题 目：Stationarity, Ergodicity and Finite/Infinite Moments of Log-ACD Models, and Comparisons

with Alternative Conditional Duration Models _

报告人：Chor-yiu SIN（副教授）

Wang Yanan Institute of Studies in Economics (WISE)

Xiamen University

时 间：2008年1月2日（周三）上午10：00-11：30

地 点：伟易博116

Abstract

Modified upon the autoregressive conditional duration (ACD) model suggested by Engle and Russell (1998), the logarithm ACD (Log-ACD) model proposed by Bauwens and Giot (2000) has been attracting a lot of attention in the literature of modelling the duration of high-frequency financial data. While ACD model resembles the GARCH model, the Log-ACD model resembles the Exponential GARCH (EGARCH) model. As pointed out by Nelson (1991), the probabilistic properties of this kind of models need to be discussed with primitive assumptions. Assuming the standard error follows a generalized gamma distribution, we prove conditions on the parameters that suffice for stationarity, ergodicity and finite moments of the duration. These conditions are compared with the corresponding ones for the power ACD model, for which distributional assumptions are not required. All these conditions complement those proposed by Carrasco and Chen (2002). We conclude with a discussion on applying these conditions to the maximum likelihood estimation (MLE) and the sample autocorrelation function.