PETER NOLAN IS not convinced that buyers and sellers of credit default swaps are fully aware of their exposure. He works for Lysis Financial, dealing with project management in operations and technology for banks that deal in structured credit products. “Banks in the industry are continuously uncovering new exposures to General Motors through credit default swaps,” he warns. “There are risks that people aren’t cognizant of because of old documentation. Every bank is trying to improve its systems but at large organizations there are still legacy trades that are difficult to track. Documentation has altered with the growth in the market and legacy exposure can be hard to work out. How much risk is hiding in those dark corners?”
The question cannot yet be answered. No one predicted that the CDS market would grow as fast it has since its inception in the 1990s and that the infrastructure would not be able to cope with the ballooning volumes. Some market observers even suggest that some trades executed years ago are still unconfirmed.
According to the International Swaps and Derivatives Association, the notional amount of CDS outstanding at the end of 2003 was $3.58 trillion. In June 2005, the figure exceeded $12 trillion. By mid-2006 it is likely to have risen to more than $20 trillion. Global participants in the market are numbered in the thousands and rising, says Isda, with dealers, banks, investment managers, hedge funds, pension funds and insurers all active.
Having a large number of market participants when there is unprocessed documentation can create chaos if there is a credit event. Often the outstanding derivatives notional can be more than the bonds notional so an effective settlement process through exchange of money or bonds is needed to allow trades to settle neatly. In the case of October 2005 bankruptcy Delphi, the total credit derivatives market was about $28 billion, compared with $2.2 billion par of bonds. If there are a sufficient number of outstanding CDS that are not fully matched, there is an increased risk that there will not be an orderly settlement in such cases. “If the settlement were not orderly and effective in its risk mitigation, the risk of legal action to achieve satisfaction would increase and that would damage market growth,” says Mark Beeston, president of trade affirmation service provider T-Zero.
In the autumn of 2005, regulators were forced to recognize the risk that had arisen from the lack of infrastructure in place in the CDS market, and swiftly drew up plans to reduce it. Buyers and sellers have since made significant efforts to improve their systems, and technology providers have been rallying to their aid. How far have we got in establishing a suitable infrastructure?
As the corporate debt market took off in the mid-1990s, credit default swaps began to gain in popularity. Initially insurance contracts taken out on single credits, CDS were bought by banks as protection in the event that companies they were lending to went bankrupt, failed to make payments or restructured. It was an over-the-counter market, with banks entering into privately tailored agreements with customers. Buyers and sellers would affirm trades by phone, with final documentation faxed or emailed through for back offices to confirm that the two sides matched. Initially, when volumes were low, this method was effective. And even though a lack of standard documentation meant frequent discrepancies, these could be ironed out quickly, and confirmations could be processed shortly after a trade.
However, as the market developed, the benefits of using CDS became apparent to a growing number of financial sector participants. Business took off. CDS enabled banks to continue to extend credit to companies they were positive about by using derivatives to reduce the net exposure. The instruments also facilitated balance sheet management. By selling assets and retaining risk through CDS, a user could move assets off a balance sheet and reduce financing needs. CDS enabled banks to hedge assets and reduce capital usage. For investors, CDS enabled access to credit risk that might be difficult to obtain in the traditional bond and loan markets. And above all, the market enabled investors to use leverage. Given that a CDS is an insurance contract, no initial financing is required. So long as the buyer meets the premium payments, he can purchase a CDS.
Too much, too quickly
With an increasing number of counterparties and credits available, CDS volumes rose rapidly. And although Isda sought to standardize documentation to ease matching, and automated matching systems were introduced by such operations as DTCC’s Deriv/SERV and Swapswire in 2003, high volumes made it hard for back offices to deal with documentation and keep track of positions. Between 2003 and the end of 2005, volumes of outstanding CDS increased five-fold. “As an OTC market, it is all too easy to execute a trade, but, prior to Deriv/SERV there was no standard electronic infrastructure in place to facilitate deal matching and confirmation,” says John Burchenal, managing director of asset class expansion at trade management firm Omgeo. “It wasn’t uncommon to see paper contracts piled from the floor to above desk height in the back offices of CDS market participants,” says a senior executive at a US broker.
