Trend shifts in the forward premium and the predictability of excess returns in currency markets

ABSTRACT

This article provides evidence that the forward premium involves structural changes in the trend function, which might affect the predictability of currency excess returns to be dependent on the choice of the sample period. Accounting for the shifts in trend for the forward premium reveals that currency excess returns for the Canadian dollar, Swiss franc, euro and pound against the US dollar are significantly predictable irrespective of the sample period selected. Another advantage of detrending the forward premium is that we can obtain more consistent slope coefficient estimates in the predictive regression, which enables us to make more consistent, dependable inferences about the excess return predictability.

KEYWORDS:

• Forward premium anomaly
• shifts in trend
• structural change
• predictability
• Currency excess returns

JEL CLASSIFICATION:

• C12
• C22
• F31

Acknowledgements

The authors are grateful to the Editor Mark Taylor, two anonymous referees, as well as seminar participants at Sungkyunkwan University for their helpful comments and suggestions. Any remaining errors are solely the authors’ responsibility.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹ Other studies have investigated the predictability of excess returns in financial markets. Bacchetta, Mertens, and Van Wincoop (2009) show that the predictability of expectational errors tends to coincide with excess return predictability in bond, currency and stock markets. Evans and Lewis (1993) provide evidence that there are stochastic trends in excess returns in both the bond and currency markets.

² As noted by Fama (1984), this is equivalent to testing whether $\gamma =1$ in the following standard forward premium regression: $\mathrm{\Delta }{s}_{t+1}=\mu +\gamma \left(f_t-s_t\right)+{\zeta }_{t+1}$, where ${\zeta }_{t+1}$ is a disturbance.

³ See Perron and Yabu (2009) for further details of the test procedure.

⁴ Before the introduction of the euro in January 1999, we use the Deutsche mark/US dollar exchange rate adjusted by the official conversion rate between the euro and the Deutsche mark.

⁵ As mentioned in Zhou (2002), based on the overshooting exchange rate theory, such a trend shift in the time series behaviour of the forward premium $\left(f_t-s_t\right)$ and interest rate differential $\left(i_t^{\ast }-i_t^{US}\right)$can be explained by a switch in monetary operating procedures by the Central bank. During the recession period, the monetary authority may implement expansionary monetary policy by lowering the level of the interest rate to boost the economy. Such monetary policy or shock may lead to a trend shift in the forward premium and interest rate differential.

⁶ As Baillie and Bollerslev (2000) explain, it also applies to the case where the forward premium is described by a fractionally integrated, or I(d) process where $0<d<1$.

⁷ We are grateful to one of the anonymous referees for suggesting this.

⁸ The results of the F-tests for the non-linear trend are available from the authors upon request.