Credit default swaps have allowed banks and investors to improve the management of their credit risk, but they may represent a lurking source of contagion in a crisis.
When credit default swaps (CDSs) were introduced in the early 1990s, they offered banks a valuable new tool for managing credit risk. In an era when banks tended to originate and hold credit risk, portfolios largely reflected the lender's regional footprints, and industry and credit-assessment expertise developed to deal with the characteristics of locally active firms. While a modest market in whole loans did exist, and syndication of large loans was a well-established practice, banks still found it hard to avoid credit portfolio concentrations driven by their particular market circumstances.
One incentive for banks to merge and broaden their market reach was to improve diversification of their portfolios. I have argued elsewhere1 that, as end-users, smaller regional banks gained more from the emergence of liquid credit risk transfer instruments than global giants. The anonymity of CDS contracts allowed regional banks to continue serving established clients, where their industry expertise and experience gave them a competitive edge, while limiting excessive portfolio concentrations by buying credit protection in the CDS market.
Of course, the above discussion relates to the impact of the CDS market on banks as end-users of these instruments. The impact of CDS contracts on banks as market-makers is a separate discussion. Market-makers stand ready to buy or sell an instrument subject to a bid-offer spread. The realities of day-to-day deal flows require them to hold temporary open positions. Nevertheless, market-makers seek to maintain a broadly balanced book in the long-run by adjusting their bid-offer spreads, to encourage the type of trades that reduce such exposures and discourage the ones that increase them. Of course, if exposures reach unacceptable levels, they may be forced to pay away the bid-offer spread to another professional market-maker to bring their positions closer to balance.
A well-known aspect of the financial crisis that struck in earnest in September 2008 was that one major player - namely, AIG Financial Products - was not behaving in the manner just described. Rather than running a more or less balanced book, it was simply writing unhedged insurance on what became almost $500 billion of largely unexamined credit exposure. It seems unlikely any player in today's CDS market is following such a foolhardy strategy. Nevertheless, the CDS market may represent a lurking source of systemic risk in the face of the snowballing eurozone crisis.
It has been interesting to see the term 'counterparty risk' tripping off the tongues of politicians and pundits of all kinds. At least it indicates a useful broadening of interest in this important subject. However, the CDS market presents a particularly dangerous form of this type of risk. In part, this is because CDS exposure represents a form of systemic wrong-way risk: CDS contracts are triggered by defaults that inevitably rise during a severe business downturn, which simultaneously strains the financial stability of major market-makers.
A second source of concern is that the sport exposures of CDS trades balloon dramatically when companies once thought to be sound begin to experience credit down-grades. The usual method of estimating potential future exposure involves simulating values at a level of confidence somewhere between 95% and 99%. Since the default likelihood of high-grade names is typically below 1%, these methods tend to yield CDS exposure estimates that are a very small proportion of the full face amount of protection sold. When an underlying reference name is suddenly downgraded, this exposure can balloon to a large multiple of the previously estimated amount.
In this situation, stress testing is an essential component of effective risk management. Fortunately, the CDS book lends itself to a simple calculation in this respect. Banks and supervisors should be examining the implications of a default by Greece and all other potentially suspect eurozone countries on the configuration of exposure of all CDS market-makers. Assuming there are no foolhardy players such as AIG that are just writing unhedged insurance policies, such a stress test will yield only a modest increase in net exposure. What is far more interesting is how such a default would alter the gross long and short positions that would underlie that net exposure.
The most likely way fallout from a European sovereign default could damage US banks would be if a significant portion of their purchased CDS hedges were written by a European bank that also had direct exposure to the defaulting country. Individual bank management should have a thorough grasp of how one or more European sovereign defaults would scramble the composition of the gross long and short positions in their CDS books. Supervisors can potentially go further, by sharing information on which banks would be worst affected by the initial default and so present a secondary threat to other market-makers to which they have sold protection.
1 See Rowe D, Competitive Implications of Holistic Balance-sheet Management, The RMA Journal, July-Aug 2008, pages 76-77.