Unless you're a big player with lots of clout, don't bother. Nowadays, you can always buy space in someone else's DPC.

By Simon Boughey

In mid-May NationsBank of North Carolina announced it was launching a separately capitalized derivatives products company (DPC). That a regional U.S. bank, albeit one with national and even global aspirations, should embark on this course proves that the lust for triple-A status among derivatives shops has not been quenched. Lesser-rated Wall Street luminaries started the trend in the early 1990s and Asian and European financial institutions quickly followed.

But several years later the economic rationale for a DPC is unproven. "It's like the clamor to become a primary dealer," says Richard Robb, global head of derivatives at Dai-ichi Kangyo Bank. "The first ones in wonder why the rest are joining and within a few months, the last eight in wonder why they've done it."

Another dealer at a firm that considered forming its own DPC, and spent several weeks getting market reaction before dropping the idea, described the process this way: "People were saying, 'Why do you want to form a triple-A subsidiary? Why don't you put your tongue in a light socket? You'll get the same effect.'" These sentiments were echoed by the New York Fed's Economic Policy Review, which in May published an article saying that the DPC approach to managing risk is costly and "unlikely to give DPCs a competitive edge in the derivatives markets as a whole."

In theory, forming a DPC made sense. If a firm could not snare the swaps deals it wanted because its credit rating wasn't stellar enough, it could set up a DPC and run the deals through it. Practically speaking it wasn't quite so easy. End-users weren't always happy to accept the DPC as a counterparty in lieu of a full institution; the rating agencies raised red flags. Though the structures of DPCs have evolved in response to end-user concerns, the ultimate question-what happens in the event of a bankruptcy of the parent-has never been tested. So uncertainties remain about the legal structure of DPCs. The growth in collateralized transactions means that lesser-rated institutions can access deal flow without a DPC. And if these concerns weren't enough, DPCs are costly to set up and administer, and often haven't provided an adequate return on capital even when successful.

All these caveats aside, DPCs can bring benefits to certain players. The original DPC, Merrill Lynch Derivatives Products (MLDP), was formed in 1991. One of the only unequivocal success stories among DPCs, MLDP was capitalized at $300 million to start and in the subsequent four years has added $125 million in capital as the portfolio has grown. But even Merrill has felt that it hasn't used its capital as efficiently as it might.

Big fees

After Merrill Lynch led the way, a number of investment firms and banks formed their own triple-A rated derivatives products companies. In the last five years entrants to the triple-A market have included Morgan Stanley, Lehman Brothers, Goldman Sachs, Bear Stearns and Salomon Brothers from the U.S., Paribas and Crédit Lyonnais from Europe, and Westpac, Tokai and Sumitomo from Asia. The cost of capitalization has rarely been less than US$150 million and has often been US$200­300 million. For most banks these are large sums, and to them must be added the hefty legal fees incurred in consultation with the rating agencies before a DPC is set up. It seems doubtful whether there is enough business to support and generate profitable returns for all these separately capitalized derivatives products vehicles.

The arrangement concluded between Dai-ichi Kangyo Bank (DKB) and Merrill Lynch in February underlined this point. Rather than set up its own derivatives products company, DKB entered into an agreement whereby its swap contracts are backed by MLDP in return for an intermediation fee. The size of the fee varies according to the size of the trade that MLDP backs, and as the trade goes onto MLDP's balance sheet, DKB provides collateral to support it. This is the first time that this type of structure has been implemented, but those in the industry are sure it will not be the last.

Like most Japanese banks, DKB's creditworthiness has been dragged down by bad loans, and the bank is now rated only single-A by S&P and A1 by Moody's. Consequently it found itself shut out of much underwriting business with highly rated counterparties because its credit was not good enough to provide currency swaps, as Merrill had discovered in the late 1980s. DKB began to investigate setting up a triple-A rated DPC in the late summer of 1995, and though the capitalization costs that would have been incurred were not overwhelming for what is still the world's biggest bank in terms of assets, it decided that such a course of action did not make sense. "After looking at special purpose vehicles," says Robb, "we decided it was not a good business to get into." He is very far from convinced that some of the most recently formed DPCs-those established by Sumitomo, Tokai and Crédit Lyonnais-have begun to generate a reasonable return on investment. He doubts whether some of them have even done more than a handful of deals.

