A recent blog on Reuters named three prominent American corporations in the post “Links to CDS measures growing common in bank credit deals.” This blog argues that borrowers should never agree to tie borrowing rates to credit default swaps in a loan agreement. It’s the 21st century equivalent of telling your banker “I’ll pay whatever rate you think is fair.” This blog explains why.
In theory, why is it a bad thing to tie interest rates in a loan agreement to a market index or price traded in a liquid and competitive market that correctly reflects the credit risk of the borrower? It is not a bad thing to do so. A more difficult question that we have addressed in seven recent blogs is this question: “Is the market for credit default swaps sufficiently liquid and competitive to correctly reflect the credit risk of a borrower?” Sadly, the answer is no.
Lack of Liquidity in the Credit Default Swap Market
We used data from the Depository Trust & Clearing Corporation to analyze the number of trades in the 77 weeks ended December 30, 2011 in these recent blogs:
van Deventer, Donald R. “CDS Trading Volume for 1,090 Reference Names,” Kamakura blog, www.kamakuraco.com, January 9, 2012.
van Deventer, Donald R. “Credit Default Swaps and Deposit Insurance,” Kamakura blog, www.kamakuraco.com, January 10, 2012.
van Deventer, Donald R. “Municipal Credit Default Swap Trading Volume,” Kamakura blog, www.kamakuraco.com, January 11, 2012.
van Deventer, Donald R. “Sovereign Credit Default Swap Trading Volume,” Kamakura blog, www.kamakuraco.com, January 12, 2012.
van Deventer, Donald R. “International Bank Credit Default Swap Trading Volume,” Kamakura blog, www.kamakuraco.com, January 18, 2012.
During this 77 week period, there were no trades at all in the credit default swap market for 29,910 of the 31,000 public firms whose default probabilities were available on the Kamakura Risk Information Services web site. In the blogs above, we focused on non-dealer trades (trades where at least one of the two parties to the trade was not a credit default swap dealer) because dealer-dealer volume can be easily manipulated at near zero cost. For the 1,090 reference names for which there was at least one credit default swap trade during this 77 week period, we reached the following conclusions:
- 76.2% of the reference names had 1 or fewer non-dealer contracts traded per day.
- 93.6% of reference names had 2 or fewer non-dealer contracts traded per day.
- 99.0% of reference names had 4 or fewer non-dealer contracts traded per day.
- Only 5 reference names averaged more than 5 non-dealer contracts traded per day.
- Only 1 reference name averaged more than 10 non-dealer contracts traded per day.
The chart below reports the daily average non-dealer trades per day for the 77 week period for the median corporation in the 1,090 reference names for which DTCC reported any trading volume:
The graph makes clear that, for the median firm on which there were any CDS traded during the 77 week period, there was an astonishing lack of liquidity:
- In only one of the 77 weeks was there a daily average number of non-dealer trades greater than 3 trades.
- There were only 4 weeks of the 77 in which there were more than 1.5 non-dealer trades per day.
- The average number of non-dealer trades per day over the full 77 week period was only 0.5 trades per day.
Lack of Competition in the Credit Default Swap Market
The lack of liquidity in the CDS market would be very troublesome if, in addition, there was a lack of competition in the credit default swap market. This blog shows that, on average, over the 77 weeks ended December 30, 2011 trades among dealers represented 81.68% of the live trades in the DTCC trade warehouse:
van Deventer, Donald R. “Collusion and CDS Dealer Volume,” Kamakura blog, www.kamakuraco.com, January 4, 2012.
The following blog shows that, among U.S. commercial banks reporting credit derivatives notional principal outstanding to the Office of the Comptroller of the Currency, the credit derivatives market is highly concentrated using the standard HHI calculation of the Department of Justice:
van Deventer, Donald R. “The Credit Default Swap Market and Anti-Trust Considerations,” Kamakura blog, www.kamakuraco.com, January 19, 2012.
In that same blog, just three banks (JPMorgan Chase, Bank of America and Citibank) had a 91.1% share of the total credit derivatives outstanding among the banks reporting to the OCC. A complete accounting of market share among all bank and non-bank dealers world-wide is not publically available, although the DTCC trade warehouse contains such data.
Risk of Collusion in the CDS Market
Is there a risk that credit default swap dealers might collude to increase the costs of borrowings tied to a given firm’s credit default swaps? There are a number of relevant facts from other financial markets that indicate this risk is not zero:
- JPMorgan Chase, Bank of America and Citibank have all been named in a lawsuit filed by Charles Schwab alleging the manipulation of U.S. dollar Libor.
- JPMorgan Chase, Bank of America and Citibank were labeled by Reuters blogger Cate Long as “serial offenders” for bid rigging in the municipal bond market.
The consequences of a lack of liquidity, lack of competition, and potential for collusion are very serious, as the following example shows.
Consequences of Tying Borrowing Costs to Credit Default Swaps
On June 16, 2010, during the Gulf Oil Spill, Bloomberg reported that the “market” for credit default swaps was indicating that BP plc was a 39% probability to default:
At the same time, the Kamakura Risk Information Services probability of default for BP plc was an annualized 0.06% for 5 years and a cumulative default risk of 0.32% over the same period. The KRIS calculation correctly reflected a market capitalization for BP plc of more than $120 billion and worst case damage estimates of about $20 billion. The 630.6 basis point CDS spread quoted by Bloomberg was more than 100 times greater than an objective assessment of default risk using a data base of more than 1.76 million observations and 2,046 defaults. The current annualized KRIS default probability for BP plc for a 5 year horizon is now 3 basis points, further confirmation of how drastically the spreads in the CDS “market” over-reacted.
How many non-dealer trades per day were reflected in that Bloomberg quotation for BP plc? Such volume should always be disclosed by journalists now that DTCC trade volume is available. Such reporting began the week of July 16, 2010. During the 77 weeks ended December 30, 2011, the average number of non-dealer trades per day in BP plc was less than 2 trades per day. In no single week during that period did the average non-dealer trades per day exceed 7 trades per day.
Conclusions
Loan agreements which include pricing based on the borrower’s credit default swap quotation are extremely dangerous to the borrower. The credit default swap market for single name corporate credits shows next to no liquidity. Almost 82 percent of recorded trades are among the dealers themselves. The three biggest American banks in the CDS market have more than 91% of the market, and the same banks have been named in numerous lawsuits alleging collusion in the Libor and municipal bond markets. Finally, the case of BP plc shows the risk of loans tied to a CDS quotation in a market that over a long period of time averaged less than 1.5 non-dealer trades per day.
Corporate treasurers should resist all such pricing proposals and eliminate them if they are in current agreements to protect the shareholders of the borrower.
Post Script: After expressing these concerns to the chief executive officer of one of the prominent firms named in the Reuters story, I was very pleased to hear that the firm had already eliminated the CDS pricing provisions from its loan agreement for similar reasons.
Donald R. van Deventer
Kamakura Corporation
Honolulu, January 24, 2012