EXTREME VOLATILITY DEPENDENCE IN EXCHANGE RATES

Abstract

This paper aims to analyse asymmetric volatility dependence in the exchange rate between the British Pound, Japanese Yen, Euro, and Mexican Peso compared to the U.S. dollar during different periods of turmoil and calm sub-periods between (1994–2018). GARCH and TARCH models are employed to model conditional variance. Once volatility is estimated, the Copula approach provides an upper and lower tail dependence measure for each subperiod: 1994–1999, 2000–2007, 2007–2012, 2013–2018. The empirical joint distribution of exchange rate volatility pairs displays high tail-dependence in the lower tail and low tail-dependence in the upper tail. Results have important implications in term of risk management and investment strategies.