Even with paper-thin margins and bruising competition, the market continues to grow.

By Simon Boughey & Margaret Eliott

Bad news and swaps go together in many people's minds. It seems these mysterious instruments have caused financial crisis across the globe: think Procter & Gamble or Orange County. Many people may not realize that the swap market is really two beasts. One is the familiar dealer-end-user market, where a bank (dealer) constructs a swap for a corporate or money manager (end-user). Far larger-some 70 percent of total swap volume by some estimates-and in some senses more sophisticated is the dealer-to-dealer or inter- dealer swap market.

By every statistic-except a crucial one-the interdealer swap market has had a banner five years. The overall notional principal outstanding of all interest rate, currency and options has grown from $4.5 trillion at the end of 1991 to $13.9 trillion at the end of the first half of 1996. The average size of the interdealer swap has jumped from less than $50 million to more than $100 million. Five years ago, the maximum tenor of a swap was 10 years; today, tenors of 50 years are possible.

The problem, not surprisingly, is that while this market has matured, spreads have shrunk dramatically. At the beginning of the decade, the interdealer market was a very different entity. Bid/offer spreads were wider, offering commensurately greater money-making opportunities. Spreads to treasuries also moved in a more volatile range, and canny and confident traders earned greater profits in markets that fluctuated. It was not uncommon for swap dealers, say, to receive five-year fixed-rate at 43 basis points against the note and then turn around the position within a few weeks at 38 basis points to the note.

That simply isn't possible today. Spreads have not only narrowed, but the trading range has stabilized within a relatively tight band. As the market has matured, traders have become more proficient and experienced and there are fewer ingenues of whom rivals can take advantage. End-users now are said to deal almost exclusively on the basis of price, not relationships. Not surprisingly, revenues have declined. According to one veteran trader, "Deals are still getting done, but the profitability per trade is down a lot."

Declining profits have led to some consolidation in the industry. It is likely to continue. Five years ago everyone wanted to be in swaps. Now some of the smaller players are pulling back. Some banks, like Republic Bank of New York, have pulled out of the business. Others, such as Westpac and some of the Japanese and French banks, have drastically curtailed operations in this area.

Only perhaps 10 dealers can claim to provide a full range of products on an international basis. Author and derivatives consultant Sayajit Das, who was treasurer of the TNT Group in Australia from 1988 to 1994, predicts that this group will shrink to as few as five full-service dealers by 2000. Deutsche Bank signalled its desire to be one of these five international powerhouses when it purchased Morgan Grenfell and started a relentless hiring binge. The gap between the haves and have-nots of the international dealer market has widened to a point where it is almost unbridgeable.

Of the second-tier banks with more of a U.S. focus, only NationsBank has made a determined effort to get promoted to the premier league. It has followed the example of Swiss Bank Corp., which made its mark in the derivatives world by buying Chicago options boutique O'Conner & Associates. NationsBank bought Chicago Research & Trading in 1994 and within several months had amassed $20 billion in new derivatives positions, from an almost standing start. This example is comparatively rare. The business remains dominated in the United States by familiar names like JP Morgan and Merrill Lynch.

The interdealer swap market has grown more efficient and sophisticated, both in terms of the instruments themselves and the greater skill of those who trade them. Most of the newest instruments-from credit derivatives to total return swaps to spread options-start in the interdealer market before being repackaged by dealers for end-users. This is a credit to the creativity of the interdealer market, but the innovation is greatly aided by two other concurrent developments: increased liquidity and a greater variety of hedging opportunities.

Liquidity is key to the growth of any market and interdealer swaps are no exception. Five years ago, activity was clumped around the shorter end of the yield curve, with 10 years the longest tenor anyone was willing to countenance. Treasury notes were the only real hedging vehicle available, although the Eurodollar strip did help out on the first three years of the swap curve. Now there's activity across the yield curve spectrum. But the behavior of the underlying cash markets has changed as well. And the behavior of the spread between the cash and swap markets is much less predictable. This meant that dealers hedging positions with cash became exposed to spread risk. So with deeper, more liquid markets, swap dealers need a greater variety of instruments and contracts with which to hedge.

The two go hand-in-hand-liquidity and available hedges-so it is hard to say which comes first. The increasing size and depth of the Eurodollar futures market, which allows swaps contracts to be hedged easily and quickly, have greatly aided the development of swap activity along the curve. The Chicago Mercantile Exchange's Eurodollar contract now extends to 10 years and is a far more attractive way to hedge swap contracts because it mirrors LIBOR price movements, as do most swap contracts.

