Generating currency trading rules from the term structure of forward foreign exchange premia

The quality of an exchange rate forecasting model has typically been judged relative to a random-walk in terms of out-of-sample forecast errors. The difficulty of outperforming this benchmark is well documented, although Clarida and Taylor have demonstrated how the random walk can be beaten in this metric by exploiting information embedded within the term structure of forward exchange rate premia. But this achievement does not guarantee success within an investment context. We therefore assess whether the Clarida-Taylor framework can be used to generate significant trading profits in combination with an acceptable degree of risk in a realistic investment portfolio context.