| A growing pile of paperwork Outstanding notional amount of CDSs |
| Source: Isda |
Exacerbating the problem was the issue of novation/assignments of CDS, where one participating member of the trade was replaced by another. When a dealer and a hedge fund traded a CDS, the two would, if signed up to the DTCC, send in their respective trade files to Deriv/SERV for confirmation. But if the two sides of the trade did not match up, the trade would be returned to the back office of the dealer and a hedge fund client to go through it and manually check it. That could take days, if not weeks. Meanwhile, the hedge fund manager might have thought that the trade was fine, and therefore might have decided to sell it, or to assign it to another party. That third party might then do the same, while the original dealer might also be assigning the trade to another party. As the original trade hadn’t been confirmed, the rest of the trades were unconfirmed. In the event of a credit default, some participants would be expecting delivery, only no trades would have been confirmed. With chains of CDS in motion, a lack of standard procedures, and paper confirmations, it is easy to understand how banks have now found themselves questioning what their positions are, and where their counterparty risk lies. “Paper started to proliferate, as did the risk,” says Omgeo’s Burchenal.
The Fed lays down the law
In September 2005, the Federal Reserve Bank of New York called in representatives of the 14 major dealers in credit default swaps. Timothy Geithner, the New York Fed’s president, urged those present to address the situation, and to reduce confirmation backlogs and improve the CDS settlement process. It got the attention of the dealers. “If the Fed bank calls you in and urges you to do something, you do it,” says Dean Barr, head of liquid alternative investments at Citigroup Alternative Investments.
The major target was to reduce confirmations outstanding more than 30 days by 30% by the end of January 2006. “Anything less than significant progress on our backlogs over and above our January 31 2006 goals will be unacceptable,” the dealers responded in a letter to Geithner on October 4. Indeed by January 31 the number of outstanding confirmations had been reduced by 54%. “I gather banks boosted back-office staff by 30% to go through outstanding transactions,” says the senior manager of a buy-side firm. “They basically conducted lock-ins with other dealers, where they would sit in a room with lawyers and the counterparty on the trades. Every trade between the two would be considered as one bloc, and the two sides would net down all the buys and sells between them to a final figure.” A back-office executive points to the sleight of hand involved. “It’s cheating really,” he says. “But the banks wanted the problem to go away. Should they fix the wall? Or paint over the cracks?”
Other participants in the market noticed an improvement. Samuel Cole, COO of Blue Mountain Capital Management, a $2.6 billion alternative asset manager in the corporate credit market, says outstanding confirmations involving his company have fallen by more than 90% since early August 2005.
The 54% reduction in outstanding confirmations reported is undoubtedly impressive, and has succeeded in quelling the growing fear of systemic risk in the CDS market. Most market participants would agree that the issue has been successfully dealt with. “Some of our investors breathed a sigh of relief and stated ‘problem solved’,” says one hedge fund manager.
| Who is using the CDS market? Sellers of credit protection predicted for 2006 |
| Buyers of credit protection predicted for 2006 |
| Source: British Bankers’ Association Credit Derivatives Report 2003/04 |
But some participants will admit that the problem is far from “solved”. Cole says: “In our view, dealers have done a terrific job of capturing the low-hanging fruit, principally dealer-to-dealer outstandings that have been processed through regular lock-ins.” Geithner, however, is asking for a total reduction in backlogs of 70% by the end of June 2006, and Cole says that getting from 54% to 70% will be the hard part. “Systemically, it is important that dealers begin to address the backlog with clients,” he says. “While there has been success in addressing the short-term backlog with limited investment by dealers in systems, this is a growing market, and that backlog could return in the absence of focused management attention and innovation.” This is a point that Geithner also emphasized. In addition to urging dealers to cut the backlog, he also suggested they implement operational tools that would lead to automated matching of buys and sells, thereby eliminating discrepancies and speeding up confirmation and settlement. The solution he had in mind was electronic confirmations and the use of the Depository Trust & Clearing Corporation’s matching capabilities. It is a start, some in the market say, but by no means the sole solution.
