A spectacular event in August 1996, apart from the Olympics in Atlanta, was the closure of two derivative product companies. The first, Fisher King Derivative Products (Fishco), a termination vehicle, was wound up when its parent Fisher King & Co fell below BBB investment grade.
That forced the termination of over 500 interest rate and currency swaps, caps, floors and other structures, with more than 200 counterparties, totalling over $100 billion in notional principal amount.
The market was hit by a wall of demand for close-outs within the next 10 days, during the least liquid month of the year. The $800 million of collateral posted by Fisher King with Fishco to cope with just such an event, together with its core capital of $100 million, had been well above the mark-to-market value of Fishco’s exposure to its clients.
But in the subsequent market disruption, spreads on over-the-counter derivatives widened so far that mid-market close-out values were highly unstable. Disputes about the mid-market price paralyzed the close-out beyond the 10 days prescribed by Fishco’s wind-up schedule. Irate counterparties reached for their lawyers, and immediately reviewed their master agreements with all other triple-A derivative product companies (DPCs).
Within days the market received a second shock. Ninja Bank, the Japanese parent of a London-based DPC, a continuation vehicle, declared bankruptcy. As expected, that triggered an orderly assignment of the clients’ swaps to a contingent manager, Bank of New Hampshire. But dealers at Ninja Derivative Products (NDP) were faced with replacing in the market all the hedges that its parent had put in place to offset those client swap positions. NDP’s stock of $200 million capital and $350 million in collateral held out for a while, but as the market realized NDP’s precarious state, the cost of the hedges escalated – until NDP, like its parent, faced bankruptcy.
In both cases the losses to the DPCs’ counterparties were not huge, and there was no great anxiety about a threat to the world financial system. But the illusion that these vehicles were much safer than the parent was shattered forever.
Fantasy? Perhaps. However, there is plenty of resistance among derivatives users and dealers to the idea that a subsidiary of a bank or company rated single-A or double-A can enjoy a triple-A rating.
Since 1991, 13 derivatives dealing houses – mostly banks or investment banks with a single-A rating – have created bankruptcy-remote subsidiaries, or guaranteed pools of collateral; and, having secured a triple-A rating for these vehicles, they have offered them as more acceptable derivatives counterparties, in order to gain access to the world’s choosiest, top-rated clients. But not everyone is impressed. Instinct rather than cold logic still governs that final decision: to deal or not to deal with them.
Gut feeling
“No company can be rated higher than the country in which it operates,” argues a British clearing banker. Logically or not, his credit department refuses to regard a triple-A rated DPC as any better than its parent. “The existence of a DPC doesn’t add to my appetite for exposure to the group,” the banker concludes.
Some end-users have a similar gut feeling. “We stripped them down and decided something wasn’t right,” says an investment manager at a major UK-based fund management firm. “That’s not to say we don’t use them,” he concedes. “But we have a mechanism whereby we can insert a true triple-A credit as intermediary.” The head of investment at a UK insurer restricts his exposure to short-term deals. “We just have a feeling there might be some fancy footwork ahead,” he says. “Whereas you can choose a counterparty that has been run successfully and strongly for years.”
DPC supporters detect most hostility among top-rated banks that are derivatives players. “Most of the criticism is sour grapes from people who had this market to themselves,” says Steve O’Connor, at Morgan Stanley Derivative Products. Certainly the big banks’ risk managers have gone about their due diligence with a vengeance. A note from one such to his colleagues, scrawled on a copy of a Salomon Swapco interim report, reads: “Let’s meet and rip this apart.”
Interestingly, this risk manager concedes: “We came to the conclusion that we had no case to make against the DPCs.” This doesn’t mean his department has stopped trying.
Top-rated banks are not target clients of the DPCs. They have little reason to deal heavily with each other. In general, DPCs have been designed to expand a lowly-rated dealer’s ability to deal with end-users who can accept only highly-rated counterparties.
