Macroeconomic Momentum in the Foreign Exchange Market

Abstract

For decades, academics and investors have searched for profitable portfolio management strat- egies while some fundamental theories in the field deny the possibility for gaining such excess returns. As a counter argument against these theories, momentum anomalies have received attention in academic literature. One of these anomalies is the momentum anomaly, which in- dicates that historical price or performance development of an asset can be used to predict fu- ture prices and performance. Several studies have shown the effectiveness of the respective strategies across several markets, but the source of excess returns is still a mystery. After mo- mentum anomaly gained a permanent foothold in equity market research, the phenomenon has been studied widely in other markets. Although excess returns have been linked to the respec- tive strategy on multiple occasions, the inherent source for such returns remains unknown. This master’s thesis examines macroeconomic momentum investment strategy expanding the exist- ing momentum research. As macroeconomic data is strongly linked to the pricing of currencies, we look for momentum trends from macroeconomic data variables. With these trends, curren- cies in the investment universe are set to a ranking which is used to create monthly balanced long short portfolios. Macroeconomic momentum portfolios show that the investment strategy produces unique ex- cess returns that well-known foreign exchange strategies cannot explain. Overall, the cumula- tive returns match almost the returns of benchmark strategies carry and momentum. However, compared to previous reports, the returns of macroeconomic momentum portfolios are rela- tively volatile. The background of the findings is influenced by possible differences in portfolio construction, differences in research samples and the lack of comparable research. The findings support the importance of further research. In addition to examining the currency market, the investment strategy could also be replaced in other markets for example fixed income products. Additionally, the results provide interesting findings on the most optimal lookback period for observing momentum trends. This study encompasses lookback periods from 1 to 60 months when typically, momentum research considers only lookback periods up to 12 months. The re- sults support the favorability of these short lookback periods. As a new discovery also excep- tionally long lookback periods from 37 to 60 months seem favorable. The findings highlight the importance of further research. Due to the lack of comparable re- search making large generalizations is still fairly early. Results of this study also provide contra- dicting evidence to previous studies finding which find relatively low volatility in the return dis- tribution of macroeconomic momentum portfolios. Through this there is an incentive for future research if momentum crashes found with other momentum strategies apply to these portfolios as well. In addition, the strategy could be applied to other markets like the fixed income market.