Sleepless nights
The main gripe among CDS users is that there are several providers of operational tools to aid confirmation and settlements but none appears to offer a complete solution. “CDS settlements is the only risk involved in my job that keeps me awake at night,” says a hedge fund manager. “There are several parties offering similar or complementary solutions, but it is hard to understand how they all fit together and which to use.”
A systems provider makes a similar point. “I think there are so many innovations occurring that some users are confused as to who is offering what and are getting their facts wrong,” he says. “Everyone knows CDS settlements are an issue, but only a few are prepared to get to grips with what is being offered as a solution. Many just presume the dealer community is handling it.”
The New York Fed’s Geithner is effectively pushing for the DTCC to be the central body for CDS confirmations and settlements. The 14 dealers agreed in the autumn of 2005 to use the DTCC’s confirmation matching system, Deriv/SERV, to confirm trades and “require all active clients to subscribe to industry-accepted electronic matching platforms, including at least DTCC” by mid-January 2006. As somewhat of an afterthought, the dealers add: “We reserve the right to support future platforms as well.” One hedge fund manager was adversely affected by the rigour of dealers’ commitment to this agreement last October; he told Euromoney that his firm, a top revenue generator, had been suspended by a dealer since it had not signed up to Deriv/SERV.
Deriv/SERV has advantages when used in the automated post-trade CDS environment. Buyers and sellers can send contract details into the DTCC, which will check that the two sides’ agreements match and will record the confirmation or return the trade. Ironing out discrepancies is seen as crucial to a speeding up of confirmation and settlement. “Our experience demonstrates that on 20% to 25% of trades there is some discrepancy in the understanding of the trade details between the client and the dealer,” says T-Zero’s Beeston, whose firm facilitates a reduction of these errors in a timely fashion.
Deriv/SERV started confirming CDS trades at the end of 2003. Before Geithner’s meeting, Deriv/SERV had taken on board all major dealers and about 90 buy-side customers. It now has more than 250 buy-side customers. Peter Axilrod, managing director for business development at the DTCC, expects this to rise to as many as 400 in the coming months. “According to major market participants, more than 60% of credit default swaps traded worldwide are now confirmed through DTCC,” he says.
The role of Deriv/SERV as a centre for confirmations has been endorsed by electronic trading platforms such as Tradeweb and MarketAxess, and interdealer brokers such as Icap, GFI and Tullett Prebon. Both Tradeweb and MarketAxess feed CDS trade information directly to the DTCC and communicate the state of confirmation to users. Icap, GFI and Tullett Prebon are piloting a multi-broker check-out service with the DTCC, where dealers can review, verify and affirm OTC derivative trade information, and automatically submit transactions to DTCC Deriv/SERV for confirmation within seconds.
Downbeat on Deriv/SERV
Some hedge fund managers are not so appreciative of the way many industry participants are leaning towards Deriv/SERV. First, to use the matching service CDS users have to sign up for it. Some managers do not want to go through the application process, says one dealer, as they presume that it will be time-consuming and involve the submission of a lot of information.
Indeed, it appears that although dealers are fully aware of what Deriv/SERV is, those on the buy side are often uninformed. One hedge fund manager says that he will not use Deriv/SERV, as the costs of signing up are too high. Yet the DTCC says that it is only dealers that are charged for the Deriv/SERV matching service; buy-side clients can sign up for free.
The success of Deriv/SERV in tackling a backlog will lie in its being able to convince as many users as possible to sign up so that the DTCC becomes a central store of CDS trade data. But one market participant has doubts about the ability of the DTCC to encourage non-US CDS buyers and sellers to sign up. “In Europe, the DTCC is less well penetrated than in the US as it is not understood in Europe what they are doing,” he says. “I get the feeling Europeans are using their own matching engines.” However, the DTCC’s Axilrod says that this perception is also incorrect as more than half of its volume comes from Europe. “There are more players in the US active on the buy side of CDS, but when it comes to dealers, half the volume of trades confirmed with us comes from Europe,” he says.
Another problem is that not all CDS products are eligible to be matched by Deriv/SERV. Since inception, the system has gone from confirming trades for single-name European and US to include eight index products and several Asia-Pacific corporate credits. Tranches, however, are still not eligible, although Axilrod says the DTCC is planning to include them.
