How Efficient are Credit Default Swap Markets? An Empirical Study of Capital Structure Arbitrage Based on Structural Pricing Models

Abstract

We investigate the efficiency of the credit default swap (CDS) market via the profitability and risk of capital structure arbitrage strategies based on observed mispricing in the CDS market over the years 2002 to 2006. We find that the CDS market has been inefficient in our observation period although this inefficiency declined over time. For this purpose, we calculate CDS premiums by means of the CreditGrades (2002) model and the models of Leland and Toft (1996) and Zhou (2001) and apply those within a capital structure arbitrage context for a dataset covering more than 800,000 observations. Our results indicate that structural credit risk models can adequately replicate market spreads but still produce significant positive arbitrage returns within our strategy. This also holds when transaction costs are incorporated. Accounting for risk via the Sharpe ratio reveals that the CDS market was inefficient at the beginning of our sample period but became efficient in the years 2004/2005 consistent with soaring trading volumes and the introduction of CDS index trading at this time. We are also able to show that the mathematically more advanced Leland and Toft and Zhou models provide larger arbitrage returns within our strategy. Considering rating classes, the arbitrage returns increase as obligors become riskier, in line with standard investment theory.