Asymmetric Stochastic Conditional Duration Model--A Mixture-of-Normal Approach

This paper extends the stochastic conditional duration model first proposed by Bauwens and Veredas (2004) by imposing mixtures of bivariate normal distributions on the innovations of the observation and latent equations of the duration process. This extension allows the model not only to capture various density shapes of the durations but also to easily accommodate a richer dependence structure between the two innovations. In addition, it applies an estimation methodology based on the empirical characteristic function. Empirical applications based on the IBM and Boeing transaction data are provided to assess and illustrate the performance of the proposed model and the estimation method. One interesting empirical finding in this paper is that there is a significantly positive correlation under both the contemporaneous and lagged intertemporal dependence structures for the IBM and Boeing duration data.