The role of permanent and transitory components in the fluctuations of Latin–American real exchange rates

Abstract

Following the approach suggested by Engel and Kim (1999), we estimate the permanent and transitory components of the real exchange rates in four Latin–American countries for the period 1957:01 to 2002:04. Results suggest that transitory component is the driving force of the real exchange rates in Argentina and Mexico. A principal role of the permanent component is observed in the real exchange rates of Brazil and Chile. Estimates probabilities of the high-variance regime allow to identify the principal events happened in these countries. This information is closely related to nominal shocks and therefore, it explains the significant role of this component in these countries.

Acknowledgements

This article is drawn from the Indira Romero's M. A. Thesis at the Department of Economics, University of Ottawa. We thank Emmanuel Apel for useful comments to an earlier version of this article. We also thank an anonymous referee and the Editor for valuable comments. Rodríguez thanks financial support from the Social Sciences Faculty of the University of Ottawa.

Notes

¹ An extensive revision of the literature is beyond the goal of this article. Good surveys are presented in Sarno and Taylor (2002), Taylor and Taylor (2004), Taylor (2006). See also the earlier and general survey by Taylor (1995).

² Gibbs-sampling, a Markov chain Monte Carlo method, is a technique for generating random variables from a distribution indirectly, without having to calculate the density.

³ Coakley et al. (2005) have shown that even if there are permanent components to real exchange rates, long-run relative PPP may still hold.

⁴ Interpreting these shocks in this way is coherent with the characteristics of these shocks in many optimizing models; see Stockman and Tesar (1995), Asea and Mendoza (1994) and Obstfeld and Rogoff (1995).

⁵ Notice that a model with heteroskedastic disturbances was estimated but not improvements were observed. In the same spirit as Engel and Kim (1999), no drift is included in the specification of the permanent component. As they argue, there is unlikely that permanent shocks could impart drift to the relative prices over a long period.

⁶ Taylor (2004) has suggested that switches in the real exchange rate between random walk and stationary regimes may be due to the interaction of noisy traders, informed traders and government intervention.

⁷ Where X ^T , Y ^T and S ^T denote the vector of T values of the transitory component, permanent component and the state S _t , respectively.

⁸ In order to save space, tables with estimations are available upon request.