One at a time

What is particularly distressing to those institutions that have spent large amounts of capital and time setting up DPCs is that they may not be accepted by end-users. Customers take DPCs on a case-by-case basis and will spend a great deal of time evaluating the pros and cons of each before deciding to accept them-or not. Some end-users still only accept MLDP, and most prefer dealing with a well-rated bank than with a DPC. Jeremy Gluck, vice president and senior analyst with Moody's, points out that "There is often more favorable BIS capital treatment when dealing with a bank rather than a nonbank." Other customers have simply got better things to do with their time than consider whether to accept the latest DPC on the market. "You can go in, plunk down a lot of money, and then go to a customer who says, 'Enough is enough, we don't want to negotiate documents with any more triple-As,'" explains DKB's Robb.

What is perhaps even more illuminating about the whole DKB/MLDP episode is that Merrill Lynch, which runs one of the only unequivocal success stories among DPCs, was just as ready to form a partnership as DKB. DKB's initiative came at a time when Merrill was trying to "maximize the use of capital by running more volume through the structure," admits Chip Goodrich, who at the time was managing director at Merrill Lynch.

If Merrill feels it is not maximizing the use of capital, the theory goes, other DPCs must really be hurting. One of the many functional problems with DPCs is that a universally acceptable formula in the event of bankruptcy of the parent has yet to be established. End-users are confronted by a bewildering array of differing structures, no two of which are completely identical. Merrill Lynch, for example, employs the so-called continuation structure, while Salomon's Swapco has a termination structure. Under the former, if the parent goes bankrupt or slips below a credit threshold, a contingent manager is called in to manage and hedge existing positions; the capital should be sufficient to cover this procedure. If a termination procedure is in place, all transactions are terminated and the counterparty must find a new hedge.

Untested structures

There are a number of variants on these two themes. Crédit Lyonnais New York, for example, which was formed in October 1994, operates a termination structure. In the event of termination, however, investment-grade counterparties can decide whether they wish to close out the positions according to independent third-party quotations or reassign the positions with other dealers of their own choosing. On the other hand, Sumitomo Bank Capital Markets Derivatives Products (SBCMDP), which opened for business in April 1995, offers a continuation structure with the added advantage that counterparties are not reassigned to the parent if it is downgraded. More capital is simply allocated. Both Crédit Lyonnais and Sumitomo claimed theirs was the most "user-friendly" yet devised at the time of launch.

Customers are supposed to like termination structures less than the continuation models. End-users don't like the idea of terminating a large swap before maturity. If one of the bigger DPCs were forced to terminate all its contracts, the market disruption would be considerable, as counterparties trying to find new hedges would find the market moving against them rapidly. "The clients that we routinely deal with far and away favor continuation vehicles," says David Milich, executive vice president and director of Merrill Lynch Financial Products. In fact, in 1994, S&P said it could not "approve" triple-A derivatives products companies with termination structures as "eligible providers of swaps, investment agreements or other financial hedges in structured finance agreements." If a triple-A vehicle with a termination structure were to conduct swaps with an issuer as part of an offering of structured debt, like a callable note, then S&P would use the rating of the parent bank. Mark Gold, senior analyst with S&P, explains that the structured finance group at S&P is worried that an issuer will not be able to rehedge derivatives with an equally good counterparty in the event of termination.

All this is powerful evidence that termination structures are less likely to be popular with certain market segments than continuation formulas. However, as Gold points out, the current number of DPCs is about evenly split between termination and continuation structures. According to Moody's Gluck, this is not only because termination structures are less costly to set up. He notes that if one looks at the portfolios of termination-structure DPCs one will find plenty of end-users, suggesting that their unpopularity has been overestimated. "For termination to come into effect," says Gluck, "the sponsor generally has to fail, which is an unlikely event. My view is that many end-users are not that concerned." The relative success of Swapco, the second-biggest DPC, would seem to support that view. For example, any end-user that is obliged to deal only with triple-A counterparties would not be happy about the prospect of suddenly dealing with a non-triple-A contingent manager.

