Real exchange rate volatility, financial crises and exchange rate regimes

Abstract

This article examines real exchange rate (RER) volatility in 80 countries around the world, during the period 1970 to 2011. Two main questions are raised: are structural breaks in RER volatility related to changes in exchange rate regimes or financial crises? And do these two events affect the permanent and transitory components of RER volatility? To answer these, we employ two complementary procedures that consist in detecting structural breaks in the RER series and decomposing volatility into its permanent and transitory components. Our results suggest that structural breaks in RER volatility coincidence with financial crises and certain changes in nominal exchange rate regimes. Moreover, our findings confirm that RER volatility does increase with the global financial crises and detect that the more flexible the exchange rate regime, the higher the volatility of the RER using a de facto exchange rate classification.

Keywords:

• financial crisis
• structural breaks
• component-GARCH model
• real exchange rates

JEL Classification:

• G01
• C22
• C54
• F33

Acknowledgements

The authors thank the editor and two anonymous referees for useful comments and suggestions. The authors also wish to thank Jushan Bai and Pierre Perron for kindly providing us with the GAUSS code for computations of their tests to detect multiple structural breaks and Ethan Ilzetzki for kindly providing us with the updated database on exchange rate arrangements. Responsibility for any remaining errors rests with the authors.

Funding

The authors gratefully acknowledge financial support from the Spanish Ministry of Economy and Competitiveness [project ECO2011-23189].

Notes

¹ Similarly, Stock and Watson (2002) use the absolute value of the fitted residuals of a VAR model to analyse changes in variance. Alternatively, Valentinyi-Endrész (2004) use the squared errors from a AR(1)-GARCH(1,1) model to compute changes in variance.

² For further analysis see Bai and Perron (1998, 2003).

³ The sample size for Nicaragua covers the period 1988:1 to 2011:12.

⁴ Data collected by Mathew Shane, Economic Research Service, the US Department of Agriculture.

⁵ The European Union (EU) was established on 1 November 1993 with 12 member states (Belgium, Denmark, France, Germany, Greece, Ireland, Italy, Luxembourg, the Netherlands, Portugal, Spain and United Kingdom). Their number has grown to the present 28 through a series of enlargements. In our analysis, we use the original EU-12 as an additional ‘country’.

⁶ We have made computations using the program TSW Caporello et al. (2012).

⁷ In order to save space, the numerical results of Bai and Perron’s tests are not reported in Table 1 but they are available upon request.

⁸ We summarize the results by pointing out the main regularities. The reader is asked to browse through Table 1a–f to find evidence for particular countries or group of countries of her/his special interest and the respective estimated break points. A more detailed account of the results by groups of countries can be found in Morales-Zumaquero and Sosvilla-Rivero (2012).

⁹ Except for in Malaysia, Egypt and Senegal.

¹⁰ This is the main regularity. The reader is asked to browse through Tables 2a–f and 3 to find evidence for particular countries or group of countries of her/his special interest. A more detailed account of the results by groups of countries can be found in Morales-Zumaquero and Sosvilla-Rivero (2012).

¹¹ To save space, we do not show here these figures. They are available in Morales-Zumaquero and Sosvilla-Rivero (2012).

¹² For details see Sokal and Rohlf (1995), Levene (1960), Conover et al. (1981), Brown and Forsythe (1974a, b) and Neter et al. (1996).