DPCs broaden the choice for such end-users. To that extent they diversify the users’ risk away from what had become a limited choice between a handful of highly-rated derivatives dealers, such as JP Morgan, SBC Warburg, Barclays, UBS, etc. Even if there is some doubt whether DPCs are copper-bottomed triple-A risk, they could still be a valid means for users with restrictions on the quality of their counterparties to diversify their exposure.
It can even be argued that DPCs’ derivative portfolios are safer, and subject to much closer scrutiny, than that of a big bank: weekly mark-to-market by an outside firm of accountants; regular oversight by one or more rating agencies; extra oversight by securities or banking supervisors. “You could say that these vehicles have a more transparent risk than that of a bank,” says Adrian Maconick at Deloitte & Touche Consulting Group. That sentiment is echoed by Stephen Mulliner at Tokai Derivative Products: “Their triple-A rating, given the conservative capital requirements and absence of a loan portfolio, is arguably better than that for a standard bank.”
Officialdom appears to share that view. Last month, the US Office of the Comptroller of the Currency (OCC) raised no objection to the creation of Nationsbanc Financial Products (NFP), the first bankruptcy-remote triple-A dealing subsidiary of a US Federal Deposit Insurance Corporation (FDIC)-insured bank. Senior deputy comptroller for capital markets, Douglas Harris, affirms: “We found some very positive reasons to approve [the application].” Four years previously the OCC took a different view of applications by Citibank and Chase Manhattan to create triple-A DPCs: Citibank and Chase withdrew their applications. Harris comments: “Over the last few years the agency has gained a much better understanding of derivatives and competitive forces in the derivative markets.” NationsBank, with its single-A rating, would be “in a better competitive position,” he says.
When Citi and Chase applied, say sources close to NationsBank, they were fighting to retain market share while their ratings were slipping. NationsBank today is getting a better chance to compete on equal terms with securities houses equipped with DPCs, says Harris. The alternative for the bank would have been to work on improving its rating to at least double-A. “NationsBank are new boys in derivatives, and trying to expand in all markets,” says Duncan Goldie-Morrison, the bank’s global manager for interest-rate products. “They don’t want to get their rating up, and risk reducing corporate flexibility, just for our [derivatives] business.”
Jim Leach, chairman of the US House of Representatives committee on banking and financial services, wrote a glowing letter to comptroller of the currency Eugene Ludwig on May 17. His argument is that by upgrading the quality of its counterparties “the credit quality of a national bank operating its derivatives portfolio in a prudent manner may rise and its overall derivatives business strengthen upon formation of such a triple-A rated subsidiary”.
With such an endorsement, who needs to rip these things apart any more?
The NationsBank development emphasizes that there is still mileage in the DPC concept. In some ways NFP is the most efficient yet, since the OCC will allow the subsidiary’s capital in NFP, initially $300 million, to be consolidated for group capital adequacy purposes.
Most DPC variations, since the creation of Merrill Lynch Derivative Products in November 1991, have been attempts to mitigate the up-front capital cost. There are three basic structures, but all of them seek to separate the credit risk of over-the-counter derivative transactions into a separate vehicle or guarantee programme, while leaving the parent to manage the market risk. The vehicle or programme carries capital, insurance support or collateral to cushion credit defaults by counterparties, and to hedge out or unwind the portfolio if the parent defaults or any other specified trigger event occurs.
Continuation
The first example of the continuation structure, Merrill Lynch Derivative Products (MLDP), the first ever triple-A DPC, is designed so that even if Merrill Lynch & Co goes bankrupt MLDP will continue to honour its obligations, until all existing contracts with its counterparties have expired. Merrill Lynch put $300 million of capital into MLDP and sold $50 million of preferred stock to investors. For any contract which MLDP writes with a client, it writes a mirror contract with Merrill Lynch Capital Services. This means that, provided there is no trigger event, all market-sensitive flows pass between the client and the Merrill group, via MLDP. MLDP remains market neutral.
