An empirical evaluation of transitory and permanent components of the exchange rate volatility

We analyse the volatility behaviour of major currency rates, using daily spot exchange rate data for sixteen different currencies relative to the US from 1999 to 2023. We apply the component-GARCH model proposed by Engle and Lee (1999) to break down volatility into permanent and transitory components. In addition, we apply a correlation analysis between permanent and transitory volatilities to examine their statistical association, disentangle their mutual influence, principal components analysis of long-run and short-run volatility; and identify clusters among them. Results suggest that (i) permanent and transitory volatility components capture the most relevant events of the turbulent 21st century, showing higher volatility persistence with long memory in the permanent component than in the transitory one; (ii) cross-country correlations are lower for the transitory component than the permanent component and principal component analysis reveals a long-run volatility trend; (iii) there exists a big group of currencies with similar traits in the permanent and transitory components of volatility; and (iv) the transitory component is closely associated with measures of market sentiment and financial tensions. Therefore, this study provides important insights into the behaviour of various components of exchange rate volatility, highlighting key aspects that have not been sufficiently explored.