An amendment to the International Swaps and Derivatives Association's collateral dispute resolution protocol, delivered to the Federal Reserve Bank of New York on September 30, answers many of the concerns voiced by buy-side firms and represents a workable compromise, say market participants.
The protocol was originally drafted on July 15 and is part of an industry effort to reduce risk and standardise the over-the-counter derivatives market following a raft of collateral disputes in the wake of Lehman Brothers' bankruptcy in September last year.
The original draft prescribed a market-polling mechanism for resolving complex disputes (Risk August 2009, pages 40-42). However, Isda amended the method originally presented for evaluating quotes between one market-making party and one rebutting party because of concern among buy-side investors.
The new method states that if the pool contains at least two quotes on a two-way, firm, executable basis, then quotes on any other basis are discarded and the remaining bids and offers are averaged, and the mid-point between the average bid and the average offer is deemed to be the result.
Otherwise, the average mid-market value of any indicative reference quotes is calculated - known as the indicative average. The average of the two mid-market values stated by each party and the indicative average (if any) is then calculated. This value is the mid-market average.
The bid and offer levels of the market-maker quote obtained during the quote gathering period are then compared with the mid-market average. If the mid-market average lies between the bid and the offer (inclusive), then the mid-market average is the result. If it is higher than the offer, then the offer is the result. If it is lower than the bid, then the bid is the result.
"There are two key principles embedded in the process," explains Michael Clarke, co-chair of the Isda collateral committee and global head of counterparty risk operations at UBS. "The first is that when no independent third-party market prices are available, an equal weight is given to the dealer view and the counterparty view of the mid-market. The second is that, when a dealer is willing to commit capital and risk appetite by making a two-way, firm, executable price, the result should lie inside the market the dealer is making. This gives parity to end-user and dealer mid-market values, provided the average is within the dealer spread."
The previous process required that all quotes submitted on a two-way, firm, executable basis be pooled together, while other quotes are discarded. If the pool contained at least two remaining quotes, the bids and offers were averaged and the mid-point would be deemed the mid-market value for the trade. If there was only a single market-maker quote and no reference quotes, the mid-point between that bid and offer was used.
This posed problems for buy-side firms: in a market that is inactive, with no trades being executed, how far can a single quote from the market-maker be considered valid? If no trades are being executed, the argument would follow that the market was not functional and hence the quote unreliable.
The collateral committee included some safeguards in the protocol to allow the rebutting party to challenge the market-marker quote if it felt it was not justifiable, such as requiring the market-making party to certify in writing that the mid-point of its quote accurately reflects the level at which it is marking the transaction for its own books and records. But the committee acknowledged this doesn't necessarily solve the problem.
The new methodology responds to many of these concerns, say market participants.
"From the perspective of buy-side clients, we believe the ultimate fallbacks in the dispute resolution represent a workable compromise," says Lauren Teigland-Hunt, managing partner at New York-based law firm Teigland-Hunt. "It is a solution the parties can implement themselves and resolve disputes in a timely manner, which was one of the major overriding considerations. Furthermore, the process always yields a result, whereas earlier iterations included the possibility that an intractable dispute may be declared."
A number of alternative mechanisms were considered, including the use of indicative quotes to moderate the market-maker quote, mediation and arbitration, although none was deemed suitable.
The collateral committee revisited arbitration and mediation after the period for public comment, but decided that arbitration, although a good idea in theory, would be difficult to put into practice. A panel of arbitrators would have to be selected, and deciding who is qualified to serve on a panel - particularly whether they should be dealers, buy-side participants or academics - proved troublesome. The committee was also concerned that arbitration might be too slow - collateral disputes must be resolved as quickly as possible - and that if a determination were required for a highly complex instrument, the panel members would effectively be arbitrating between two models to determine which is more valid.
"Although arbitration has certain attractive qualities to it, such as objective determination, there was a recognition among working group members that a formal arbitration solution was not ultimately going to be practical or feasible," says Teigland-Hunt.
The protocol is now in an experimental pilot phase, during which a small number of volunteer firms will apply the procedure in parallel with actual disputes (and possibly to hypothetical test cases) in order to detect and correct any obvious deficiencies with the procedure. The pilot phase will run until December 15. An extended trial phase will then commence on January 15, conducted on a mandatory basis by 15 major dealers, with voluntary participation by other firms. Market adoption is expected to take place from July 15, 2010, onwards.