However, it does have a potential market exposure. If there is a trigger event, such as a default by MLCS, MLDP ceases to take on new business. Its portfolio passes to its contingent manager (specified in early documents as Sumitomo Bank and Mitsubishi Bank but now named only as at least a single-A credit). MLDP’s capital and any collateral posted with it by MLCS are then used to replicate the hedges that were lost when MLCS fell away as a mirror counterparty. MLDP must always carry enough capital and collateral (sometimes called sensitivity collateral) to replace those hedges within a minimum of three weeks even in the most turbulent markets. At the end of last year, in addition to its $410 million stockholders’ equity, MLDP carried $1,204 million of collateral to offset its estimated potential exposure to MLCS. At that date the positive mark-to-market value of its swaps with MLCS and an affiliate was around $400 million.
On the basis of stringent checks by accountants and satisfactory risk-modelling, Moody’s Investors Service, Standard & Poor’s and Fitch have consistently given MLDP a triple-A rating. Their criterion is not that MLDP is 100% immune from default, but that the likelihood of default is consistent with that of conventional triple-A ratings. However, the ratings don’t address the snag that, after a trigger event, MLDP’s counterparties will find themselves exposed to a counterparty with an inferior rating.
Since Merrill pioneered the structure it has been imitated by five houses: Sumitomo Bank (SBCM Derivative Products Ltd), Tokai Bank (Tokai Derivative Products Ltd), Lehman Brothers (Lehman Brothers Financial Products), Bear Stearns (Bear Stearns Financial Products) and NationsBank (Nationsbanc Financial Products). The Japanese versions were more-or-less direct copies, once they had been approved by Japan’s ministry of finance. Both are based in London.
MLDP’s successors have less capital, but in most cases their capital is handled dynamically, being allowed to fluctuate according to the size of the derivative book. LBFP has fine-tuned its use of capital and collateral further by having the ability to reduce its potential market exposure to its parent by means of its own internal hedges – taking positions in cash instruments or exchange-traded products. “The capital requirement is potentially more economical,” says Reza Bahar, managing director of S&P’s derivative ratings group, “because it decreases the volatility of the portfolio value.”
The continuation structure appears to be the preferred choice for end-users which use accrual accounting and which would be uncomfortable replacing hedges if they were suddenly terminated.
But there is fierce, almost religious, debate about the relative merits of continuation and termination structures.
Termination
Salomon Brothers introduced the first termination structure, Salomon Swapco Inc in February 1993. Instead of the complication of a contingent manager and keeping enough capital to run the portfolio after a trigger event, Salomon opted for a portfolio that forces termination at mid-market prices if a trigger event occurs. Trigger events include bankruptcy of any part of the Salomon group, failure by Salomon to post sufficient collateral with Swapco, or failure by Swapco to maintain sufficient capital, liquidity or a rating of at least single-A.
A counterparty that does a swap with a termination vehicle is therefore taking a risk on the parent. If the parent defaults, it has a triple-A chance of being paid the value of the swap at mid-market price; but if it wants to replace that swap in the market it has to pay the bid spread.
End-users, such as the World Bank, tend not to like this risk. But dealers which run a liquid portfolio of swaps at mark-to-market value are more-or-less indifferent to the risk of termination and see a price advantage. “There’s no pain and no gain, unless we take that kind of potential risk,” says Yuichi Hiromoto, general manager of capital and money markets at London-based Mitsubishi Corporation Finance, an arbitrage house.
Swapco and other termination structures are, in theory, less costly on capital. Instead of keeping enough capital and collateral to cover the potential credit risk of counterparty contracts until maturity, and sensitivity collateral from the parent – like the continuation DPCs – termination vehicles need only enough capital to cover counterparty credit risk and market risk during the close-out period, which is usually around two weeks. “The termination structure has a sort of rolling life,” says Ken Nadel, president of LBFP. “It rolls up in windows of several weeks and it says, ‘You’ll be repaid, but at any time you could be taken out of your trade'”. For this reason the capital charge on termination structures tends to be less than for continuation structures like LBFP, Nadel says, particularly for something as uncertain as a long-dated currency swap. S&P alerts potential counterparties to the special properties of termination vehicles by rating them AAAt. Moody’s gives them a straight Aaa, arguing that there is a triple-A likelihood of default, and that the procedure after a trigger event is irrelevant to that rating.
