Kamakura Blog: Daily Trading Volumes in Credit Default Swaps

In a press release today, the Depository Trust and Clearing Corporation announced that it was supplementing its weekly disclosures on credit default swap (“CDS”) volumes with an analysis of volumes over the period from June 20, 2009 to March 19, 2010.  A copy of the press release is available here:

http://www.dtcc.com/news/press/releases/2010/credit_derivative_data.php

The explanation of the data is available here:

http://www.dtcc.com/products/derivserv/data/index.php

The data itself is available on this page:

http://www.dtcc.com/products/derivserv/data_table_snapshot.php

The documentation on the data elements is given here:

http://www.dtcc.com/downloads/products/derivserv/CDS_Snapshot_Analysis.pdf

What the data shows is that daily trading volume is very significantly less than a casual reader of the financial press and many market participants would expect.  The trading volumes reported by the DTCC include dealer to dealer trades.  In this blog entry,

van Deventer, Donald R. “Sovereign Credit Default Swaps and Lessons from Used Car Dealers,” Kamakura blog, www.kamakuraco.com, May 11, 2010. Redistributed on www.riskcenter.com on May 13, 2010.

we used DTCC data from the week of April 23, 2010 to find that 83.7% of all new CDS trades that week were dealer-to-dealer trades.  If one believes that non-dealer trades are the best indicator of true end user sentiment and interest, then the adjusted non-dealer volumes are much less than what was reported by the DTCC today.  The DTCC classifies the following 14 “families” of companies as dealers:

• Bank of America Merrill Lynch
• Barclays
• BNP Paribas
• Citibank
• Credit Suisse
• Deutsche Bank
• Goldman Sachs
• HSBC
• JPMorgan
• Morgan Stanley
• Nomura
• Royal Bank of Scotland
• UBS
• UniCredit

934 non-sovereign reference names are reported as most active by DTCC over the June 20, 2009 to March 19, 2010 period.  The distribution of average CDS trades per day for those 934 reference names shows there is next to no volume in the overwhelming majority of reference names:

The histogram below shows a number of interesting facts:

• Only 241 corporate reference names averaged more than 5 trades per day
• Only 63 reference names averaged more than 10 trades per day
• Only 14 reference names averaged more than 15 trades per day
• No reference names averaged more than 23 trades per day

When we look at the deepest bond markets in the world, those for sovereign debt, we see a similar picture.  The DTCC data shows that, for the 62 most active sovereigns and sub-sovereigns, there was surprisingly low volume:

As the histogram indicates,

• Only 26 sovereigns and sub-sovereigns averaged more than 5 trades per day
• Only 17 sovereigns and sub-sovereigns averaged more than 10 trades per day
• Only 6 sovereigns averaged more than 20 trades per day
• No sovereign averaged more than 42 trades per day

Implications of Low Volume in Sovereign and Corporate CDS Contracts

Especially when one takes into account that the overwhelming majority of trading volume is dealer to dealer, there are some important caveats of which potential participants in the market for corporate and sovereign CDS should be aware.  The messages are exactly the same as those explained so skillfully by Michael Lewis in his 2010 book The Big Short:

• The market is controlled by 14 dealers world-wide.
• Only five of the 14 dealers reported by the DTCC have US headquarters
• Therefore the market is ripe for oligopolistic behavior and collusion
• The market has a history of intentional non-disclosure, which is a very big red flag.  The disclosure by DTCC today is a step forward but one that the industry resisted for years
• As in The Big Short, a non-dealer with a position in an illiquid credit default swap is at the mercy of the dealers for a “mark to market.”  As Michael Lewis explains, even if the real market has moved in the non-dealer’s favor, the dealer has no pressure to show a true mark unless the dealer, overall, has the same net long or short position as the non-dealer.  When their stances are the opposite, the dealer will understate the non-dealer’s gains so as to minimize the dealer’s own reporting of its mark to market losses
• Financial journalists’ reporting of credit default swap levels typically ignores this lack of volume and ignores the ability of dealers to set CDS prices freely without the discipline of non-dealer traders picking off “off the market traders.”
• The potential for abuse is great:
  • Step 1: Take a short position in the currency of Country X
  • Step 2: Boost the quotes for the CDS on Country X dramatically over a series of days on no volume
  • Step 3:  Financial journalists report that Country X is in dire financial straits and the currency depreciates sharply
  • Step 4: Close out all trades and take profits

A market participant who ignores these dangers, especially after reading The Big Short, is at severe risk of loss.

Ranking of Sovereigns by Daily Average CDS Volume

The charts below rank sovereign and sub-sovereign reference names from most active to least active according to data from DTCC analyzed by Kamakura. Keep in mind that for more than 120 sovereigns on which the KRIS default probability service reports default risk, there were no CDS trades at all reported by DTCC.

Rankings of Corporates by Average Daily CDS Volume

The tables that follow report trading volume in order from most active to least active for the top 300 corporate reference names.  For an alphabetical listing, see the link to DTCC data given above.

Never has the expression “buyer beware” been more appropriate than in the credit default swap market, as the DTCC data released today shows.

Donald R. van Deventer
Kamakura Corporation
Honolulu, June 2, 2010