THE PROFITABILITY OF CARRY TRADES

Abstract

The carry trade phenomenon has been extensively covered in the media and it is widely held to be profitable. Hence, this thesis seeks to examine the profitability of carry trades. The carry trade strategy of borrowing low-interest rate currencies to reinvest in high yield currencies is implemented in this study. I demonstrated the importance of incorporating bid-offer spreads into the analysis of carry trade profits as the transaction costs can be quite considerable compared to the raw carry trade returns. My results indicate that carry trades are generally not profitable at the daily and weekly investment horizons when transaction costs are included. Positive returns can be earned at the monthly and quarterly investment horizons. However, these carry trade returns are relatively small when viewed in relation to the standard deviations of the returns. Trading portfolios based on interest differentials as well as an equally-weighted strategy are also initiated to exploit the carry trade phenomenon. The trading portfolios in general deliver a better performance compared to the carry trades of the individual currencies. This provides evidence that trading rules can be used in improving the carry trade profits. In addition, many of the monthly and quarterly carry trade returns are statistically significant under the conventional t-tests. However, the more robust bootstrap tests indicate that the monthly returns are not statistically significant. A random walk process is able to generate returns similar to that obtained by the monthly carry trades. Despite all the publicity within the media on the carry trade phenomenon, the carry trade profits are generally small and the monthly returns are not inconsistent with a random walk process. Thus, the popular media might have over-exaggerated the profitability of carry trades.