Apart from Salomon’s Swapco, there are three similar termination structures: Morgan Stanley Derivative Products (MSDP), Westpac Derivative Products (WDP), and Bear Stearns Trading Risk Management Inc (BSTRM) – note that Bear Stearns has a continuation and a termination structure.
MSDP claims that it has developed a more efficient use of capital: its calculation requires no additional collateral. Its termination procedure is also more transparent, argues MSDP, since it will take quotes from a list of major dealers to determine the mid-market price. Swapco takes it on itself to determine the mid-market price.
Westpac created its termination vehicle WDP in March 1994, a time when it was rated only single-A. Now that its rating is on the mend – it bought back shares last year and acquired Challenge Bank – its derivatives counterparties are less insistent on using WDP. Apart from its $100 million capital WDP has the additional comfort of a $100 million surety bond from Capital Markets Assurance Corporation (CapMac), and, because of concerns about Australia’s own AA rating, all WDP’s capital and flawed deposits (collateral irrecoverably posted by the parent), are held in an English trust under English law. Even so, one of WDP’s trigger events is a downgrade of Australia’s sovereign debt rating below A-.
Bears Stearns’ second vehicle, BSTRM, is a subsidiary of the first, BSFP, and would attract those who prefer the clean exit offered by a termination structure. “We wanted to appeal to as many customers as possible,” says Wendy de Monchaux, senior managing director at Bear Stearns. Bahar at S&P sees the Bear Stearns solution announced in January as a testament to how far the concept of the triple-A DPC has developed: “Clearly the market has an appetite for both structures,” he says. Nadel at Lehman admits the value of termination for some deals: “We’re working on putting together a programme to allow termination trades – guaranteeing trades of other Lehman entities,” he says.
Guarantee programmes
Some commercial banks chose a different route to a triple-A rating by guaranteeing a pool of transactions done within the bank. The advantage is that counterparties still deal directly with the bank but have the comfort of a triple-A rating. These programmes are termination structures, rated AAAt by S&P, although one of them, Crédit Lyonnais, offers continuation as an option.
Paribas Dérivés Garantis (PDG), established in January 1994 as a subsidiary of Compagnie Financière de Paribas, is a financial institution supervised by France’s Commission Bancaire, but cannot take deposits or make loans. PDG has a capital of Ffr800 million ($154 million) but has a call on subordinated debt from its parent up to Ffr2.7 billion. Some programmes have mutual collateral agreements with counterparties to reduce their credit exposure, but PDG doesn’t. “If the counterparty wants to limit exposure through collateral, it accepts [Banque] Paribas,” says Denis Bertin, financial controller at PDG. “We don’t have counterparties that require collateral and PDG,” says Bertin, but he adds: “We may do it soon if there’s a request.”
Crédit Lyonnais established a similar vehicle, the Crédit Lyonnais Derivatives Program (CLDP), a year later. Guy Laffineur, global head of derivatives at Crédit Lyonnais, argues that CLDP has the ultimate in user-friendliness. On a trigger event the counterparty may choose either to have its contracts assigned to another dealer of its choice, or to cash-settle at “full replacement value”. The guarantor under the programme, CLFG Corp, is a wholly-owned subsidiary of triple-A rated Financial Security Assurance Holdings (FSA), which has provided a $200 million to $500 million 35-year policy to cover CLFG’s guarantee. Initially, deals were done solely with Crédit Lyonnais New York (CLNY), and collateral was placed with a trustee (Chemical Bank). Now the scope of the programme includes Europe, and it will be possible to do deals with Crédit Lyonnais in Paris.
