Does Limited Attention Drive Momentum Effect in FX Markets?
Posted: 25 Jan 2018 Last revised: 1 Jul 2022
Date Written: January 16, 2018
Abstract
This paper studies the momentum effect in foreign exchange (FX) markets. We test the frog-in-the-pan (FIP) hypothesis of Da et al. (2014) in FX markets that states investor limited attention drives momentum effects. FIP-like investors underreact to information arriving continuously in gradual changes, leading to stronger momentum effects. We construct currency momentum portfolios by extending the data period studied in Menkhoff et al. (2012b). Data include 48 currencies, covering the period of 40 years. The empirical results show that the average of returns for momentum strategies are up to 15% per annum (p.a.). As we further sort momentum portfolios by information discreteness, momentum returns are up to 40% following continuous information. In contrast, momentum returns become insignificantly different from zero following discrete information. The cross-sectional regressions and robustness checks show three main conclusions as follows. First, momentum effect still exists in FX markets until 2016; Second, discrete information mitigates momentum effects due to more attraction to investors’ attention, while continuous information induces stronger momentum effects; Third, investor limited attention causes underreaction and further drives momentum effects in FX markets. The extent of FIP phenomena varies with currencies. In the currency market features, high liquidity and large transaction volume, behavioral biases are more likely to influence investors’ decision-making and lead to momentum effects.
Keywords: Limited Attention, Momentum, Foreign Exchange (FX) Markets
JEL Classification: F31, G02, G12, G14
Suggested Citation: Suggested Citation