Dispute looms over clearing
Natasha de Terán
08 Aug 2008
The fact clearing is on the way for credit derivatives seems unavoidable. Yet in all the hours of discussion that have led us to this point, one of the fundamental considerations has received little attention—pricing.
Where will the exchanges and clearing houses that are to provide these services obtain their reference pricing data? And will these sources be sufficiently reliable? Therein lies the rub.
Credit derivatives reference over-the-counter instruments. Unlike equity-linked or interest rate contracts, which reference liquid, observable, transparent underlying markets, the corporate bond and the credit derivatives markets can be illiquid and opaque.
Although transparency has increased, there is little observable transactional pricing and there is no centralized marketplace from which to gather accurate, timely data.
By clearing credit derivatives, the exchanges and clearing houses will, in effect, be institutionalizing whichever pricing sources they use.
If the source is good, no problem. But if it has any shortfalls, inconsistencies or biases, those will be embedded into the clearing house’s risk management process and could present a problem.
Markit is the largest credit derivatives pricing provider. It is tried and tested and, we must presume, reliable. But is it reliable enough for clearing houses to depend solely on its prices, irrespective of any differences that these may have to other pricing providers?
Some argue that Markit’s pricing is sufficiently clean, accurate and independent to be used in isolation. Others argue CMA, a rival provider, is superior. The differences are not inconsiderable.
Markit derives its reference pricing from contributed data extracted from dealers’ official books and records. CMA’s DataVision is based on observed, real-time indicative prices given by the front office desks of buy-side firms.
The CMA camp claims Markit’s data is clouded because, if and when dealers mismark their books, these can “pollute” its calculations. Markit would refute this on the grounds that it “cleans” the received data, but recent mispricing episodes at two of Markit’s shareholders and contributing banks have highlighted the risk.
In the past few months, both Morgan Stanley and Toronto-Dominion Bank’s investment arm TD Securities have each discovered mismarks on credit traders’ books, resulting in Morgan Stanley having to write off $120m and TD $94m.
There is another difference. Markit is dealer-backed and governed, although some large and powerful buy-side firms have stakes in it. CMA has a following from the buy-side and is owned by the Chicago Mercantile Exchange.
Some of the buy-siders using CMA say its data is more accurate, that it does not contain the dealer biases that appear within Markit and, critically, that it is more useful since it is based on quotes rather on marks. The Markit camp would argue theirs is the more accurate, that it is more consistent and reliable and that it contains no bias.
Most sellsiders appear happy with Markit and so do the majority of the purported credit clearers. All the evidence suggests they will be getting their pricing from Markit – and Markit only.
Markit is a stakeholder in the Clearing Corporation, whose subsidiary, the Clearing Corporation Trust Company, is to launch credit derivatives clearing. Liffe, which is to launch credit clearing through its Bclear OTC platform, is understood to be agreeing a deal to use Markit. Eurex already references Markit data for its other products.
The CMA’s pricing will be used by CME, but the data provider has yet to issue a licence to another exchange or clearing house.
But if there is the smallest shred of uncertainty surrounding credit derivatives’ pricing, it would seem sensible that any nascent clearers should give full consideration to the alternative pricing providers, their differences and whether they should be using them.
And pricing is not the only rift emerging between the Markit and CMA camps. A fissure is opening on another front—identifiers.
In 2002, three dealers formed a consortium under the name Project Red.
Together with PricewaterhouseCoopers and Allen & Overy, they designed the Reference Entity Database—a catalogue of reference entities and obligations traded in the credit derivatives market with identifiers, that is akin to the ISIN system used in the cash markets. Soon after, the Red database was transferred to Markit, which now owns and runs it.
This system ensures that when CDS trades are struck, both parties can ascertain that the entity in which they think they are buying or selling protection is the same one that their trade counterpart is referencing, and which of its obligations that protection rests against. Without it, the scope for errors and legal risk within the CDS market would be substantial.
But Red’s services are not free and not all of its users are entirely happy with the costs.
Markit’s shareholders are neutral on the issue. They are large users and have a stake in Markit’s fortunes. But some who are not Markit shareholders believe the service, if not free, should be cheaper – ideally, run at cost.
Such is the extent of dissatisfaction, that the door appears to have opened sufficiently for another provider to step in. That seems to be what CMA plans to do. Project Blue, as its purported service is being dubbed, is understood to be at the conceptual stage but if it goes ahead, the credit markets will soon have two different identifiers.
This raises several questions: is it desirable to have competing offerings in such vital pieces of market infrastructure? Should such providers run on an at-cost or a for-profit basis? Will the emergence of a rival provider confuse matters and result in increased costs for those operating both systems?
And finally, has Markit got too powerful for its own good? The answers to some of these questions will become apparent if Project Blue gets past the planning stages. |
