A Range-Based Multivariate Model for Exchange Rate Volatility

In this paper we present a parsimonious multivariate model for exchange rate volatilities based on logarithmic high-low ranges of daily exchange rates. The multivariate stochastic volatility model divides the log range of each exchange rate into two independent latent factors, which are interpreted as the underlying currency specific components. Due to the normality of logarithmic volatilities the model can be estimated conveniently with standard Kalman filter techniques. Our results show that our model fits the exchange rate data quite well. Exchange rate news seems to be very currency-specific and allows us to identify which currency contributes most to both exchange rate levels and exchange rate volatilities.