Bankers Trust unveiled a similar programme this year – BT Credit Plus, with a $75 million initial pool of collateral posted with Bank of New York and a $25 million insurance policy from FSA. “The client has access to the bank’s liquidity and capital strength, in addition to the segregated collateral pool,” says Jeff Bardos, managing director at Bankers Trust, New York. At any trigger event, “clients can choose to terminate at their side of the market, or to have BT’s obligations assured by a well-rated dealer, based on prices provided by independent dealers,” says Bardos. “We’ve benefited from comments on other termination vehicles,” he adds.
Some counterparties judge the outside guarantee from an insurance company as the most solid DPC structure. “Those guaranteed by an outside insurance company are a better deal,” says one former derivatives risk manager.
Goldman Sachs
Goldman Sachs has three derivatives vehicles, all of which are continuation structures. The last to be created, in March 1994, Goldman Sachs Mitsui Marine Derivative Products LP (GSMMDP), is now its “primary triple-A entity”, says Andrew Devenport, executive director at Goldman Sachs International in London. GSMMDP differs from all previously mentioned vehicles, since one of its 50% joint-venture partners, Mitsui Marine & Fire Insurance Co Ltd, is triple-A rated and stands behind it “jointly and severally”; the other 50% partner is The Goldman Sachs Group LP. In this sense GSMMDP is more like the derivatives dealing subsidiaries of insurance groups AIG (AIG Financial Products) and General Re (General Re Financial Products). “Counterparties are very comfortable with GSMMDP because of the combined Goldman and Mitsui Marine credit,” says Devenport.
By this means, like other securities firms, A+-rated Goldman has been able to enhance its derivatives activities to deal with counterparties that can accept only triple-A or double-A counterparties. GSMMDP is triple-A rated by Moody’s and S&P. Moody’s notes: “The rating of the partnership depends solely upon the nature and extent of its support from Mitsui.” There is no special trigger or termination procedure: the two parents’ support agreement cannot be terminated as long as GSMMDP has any outstanding obligations.
Goldman took some time to achieve the perfect model. Its first triple-A vehicle, GS Financial Products International LP (GSFPI), was little more than a means of raising cash against a portfolio of puts, calls and warrants on the Nikkei 225 index, although it has since added further derivatives exposure. A second vehicle, GS Financial Products US (GSFPUS), is liable for the obligations of GSFPI, and is more flexible. GSFPUS with a capital of around $125 million bought a portfolio of in-the-money swaps from a Goldman affiliate and has since written new derivatives business. It is the only DPC that has issued publicly traded securities.
All three Goldman vehicles are limited partnerships, so their financials aren’t open to public scrutiny. It seems that GSFPI and GSFPUS were early attempts at creating an effective DPC which were later eclipsed by GSMMDP. The importance of Goldman’s bond underwriting business was demanding an improvement of its acceptability as a credit. “We sent two people off [to find a solution] and they went down different routes,” says Devenport. “GSFPI and GSFPUS were initially created for liquidity purposes.” GSMMDP has been “highly successful” as the preferred Goldman vehicle, says Devenport.
Comparisons are odious
Comparing triple-A DPCs can be highly misleading. Each was formed at a different time, for a slightly different purpose, and aimed at a different clientele. Volume of nominal principal outstanding doesn’t equal success, although it might hint at a vehicle’s acceptability. More important are the types of instruments held, whether the portfolio is balanced, how much capital it demands, and what additional business and reputation the enhanced credit standing has brought the firm.