Vendors such as T-Zero and Communicator, however, do offer solutions to some of the shortcomings of Deriv/SERV and are increasingly being used as complementary tools.
T-Zero was launched at the end of July 2005, and provides a credit derivatives affirmation, workflow and connectivity tool. Since its inception, the company has taken more than 40 clients onto its platform. T-Zero takes trade details and shares them between the involved counterparties, enabling them to agree all the details as soon as a trade is executed, ensuring that they are error-free. T-Zero supports trades and counterparties that are DTCC eligible as well as trades that fall outside the DTCC process. Once affirmed, the trade details can be passed down to the counterparties’ own risk systems, to the DTCC, to external fund administrators such as GlobeOp and to a range of other vendors. “The advantage is that by affirming all the economic details, T-Zero removes the operational risk on the trade date. This process facilitates a perfect match at DTCC if eligible or facilitates easier execution of a paper document with the risk already auditably mitigated,” explains Beeston.
Communicator provides tools to reduce confirmation backlogs through its OTC derivative post-trade processing platform. It offers a two-way trade-date tie-out service with full support for all derivative product classes – not solely CDS. It also offers an electronic link to the DTCC’s Deriv/SERV and the ability to track novations. As a result, Communicator states that clients have experienced a 75% reduction in confirmation turnaround and a 50% reduction in backlogs.
Worries about a monopoly
Having just one entity to provide trade confirmation and settlements is not to everyone’s liking. “The DTCC has a monopoly position really, which is worrying,” says a hedge fund manager. “And now it will be difficult for other parties to infiltrate the market.” He suggests that vendors such as T-Zero and Communicator could be more nimble in building systems to manage the CDS confirmations and settlements process but that given the backing of the New York Fed for Deriv/SERV, any other parties would be afraid of investing the money to compete. T-Zero and Communicator, however, say they are not competitors of Deriv/SERV but symbiotic with the DTCC.
Before Deriv/SERV’s soft endorsement by the New York Fed, it had a serious competitor in Swapswire. A senior executive at a hedge fund says: “Swapswire was a dominant force in the CDS matching and processing space alongside Deriv/SERV but it has fallen off the radar a little and now is seen as a leader in the interest rate swaps processing space instead.”
Launched at the same time as Deriv/SERV, Swapswire was highly regarded for its CDS matching capabilities and still is. Unlike Deriv/SERV, which matches trades after the trade details have made it through various systems at each dealer, Swapswire has trade capture and starts the confirmation process as soon as the trade is executed. Although Deriv/SERV has successfully reduced settlement time from T+20 (trade day plus 20 days) to T+5, 99% of CDS trades through Swapswire are confirmed on T+0. Chip Carver, CEO of Swapswire, says that it suffered a decline in CDS volumes after the Fed meeting in September but interest has recently increased. The firm is also offering backloading, where users can use Swapswire to confirm legacy trades.
Whether there is room for both Deriv/SERV and Swapswire, or indeed other competitors in the CDS market, is being debated by some in the industry. “It will be interesting to see the landscape in one year’s time,” says a prime broker. “Will there be only Deriv/SERV in the CDS confirmations space? Will Deriv/SERV have ironed out its flaws, and will the complementary vendors be redundant? Will there be some consolidation?” Andy Brindle, senior adviser at MarketAxess, which launched a fully electronic CDS trading platform in September 2005 with a direct feed into Deri/SERV, says: “There may be room for more entrepreneurial enterprises to spring up but we are confident that DTCC will continue to play a critical role in the confirmation and matching process.” Swapswire’s Carver responds: “Basically success will depend on how big a network you can create and how many people you can keep on it.”
Although the number of different players enabling trade confirmation of CDS players in the industry and their relationships might be confusing for users, it does point to a commitment to clear up the backlog and provide a sturdy infrastructure. “The risk is still out there but the market is in a better position now to handle a huge credit event such as a GM default,” says one dealer. But there is still ground to cover, and solutions should be fast coming. Blue Mountain’s Cole says: “The market needs a scaleable approach to preventing this in the future.”
