terms of our wholesale mix.
Over the last few years, we've used credit derivatives in our wholesale area. We rely primarily on our distribution side, but we look at credit derivatives on the margin to help us manage some of these concentration issues, and within individual business units we use derivatives quite a bit more to bring down individual positions and escape the hog share tax. We also expect credit derivatives to play an increasing role in the secondary market at Chase, and we're looking forward to the enhanced liquidity we think will come with that product over time. Liquidity has improved in the last year or two, and we think it will continue.
There are also a few limitations with some of these products. The first has to do with model risk. We spend an awful lot of time taking inventory on and validating our models. The more time we spend on them, it seems, the more time we've got people working late at night creating new models that allegedly show returns in excess of our minimums. We need to be careful we've got the models in shape, however, since this is a market and credit risk issue. At Chase the central areas of both functions are responsible for inventory and validating all the models we use in the bank.
Another concern is stress testing. Stress testing has been particularly robust in our market risk environment, where we do 10 scenarios. We run the portfolio monthly and try to figure out what we think the worst cases are. Stress testing is a key component of our risk management process.