Timing Foreign Exchange Markets

Priced level, slope, and volatility risk factors recently proposed in the finance literature help explain long-standing puzzles related to the cross-section of carry trade returns. In this paper, we examine whether the information contained in these global factors allows foreign exchange market speculators in individual currencies to successfully time the direction of their carry trades, both when they can foresee the realizations of these factors one-month in advance, as in the classic exchange rate forecasting literature, and when they cannot. We find that, in stark contrast to most previous attempts to forecast monthly exchange rates, perfect foresight of these fundamentals confers statistically and economically significant market timing ability upon speculators. Conditional linear and nonparametric models based on these factors outperform the random walk. Without perfect foresight of these fundamentals, simple strategies based on the directional forecasts of conditional models still manage to outperform the random walk with respect to market timing statistics and realized Sharpe ratios for a large minority of currencies, especially when combined with information on global liquidity factors.