Modeling & testing for volatility of the monthly rate of return on the US($)/AUS($) exchange rate

Abstract

The role of news is found to be fundamentally useful in understanding the behaviour of financial market volatility. It has been found that the Engle’s basic ARCH models are incapable of capturing all observed phenomena, such as asymmetric effect, excess kurtosis and high degree of nonlinearity, which are often the stylized facts exhibited by most financial and economic time series. Bollerslev’s GARCH has the similar cavities as the ARCH. Although the EGARCH and GJR models capture the asymmetric news and nonlinearities, performances of these two models are quite different. News Impact curve approach to the GARCH, EGARCH, and the GJR models & the diagnostic tests of asymmetry of this paper indicate that the EGARCH model is preferred to the other two models in explaining the asymmetric movements in the rate of return on the US/AUS exchange rates.