Salomon’s Swapco was formed at a time when the firm’s reputation and credit rating were still reeling from its 1991 US treasury auction scandal. “It may have been the only way they could keep dealing,” says a former senior derivatives trader; “it may not have been a waste of money at the time.” Swapco has built up a portfolio of around $136 billion nominal, the fastest-growing of them all. Salomon’s triple-B rating means that more of its customer business, apart from that with customers prepared to use a collateral agreement as an alternative, would have migrated to Swapco than with other DPCs. Most swap houses agree that collateral arrangements, except in certain jurisdictions such as Germany and Japan, are cheaper than using their triple-A DPCs, so they steer their inter-dealer and interbank business in that direction. Westpac, for example, plans to run down the portfolio of Westpac Derivative Products as more counterparties accept its own improved rating.
Morgan Stanley says it has gone for quality not quantity in its DPC. “We had 80 target clients at launch [in January 1994] and we’ve signed up 60 of those,” says MSDP’s O’Connor. “We’ve limited the use of the subsidiary to those who previously wouldn’t deal with Morgan Stanley. We’ve been very strict with that: about 90% are triple-A or double-A. It’s more of a niche subsidiary for high-rated customers.”
Merrill’s higher fixed-capital cost – the initial $350 million – has driven it to find ways of working its subsidiary harder. In February it announced an agreement for DKB Financial Products’ (DKBFP) subsidiary of single-A rated Dai-Ichi Kangyo Bank, to use MLDP as a credit-enhancing intermediary. DKBFP is simply a large customer of MLCS, bringing swap counterparties to MLDP. DKBFP’s customer has MLDP as its counterparty. MLDP’s mirror trade with Merrill Lynch Capital Services (MLCS) is matched again with a back-to-back trade between MLCS and DKBFP. “We can’t even distinguish between a trade with MLCS and a trade with DKBFP,” says Jeremy Gluck, at Moody’s structured financing group.
Morgan Stanley says it has done a few similar deals already. O’Connor at MSDP comments, “It makes sense, especially for a smaller bank which doesn’t want to spend two years setting up a triple-A subsidiary.” But there are two problems with the arrangement. One is how to charge the customer. MLDP, for example, is understood to charge Merrill Lynch Capital Services (MLCS) a flat three basis points a year per deal, regardless of the deal’s riskiness. On a portfolio basis that may work out as fair, but with multiple clients a more risk-sensitive pricing mechanism would have to be developed. The second issue is confidentiality. DKB and Merrill have different customer bases, but their arrangement relies on them not poaching business from each other.
Dreams of a multilateral clearing mechanism are a little premature. In MLDP’s case “the vehicle isn’t isolated from the credit risk of Merrill Lynch”, says a ratings specialist. “If MLCS goes into bankruptcy then there are no more trades through MLDP.” But there have been serious discussions among Wall Street firms about multilateral credit enhancement. “At Salomon,” says Douglas Lucas, formerly with Moody’s structured financing group and now co-chief executive of Salomon’s Swapco, “we see in the future a truly independent entity owned by a group of sponsors separated from the credit risk of any one sponsor.” But it would have to be regulated by the CFTC (Commodity Futures Trading Commission), he warns. Today the strategy is for each house to “go it alone”, but “when everyone has a DPC or can get one”, Lucas adds, “it seems to me that would produce an incentive to combine to reduce costs and provide customers with a multilateral vehicle.”
The Chicago Mercantile Exchange (CME) is trying to get there first with its project to value and handle collateral for over-the-counter derivative contracts. This won’t take away the credit risk, but it should help end-users not comfortable with marking their positions to market and with handling collateral themselves, to deal with lesser credits as counterparties. The CME intends to go live with its project in September.
Dinosaurs?
Does this mean that triple-A DPCs are on the way out before they even got established? A paper* by three researchers at the Federal Reserve Bank of New York asks why, four years after they emerged, DPCs had taken such a small share of the overall derivatives market. Critics of the paper note that it erroneously compares the DPCs’ over-the-counter derivatives books with banks’ entire derivative positions, including FX forwards and exchange-traded derivatives.
