The Effects of Sentiment on Extreme Movements in Exchange Rates

Abstract

We investigate details of the effects of sentiment on dollar exchange rates in six countries, in particular, during periods of extreme movements. Using the Generalized Sup Augmented Dickey-Fuller (GSADF) test, we first show evidence that there are periods of explosive patterns in the exchange rates during the sample period from January 2000 to December 2020. We then examine how sentiment affects the exchange rates during the periods of the extreme movements. Our results show that sentiment has significant effects on exchange rates only when the exchange rates decrease in extreme ways, i.e. the currencies appreciate sharply against the USD. Moreover, these sharp appreciations occur when sentiment in the six countries is more optimistic than that in the US. Therefore, our study shows that currency values appreciate by sentiment and this happens rapidly over a short period.

KEYWORDS:

• GSADF Test
• sentiment
• extreme movements of exchange rates
• Sharp decrease in exchange rates

JEL Classifications:

• G41
• E71
• F31

Acknowledgment

We would like to thank two anonymous referees for helpful comments.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 As we use dollar rates in this study, i.e., the exchange rates of currencies against the USD, a decrease and increase in an exchange rate indicates an appreciation and depreciation of the currency against the USD, respectively.

2 The panel regression is conducted to examine it under the control of macroeconomic variables. Analyses using panel data can complement each other because the data hold not only cross-sectional data information but also time series data information. Furthermore, individual and time effects, which are not possible in time series or cross-sectional analyses, can be controlled, and the problem of multicollinearity can be reduced (Baltagi, 2001).

3 In the case of Switzerland, the IPI exists only on a quarterly basis, and thus, interpolation is used to convert it into monthly data. The 91-day CDs are used for short-term interest rates in the case of South Korean won.

4 The lag selection criteria result in significant size distortion and reduce the power of both the SADF and GSADF tests (Phillips et al., 2011).

Additional information

Notes on contributors

Soosung Hwang

Soosung Hwang is Professor of Department of Economics, Sungkyunkwan University.

Eunji Lee

Eunji Lee is Ph.D. candidate of Department of Economics, Sungkyunkwan University.