Predicting currency prices and informational efficiency: Equity markets vs. credit default swaps

Abstract

This study examines the predictive power of the Credit Default Swaps (CDS) and the equity markets on the currency exchange rate to determine whether the CDS is a better predictor as compared to the equity markets. Data sets used for the study include the Investment Grade (IG) and High Yield (HY) North American CDS indices, and the iTraxx Europe index as a representative of the overall credit market conditions in Europe. The Vanguard Total Bond Market Index is included to see if the CDS spread is a more powerful information container than the bond market. The S&P500 index is used as controller for the effects of the US equity market and the Vanguard European Stock Index for Europe. ASX200 and NZ50 are chosen to represent the equity market conditions in Australia and New Zealand respectively. The Vector Autoregressive (VAR) model is used to analyse the simultaneous relationships between the exchange rates and the CDS index spreads. The Granger causality test is conducted to determine the causal relationship between currency values and CDS spreads. Variance Decomposition or Forecast error variance decomposition is also used to complement the VAR analysis. The VAR analysis considers that CDS can better capture the information in the market than other investment instruments such as bonds. CDS thus may offer arbitrage opportunities for investors. In addition, significant Grange-causality effects were found from IG and HY CDS spreads to currencies, which support the CDS spreads as a leading indicator of the several currencies versus US Dollar, even in the financial crisis. The results of variance decomposition indicate that the contribution of the CDS market to the currency market is higher in the case of the Australian dollar, implying more carry-trades in the market.