More usefully, the paper notes that DPCs have brought an important innovation – the pre-specification of the workout process: “Structured DPCs let customers know from the outset which workout process will apply.” But it sympathizes with the view of some potential counterparties who “may not regard a structured DPC rating as the equal of other triple-A ratings. The DPC’s rating may be regarded differently because the workout risk of such a DPC corresponds more closely to the parent’s typically single-A rating.”
However, paradoxically, the capital standards applied to triple-A DPCs by the rating agencies appear far more stringent than those currently applied to banks under the 1988 Basle Accord. In 1994, the researchers calculate, MLDP would have needed a minimum of $40 million in tier 1 capital, only about a ninth of its actual capital; Swapco would have needed only a seventh. A calculation by another investment bank suggested that it could run a DPC on capital of as little as $50 million; but that would not have been enough for the rating agencies.
Part of the cost of establishing these early DPCs was to pre-empt criticism that they might be flimsy and under-capitalized. Now that the concept is understood, a major exercise is gradually to make their use of capital more efficient. Even Merrill Lynch is rumoured to be working on changing the maintenance of MLDP’s capital from static to dynamic. However, reducing the credit standards of these vehicles from triple-A to double-A – theoretically a good idea – would have saved Merrill Lynch, for example, only around $50 million of its $368 million capital at the end of 1994, the researchers calculate, according to Moody’s comparative expected loss rates for companies with AAA and AA ratings.
The NY Fed authors’ conclusion? New DPCs will continue to be formed, but they “will serve a market niche consisting of the relatively few, albeit important, customers who insist on triple-A ratings.”
The economic value of DPCs is not transparent to outsiders, even if the subsidiary publishes a balance sheet and profit-and-loss account. An inventory of additional business done with clients not previously accessible to the parent would be a start, but what firm will volunteer that information?
Euromoney Bondware’s data on bookrunners of triple-A and double-A issues over the last five years suggest that probably Merrill Lynch, and possibly Morgan Stanley and Goldman Sachs, have benefited in that business from the added clout of a DPC (see chart on page 124). But such evidence isn’t conclusive, and Eurobond rankings are not necessarily the main objective of a house with a triple-A DPC.
DPCs have been useful especially in jurisdictions where using collateral and netting of swap positions are still a problem, notably Germany and Japan.
DPCs have helped Salomon, Westpac, Paribas, Crédit Lyonnais, Lehman Brothers, Bankers Trust, stay alive in markets where institutions are increasingly competing to offer a one-stop risk management service, and where a high credit rating is a persuasive additional tool. But as the use of collateral and other means of credit mitigation grow, the DPC may be put out to grass – less a dinosaur, perhaps, rather a once trusty workhorse.
And there is the final conceptual question. Can the rating agencies detach themselves sufficiently from structures that are partly their own children? The agencies say they were careful, when the first DPCs were constructed, not to define goals, propose solutions, or prefer brands of risk management models.
But now that 16 structures have their approval, there are several templates for successors to follow. Bankers Trust found it was simpler to buy a risk management model from a third party for its BT Credit Plus programme than to harness or redevelop its in-house system. Why? Because a) the model had already been vetted by the rating agencies, and b) it would emphasize the independence of its verification process; it could not be accused of producing valuations favouring Bankers Trust.
Superficially this may be reassuring. But more detached observers may worry that the rating agencies, together with a few firms of accountants, are de facto supervisors of these structures, as well as their handmaidens and midwives. Is this a relationship with which regulators are comfortable? And will lesser players be tempted to buy them off the peg?
The structures, like many creations in the derivative markets, have hardly been tested in the heat of battle. Some dealers believe they are ridiculously over-secure. But it is interesting, and maybe reassuring, how experts at the rating agencies were biting their nails in the darkest days of the Salomon downgrading. “Sensitivity at the rating agencies gets higher as the rating of Salomon goes lower,” recalls a former S&P specialist vividly. “Each notch means, are we still sure? Let’s stress the portfolio even further. As it goes lower, our franchise is on the line.” However much the agencies may sympathize with DPC structures, says the former specialist, a far greater concern is their own franchise and their own reputation for independence.
