MODELING FOREIGN EXCHANGE RATES USING COPULA-BASED AUTOREGRESSIVE CONDITIONAL DURATION MODELS

摘要

This paper formulates a copula-based generalization of bivariate autoregressive conditional duration models for irregularly time-spaced data. To construct bivariate durations from two irregularly spaced dependent point processes the approaches introduced by E ngleand Lunde (2003) and Mosconi and Olivetti (2005) are both considered. The models are illustrated using exchange rates of three major currencies in Indian Rupees. The findings on the degree of dependence between foreign exchange rates are indicators for managing future exchange rate risks.