Edith Cowan University
Place of Publication
Perth, Western Australia
School of Accounting, Finance and Economics
We introduce in this paper a new semiparametric approach to a nonlinear ACD model, namely the Semiparametric ACD (SEMI-ACD) model. This new model is more flexible in the sense that the data are allowed to speak for themselves, without a hypothetical assumption being imposed arbitrarily on its key component. Moreover, it enables a much more thorough examination of the intertemporal importance of the conditional duration on the ACD process. Our experimental analysis suggests that the new model possesses a sound asymptotic character, while its performance is also robust across data generating processes and assumptions about the conditional distribution of the durations. Furthermore, the empirical analysis in this paper illustrates the advantage of our model over its parametric counterparts. Finally, the paper discusses some important theoretical issues, especially its asymptotic properties, in order to pave the way for a more detailed analysis, which will be presented in a future paper.