Timing Foreign Exchange Markets
Robert B. Gramacy
University of Chicago - Booth School of Business
Samuel W. Malone
University of the Andes
Enrique Ter Horst
September 30, 2012
Midwest Finance Association 2013 Annual Meeting Paper
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.
Number of Pages in PDF File: 18
Keywords: foreign exchange, speculation, Bayesian treed Gaussian process, Anatolyev-Gerko statistic
JEL Classification: F31, G15, G17working papers series
Date posted: October 2, 2012
© 2013 Social Science Electronic Publishing, Inc. All Rights Reserved.
This page was processed by apollo5 in 0.375 seconds