The swap dealers with triple-A DPCs cannot use approval from the rating agencies to argue that they have made a good business decision. One UK merchant banker prides himself on having steered his board away from creating a me-too triple-A subsidiary: “For me it would have been putting the cart before the horse. We had to know first what we were intending to do in derivatives.”
The product-range of triple-A DPCs has necessarily been limited. Swaps and options on single equities, equity baskets, and commodities have so far been regarded as too volatile and too thin on price data to be admitted. That may change. The focus of many top dealers is moving to credit derivatives. Will DPCs evolve to accommodate the multiple slicing and swapping of credit exposures, or will they be put on the shelf?
*Risk Management by Structured Derivative Product Companies, by Eli Remolona, William Bassett, and In Sun Geoum, Federal Reserve Bank of New York Economic Policy Review, April 1996.
| Triple-A derivative product companies – how the structures compare | |||||||||||
| Went | Domicile | Initial | Nominal | No of | S&P | Moody’s | Other | Parent | Minimum | Regulator | |
| active | capital | outstanding | counter | rating | rating | rating | rating | c’party | |||
| (m) | parties | rating | |||||||||
| Bear Stearns Financial Products | Mar-96 | Delaware | $150 | AAA | NR | A/A2 | None | None | |||
| Bear Stearns Trading Risk Management | Mar-96 | Delaware | AAAt | NR | A/A2 | None | None | ||||
| BT Credit Plus Program | Jun-96 | New York | $100 | AAAt | Aaa | A/A1 | None | NY State/US Fed | |||
| Crédit Lyonnais Derivatives Program | May-95 | New York | $200 | $13.6bn | 104 | AAAt | Aaa | BBB+/A3 | None | C Bancaire | |
| GS Financial Products International LP | Mar-92 | Cayman Is. | c $75 | $3bn | AAA | Aaa | Fitch AAA | A+/A1 | None | ||
| GS Financial Products US LP | Nov-92 | Cayman Is. | $102 | $11bn | AAA | NR | Fitch AAA | A+/A1 | None | ||
| GS Mitsui Marine Derivative Products LP | Oct-93 | Delaware | $75 | $40bn | AAA | Aaa | AAA/Aaa | A-/A3 | MoF | ||
| Lehman Brothers Financial Products Inc | Nov-94 | Delaware | $200 | $62bn | 115 | AAA | Aaa | Fitch AAA | A/Baa1 | A-/A3 | None |
| Merrill Lynch Derivative Products | Nov-91 | Zurich* | $350 | $146.1bn | 282 | AAA | Aaa | Fitch AAA | A+/A1 | A-/A3 | None |
| Morgan Stanley Derivative Products | Jan-94 | Delaware | $150 | $51.5bn | 60 | AAAt | Aaa | A+/A1 | None | ||
| Nationsbanc Financial Products | n/a | Delaware | $300 | AAA** | Aaa** | A+/A1 | OCC | ||||
| Paribas Dérivés Garantis | Jan-94 | Paris | Ffr800 | $47bn | AAAt | Aaa | Fitch AAA | A | A | C Bancaire | |
| SBCM Derivative Products Ltd | Apr-95 | London | $300 | $8bn | 32 | AAA | Aaa | A/A1 | None | SFA | |
| Salomon Swapco Inc | Feb-93 | $175 | $136bn | 225 | AAAt | Aaa | Fitch AAA | BBB/Baa1 | BBB-/Baa3 | None | |
| Tokai Derivative Products Ltd | Feb-95 | London | £100 | AAA | Aaa | A- | AA-/Aa3 | SFA | |||
| Westpac Derivative Products Ltd | Mar-94 | Sydney | $200 | $3bn | 8 | AAAt | Aaa | A+/Aa3 | A-/A3 | AusSecCom | |
| * in Delaware, US, until April 1996 | **Expected | ||||||||||