Double duty

In any case the argument about the popularity of continuation vehicles versus termination vehicles may be moot in light of the most recent DPC. With an initial capitalization of $150 million, Bear Stearns Financial Products (BSFP) offers potential clients a choice between the two approaches by creating a unit within a unit. In the event of default, swaps contracts with BSFP would be continued under a contingent manager, but swaps with Bear Stearns Trading Risk Management would be terminated. S&P's Gold believes that triple-A subsidiaries are developing methods to establish greater flexibility, as continuation vehicles may introduce structural elements offering listed derivatives features while termination structures may develop features for counterparties with certain continuation qualities. Gold believes the Bear Stearns vehicle has embedded this type of flexibility "as an explicit structural element."

Even if differences of structure are ironed out, the growth of the use of collateralization to support swaps has threatened to undercut the rationale for DPCs in the first place. In the last few years the restrictions imposed by credit weakness have been to some extent overcome by the increasing popularity and acceptance of collateral agreements. If deals are collateralized even the most choosy end-users, like the World Bank, have been prepared to ease their credit thresholds for currency swaps. Bankers estimate that fully 75 percent of all interdealer swaps are now collateralized.

But collateral is not a complete answer and its growth is not likely to make DPCs worthless. Some customers still require triple-A counterparties, with or without collateral. Clients also may not want the administrative hassle and cost, not to mention the possibility of error, connected with the marking to market of collateral. Beyond that, as S&P's Gold points out, "not every jurisdiction is collateral-friendly from an enforcement point of view." Even in the U.S., legal requirements have to be met before collateral agreements can be considered binding. The differing collateral management schemes that are about to be offered by the Chicago Mercantile Exchange (CME), the Chicago Board of Trade (CBOT) and Cedel may clear up some of these headaches, but every client of these projects will have to pay fees. All in all, collateralization has eased the credit difficulties experienced by some banks in the early 1990s, but is by no means a total solution for those shops seeking to be global derivatives operations.

In February Bankers Trust (BT) came up with a triple-A derivatives program which utilized both the philosophy of collateralization and that of separately capitalized special purpose vehicles. The BT Credit Plus Program has $100 million of credit support, consisting of $75 million of collateral and a unconditional insurance policy of $25 million furnished by a triple-A insurer. Contracts within the program are collateralized not only by the capital BT has invested but also by receivable swap contracts. Net assets are matched with net liabilities and collateral is called upon only if the latter outweighs the former. Normally a DPC conducts mirror swaps with its parent to offset transactions with a counterparty, but BT avoids this by running a matched swap book within the program.

Although the structures are clearly evolving, the fact remains that many triple-A vehicles have often failed to return what are normally thought of as adequate returns on capital. MLDP, which has a portfolio of $149 billion in terms of notional principal, earned a 3 percent (non-annualized) return on capital in the first quarter of 1996, but the others are probably some way behind that. Satyajit Das, who was treasurer of the TNT Group between 1988 and 1994 and is now a leading derivatives consultant and writer, says that most firms underestimate the costs of running a DPC, in terms of the collateral required and the separate auditing and accounting.

SBCMDP is one of the less high profile DPCs, and managing director Joe Brennan frankly admits, "In the case of ours, if I were to be asked 'Has it gotten to the size and trading we want,' the answer would have to be no." He says that the DPC has been successful in picking up swap business with highly rated counterparties in Europe, but as for reclaiming business in the U.S. with clients who dropped Sumitomo Bank Capital Markets when it was downgraded below double-A, "it is still a work in progress." Brennan concedes it is a "challenge to find applications in the U.S." In the week ending May 3, exactly 12 months after start-up, SBCMDP had 108 outstanding deals on its books worth $7.9 billion in notional principal. This is an effective rebuff to those who sniff that SBCMDP has hardly ever done a deal, but it is not vast.

Hidden profit

It should be remembered that most of the exposure is to highly rated counterparties and so return on capital can be considered virtually risk-free, but strict calculations of volume and capital returns probably give a misleading picture of the value of any DPC. These vehicles were designed as keys to unlock business from which weak banks and investment firms were excluded. In the case of Merrill, it found it could not compete with better-rated rivals for lucrative and prestigious Eurobond business where the issuer required a currency swap. Any number of figures expressing returns on capital do not capture the profit Merrill earns through underwriting, or the enhancement to its prestige and global presence it earns through the capacity to lead manage, say, a global bond issue for the World Bank. It would not have been able to underwrite the hundreds of structured offerings for triple-A rated agencies during the heyday of that market between 1992 and 1993-from which it made lots of money-if there had been no MLDP. "The services of a triple-A subsidiary may be a vital component in concluding a broader financial transaction such as a resecuritization," says S&P's Gold. "The utility of these vehicles for the parent in such circumstances will transcend the relatively small intermediation fee the vehicle earns for its involvement."

