Are credit derivatives markets particularly prone to speculation and contagion? The answer is certainly yes if we take a look at the credit default swaps (CDSs), which are the most widely traded credit derivatives.
Hedging or speculation?
As most derivative instruments, such as forward markets or options, CDSs were originally designed to hedge out investors’ risks. CDSs are aimed at insuring against the risk of borrowers’ default. Suppose that an investor holds a bond on a given borrower (called X). He can buy a CDS on X to get his money back in case the borrower defaults. In the CDS contract on X, the buyer (A) agrees to pay a certain amount of money, called “premium”, to the seller (B) over a given period of time. The seller B receives the premium and is committed to give the buyer a pay-off only in case X defaults on his debt. This pay-off offsets investor A’s loss as it is equal to the difference between the face value of the bond and its depreciated market value.
In practice, as most other derivative products, the huge development of the market is not to be imputed to hedging purposes, but to arbitrage and speculation. Suppose that an investor holds neither bonds nor debts on company X, but thinks that this company may default. He can buy a CDS on X and pockets the pay-off in case of default. This strategy is equivalent, but more straightforward than short-selling bonds. The CDS market is also reported to have recently been used to bet on companies’ default after the ban on short-selling in financial stocks.
The size of the market has soared well above the value of the underlying debt that they are supposed to insure (reaching about USD 62 trillion at the end of 2007). This has become clear since 2005, when Delphi, the auto parts maker, went bankrupt: the CDS on Delphi’s debt in the market exceeded the value of its bonds tenfold.
The rise in correlations
All major financial actors are deeply involved in the market. From the beginning, in the mid-nineties, banks have used CDSs to escape from their capital requirements. But nowadays, banks, hedge funds, insurance companies and pension funds are hugely exposed as buyers or sellers, or both. By transferring the risk, the CDSs have acted as a kind of insurance and provided incentives for risk-taking. They are therefore at the heart of the present crisis.
Since the start of the crisis last summer, the CDS market has been especially affected. Premia have been driven upward, and reached all-time peaks, especially in the high-yield segment. As the whole market has been hit, not only distressed firms, one can assume that there have been contagion effects. To check this, contagion can be defined as a simultaneous move in asset prices which results in a rise in correlations (Kaminsky and Reinhart (2000), Forbes and Rigobon (2002)). In fact, the CDS market has exhibited a stunning rise in its correlations since the start of the crisis in August 2007 (see Graph below).
Figure 1. Average correlations between CDS premia
Note: Europe : 66 European CDS (in the iTraxx Main since 2004), US : 73 North American CDS (in the CDX NA IG since 2004). Correlations are calculated by EWMA and averaged accross all pairs of CDS
Fears on counterparty risk
The CDS market is now in the eye of the storm. The reason is straightforward, because this crisis is about credit risk. A credit bubble has ballooned for years, being enhanced by the existence of credit derivatives. As credit originators can pass their risk to other agents, they have been less careful about the quality of their loans. In that sense, CDS have given an incentive for distributing more credit to more risky borrowers.
As banks and all financial institutions have massively committed themselves in the CDS market, they are now highly dependant on market continuity and its smooth functioning. The failure of a major participant, or worse a whole set of participants, can put at stake all the others. Bankruptcies of Bear Sterns, then that of AIG, two key counterparties, could have brought about a complete meltdown of the market. This has certainly been taken into account when the US government decided to step in and prevent them from going bankrupt. However, faith in the reliability of the market has been deeply shaken by these events. Lehman Brothers’ bankruptcy has also severely hit the whole market, as it was among the ten main participants.
As long as the financial system was sound, there was no fear that counterparty risk could be a problem. Today, after the near-failure of three key participants, the counterparty risk is the major worry and raises widespread fears of a market collapse. The main drawback stems from the very nature of this market, which is over-the-counter (OTC). Each buyer negotiates with a seller directly, without global clearing. The buyer thinks he escapes from the default risk of company X, but is still exposed to the counterparty risk of the CDS seller. If no netting is done, when trying to exit from a contract, the buyer has to sell another contract, which theoretically offsets its default risk but does not cancel the counterparty risk. In this framework, counterparty risks are multiplied by the lack of global netting. On-going projects to switch to an organised market with a global clearing-house seek to tackle this issue, as it would drastically reduce the counterparty risk. However, they are not likely to solve the problem in the short run, as the bulk of contracts have been made at a 5-year maturity on an OTC basis.
Another cause for concern is that the market is unregulated. CDSs act as insurance against default, but they are not submitted to any regulations as is the case for insurance companies. The latter have to meet required reserves and are closely monitored by public authorities. On the CDS market, no reserves are required from the sellers of protection, only very thin margins, ranging from 2% to 5% of the amount insured. However, the danger is even greater than insuring against natural catastrophes for example, because of the high correlation of default risk, which is linked to the business cycle.
The first warning: the May 2005 crisis of GM and Ford
Looking back on its short historical evolution, this is not the first time the CDS market has been routed. There was already a big meltdown of the CDS market in May 2005. At that time, General Motors (GM) and Ford were downgraded by the rating agencies from investment grade to “junk” grade, which triggered a violent crisis.
The 2005 crisis was a premonitory event. The GM and Ford downgrades had a large impact on the market due to the huge size of the two leading multinational firms. At that time, the CDS premia of both firms posted a sharp rise and the whole of the CDS market was affected, as well as the bond market. Acharya et al. (2007) have highlighted the liquidity shock that this crisis brought about on the market.
Coudert and Gex (2008) investigate the contagion effects of the GM and Ford crisis within the market. To do so, they calculate the conditional correlations between each of 226 CDSs on major US and European firms that are included into the main indices. Their results show that the correlations significantly increased during the crisis, especially in the first week. Both the US and the European markets were affected, pointing to the strong international integration of the credit markets.
They also analyze the links with the other financial markets. Theoretically, as a CDS is aimed at protecting investors against a borrower’s default, its premium should be close to the borrower’s bond spread, for a given maturity, even if in practice, they differ slightly. Usually, the CDS market is considered to lead the bond market, in the sense that price innovations go from the CDS market to the bond price (Blanco et al., 2004). This relationship between the two markets was somewhat mitigated during the 2005 crisis. At that time, CDS spreads tended to increase more than bond spreads, as investors bid up the price of protection; this points to the speculative nature of the market.
Acharya V. V., S. Schaefer and Y. Zhang, 2007. “Liquidity risk and correlation risk: A clinical study of the General Motors and Ford Downgrade of May 2005”, Working Paper, SSRN.
Blanco R., S. Brennan and I. W. Marsh. “An empirical analysis of the dynamic relationship between investment grade bonds and credit default swaps”, Research Paper, Cass Business School, 2004.
Coudert, Virginie and Mathieu Gex (2008), Contagion in the Credit Default Swap Market: the case of the GM and Ford Crisis in 2005, CEPII Working paper n°2008-14.
Forbes K. and R. Rigobon. “No contagion, only interdependence: Measuring stock market co-movements”, Journal of Finance 57 (2002), no. 5, p. 2223-2261.
Kaminsky G. L. and C. M. Reinhart. “On crises, contagion, and confusion”, Journal of International Economics, Elsevier, vol. 51(1), p. 145-168, June, 2000.