Average daily volume in the Eurodollar contract frequently tops 500,000 and open interest can be between 2 million and 3 million contracts. While liquidity in the back months of the contract may not be what the CME had hoped for, the front years are very heavily traded.

The increasing liquidity has helped pump up the average trade size. A few years ago, trades of $500 million were comparatively rare and, when pushed through, would often cause a market movement. These days trades this size are digested without disturbance. Gary Grella, executive vice president of Prebon Yamane, says, "Deals even larger than that are being done, but not always through a broker, so they can be hard to quantify."

Grella suggests another reason for larger deal sizes. "With profit margins so thin, you need to do size to make any money. Ten $50 million deals will never be as profitable as one $500 million deal because of the resources it takes to get a deal done. Only the big trades let you capitalize on economies of scale." Also, end-user demands from the likes of GMAC, GECC, FHLB and Sallie Mae to swap bond issues from fixed to floating is also driving up deal size.

As the market has grown more mature it has also been able to cope with very long-term deals more comfortably. It is still the exception, but is now more common, to see 15-, 20- and 30-year trades. Dealers sometimes now quote spreads out to 30 years as a matter of course. At the beginning of the decade, dealers could never be sure that they could find an offsetting swap for a 20- or 30-year position and were reluctant to get involved. But the pioneering long end position-taking of shops like AIG and Gen Re has gradually built up a long-term market. "We are now in a position to put on a 30-year trade and feel comfortable about it," comments Bill Huth, head trader at BankAmerica. Nonetheless, savvy practitioners note that the great majority of deals are still transacted in the belly of the curve out to 10 years, and the drastic shrinkage in aircraft financing has cut down the opportunities for long-term swaps.

Dark Days

These days, trading ranges are measured in only a couple of basis points and volatility has declined to such an extent that, for example, 10-year spreads sit at 36-40 for months. In many ways this resilience to external events and large trades is an indication of increasing sophistication and maturity, but opportunities for money making have shrunk. In the period 1992­93 these sea-changes were somewhat concealed by the boom in the structured note business and the revolution in option products, but as these forms of client business have dried up, the stark realities of the interdealer market are evident. "More and more dealers are asking where the next trade is coming from," says Patrick Britt, senior vice president of capital markets at Prebon Yamane.

Interdealer brokers, which have always had to battle tenaciously for business, have been hit particularly hard. In the old days, brokers charged a full basis point in fees, but these days half-point and even quarter-point fees are the norm. Today, brokers are working harder for less money, and their ranks are now tiered between the front rank players and the also-rans. An informal survey of dealers finds the front rank composed of Prebon Yamane, Tullett & Tokyo and Eurobrokers, with Tradition, Garvin Guy Butler, Intercapital, Lasser Marshall and Martin Bierbaum close behind.

In this scrappy world, interdealer brokers have been trying to grow by gaining market share-a sometimes dangerous game. "As an industry," says David Gelber, group managing director of Intercapital, "we have not behaved very cleverly. We repeatedly see brokers seeking to gain maket share by cutting prices, but within a short period of time the competitive pressure builds up and the industry as a whole ends up with the lower fees." Gelber hasn't seen much in the way of consolidation in the broker business. "It's all been on the fringes-Prebon Yamane and Liberty merging their repo businesses, for example. As long as volumes continue to go up-and it has been a very active year for fixed income-interdealer brokers will stick it out," says Gelber.

Brokers are also helped by the easing of credit concerns in the interdealer market. As Prebon Yamane's Grella says, "Even if we the broker are perfect-producing timely prices and good deals-we can't control the credit issues when the counterparties are revealed to one another." In 1990, this was a broker's worst and far too common nightmare-having a deal scuppered at the last moment by credit difficulties, requiring a weak counterparty to step up its bid. U.S. banks and investment firms were the lowest rated in the world and had difficulty getting their names accepted by a wide variety of European houses and virtually all the Japanese houses. Their tiny credit lines to foreign and American banks would quickly be used up. At one stage it looked as if only JP Morgan could compete with the non-American names and trade regularly in the interbank market.