But to take advantage of the kind of new opportunities that a triple-A DPC offers, a bank must have a sophisticated and well-established derivatives operation. No one is prepared to admit that their DPC has been a waste of time, but Westpac, Crédit Lyonnais and Tokai were not in the front rank of the derivatives business before the formation of a DPC and have not graduated to this level subsequently. "What is Tokai doing with a triple-A?" asks one banker. Jeremy Gluck of Moody's remarks: "For major players in the market, they make sense; for a smaller one, they don't. It depends entirely on the entity in question."

To its credit DKB realized that, as a smaller player, it was not worth the capital expenditure and the time and expense of dealing with the rating agencies. And its credit intermediation arrangement with Merrill was out of the blocks very quickly. Within 12 hours of its announcement at the end of February, DKB swapped the proceeds of the European Bank for Reconstruction and Development's A$150 million three-year Euronote into floating U.S. dollars-exactly the type of business that the agreement was designed to capture. This instant acceptance of the structure proved its validity and contrasts dramatically with the weeks of deliberation most end-users would have needed before putting on a deal with a new DPC. One or two bankers at DPCs must be wishing they had thought of intermediation first, but at least second-tier names now have an alternative to costly and ultimately ineffective DPCs.

Why NationsBank Is Going It Alone

Given the expense and aggravation of convincing counterparties to use a derivatives product company (DPC), why would NationsBank pick now to set one up? The answer, says William Fall, senior vice president of NationsBank in Chicago, is simple.

NationsBank was able to set up NationsBank Financial Products (NFP) as a separately capitalized company, but still consolidate the $300 million in capital with the parent company for regulatory purposes. That "significantly reduces the cost of our vehicle," says Fall. The other DPCs have been set up by investment banks and must segregate the capital allotted to the DPC. This is one of the reasons that Merrill Lynch has been so interested in opening up its vehicle to other users-to increase flow-through and thus increase return on capital. At NationsBank the capital treatment means that NFP has a lower hurdle rate to profitability.

And of course, NFP brings all the other advantages NationsBank was looking for from a triple-A subsidiary. "As a single-A bank, we were unable to complete over-the-counter transactions with institutions that require a double-A or better counterparty," Fall explains. "We are also seeking to build access to derivatives capability among domestic and international clients of our section 20 company." This subsidiary allows NationsBank to underwrite fixed income issues.

NationsBank considered using another institution's DPC, but rejected the option on two grounds. "It would not have been cost-effective," says Fall, "and it wouldn't have given the right impression of our commitment to be a premier derivatives player."

NationsBank reviewed the other structures used in DPCs and chose to construct one that is most similar to the Merrill Lynch sub. It is an intermediation and continuation structure. That means that a swap written with NFP will immediately generate a mirror-image transaction between NFP and NationsBank. Should NationsBank fail, NFP would continue; the swaps wouldn't unwind.

Perhaps the most interesting feature of NFP, from the viewpoint of other commercial banks, is that it is the first DPC to be approved by the Office of the Comptroller of the Currency (OCC). In the past five years several commercial banks have attempted to set up such subsidiaries but were turned down by the regulator. In commenting on the NFP approval, a spokesman for the OCC says: "We as regulators understand derivatives better than we did five years ago. Since 1993 we have implemented the Group of Thirty recommendations and our comfort level is much greater. That said, we don't consider this structure to be any different conceptually than other operating subsidiaries." NationsBank's Fall says that the application process took about six months to complete.

So as far as the OCC is concerned, a DPC is just like any other operating subsidiary-a securitization vehicle for instance. Sounds like an opportunity for commercial banks to beat the investment banks on their own territory. But the OCC confirms that it has no other official applications on its books at this time. Maybe the rest of the gang is waiting to see how NationsBank fares.