Much to the relief of Wall Street, the situation has almost completely reversed. The Japanese became the poor relations leaders when the Nikkei went into a nosedive and billions of dollars of bad loans were unearthed. At the same time, the credit of U.S. institutions has steadily improved. Last year, Bear Stearns wrote a $200 million two-year swap with German bank Bayerische Hypotheken, an event unthinkable several years ago.

The growth in sophistication of collateral agreements is perhaps even more important. Whereas the chief use of SPVs is still with the client, interbank traders are now frequently collateralized. Until three or four years ago, such agreements were uncommon, but now it is customary for a lower rated counterparty to post collateral to support a trade. Additional capital, usually in the form of cash, is deposited as changes in the mark to market value of the position are recorded. ISDA documentation covering collateral and credit enhancements have kept pace with the market and there is now an established pattern to follow. According to some estimates, perhaps 75 percent of deals in the interbank market are now collateralized.

Many banks have cross-margining agreements with each other because of the volume of interbank dealing between certain names. By netting off a $100 million swap at one tenor against a similar one going the other way at a different tenor, the banks avoid the problem of double counting deals and are left with simply the collateral exposure. Intercapital's Gelber says that often, when the counterparties are revealed to one another as a deal is being consummated, "we will be told that the deal forms part of the mark-to-market deal for these banks." That can also mean that these banks have negotiated additional features, such as the ability to unwind the swap without penalty at some point, say five years into a seven-year swap.

The interdealer swap market will continue to grow, particularly as some of the smaller currencies, like the Austrian schilling and Irish punt, develop into full-fledged markets. Intercapital's Gelber cites the growth of interest in the ASEAN currencies-Thai baht, Singapore ringgit and new Taiwan dollar-as a positive indication of development.

But increasingly, interest is focused on the European currencies because European monetary union looks increasingly likely to happen. The effect is anybody's guess, says Gelber. "It could provoke an explosion in volumes, particularly between the currencies that go in versus those that stay out. Or if most currencies do go in, will the Euro become the third huge currency after the yen and the dollar?"

So expect volumes to increase. Good news for banks and brokers with growing market share. Hard work for those facing shrinking margins and market shares.

Making Money from the Convex Curve

Banks and investment firms trade U.S. dollar swaps to offset positions acquired through customer transactions. They also trade on a proprietary basis using the firm's own capital. Indeed, some of the investment firms and overseas banks in New York run almost exclusively proprietary books. In a classic spread-trading transaction, they would hope to, say, receive fixed rate for five years at 37 basis points over the five-year treasury note, and then pay fixed-rate at 32 a day, a week or several weeks later. Before the offsetting deal is completed, the position is hedged by selling five-year notes or Eurodollar futures. Of course, spreads may actually rise, and the dealer may have to scramble to get out of the position at a loss.

While Euro-dollar prices closely match those in the swap market, the Eurodollar curve is linear, while the swap curve is nonlinear, or convex. The trading opportunities imposed by this mismatch are much more clearly understood than they were in 1990. "The market has a lot more precision and understanding of this issue that it did five years ago," avers Keith Bailey, the head of the trading desk at Merrill Lynch.

"Five or six years ago, only a handful of houses knew what convexity was," says Patrick Britt of Prebon Yamane. Bill Huth of BankAmerica made the point even more forcefully: "Our starting point was total obscurity. I would say that the market was almost totally mispriced five years ago."

Astute dealers were quick to realize that there was an advantage to receiving fixed-rate in the swap market and hedging this position by selling futures. The narrow convexity bias of this position essentially confers a long straddle on interest rates because gains on the hedge will outpace losses on the swap if rates fall. If the convexity bias is wide, the trader could pursue the opposite strategy and sell volatility by paying into a swap and buying futures.

It took the market some time to become aware of these possibilities, and five years ago pricing did not take account of them. Huth recalls that traders were mystified by the willingness of their rivals to receive at what looked like very low levels-either at or just above the Eurodollar strip. But the traders that had calculated the value of the option inherent in the mismatch between a linear futures curve and a convex swap curve knew what they were doing. "Essentially these people were being paid to get long an option. Some people made a lot of money out of people's inability to get to grips with this,' says Huth.

Though the opportunities for this kind of profit-making have diminished as traders have become better educated about convexity, there are still more than a few on the Street that have only a precursory understanding of it. Moreover, estimates of fair value vary from shop to shop, which leads to position-taking that looks profitable to some but not to others.