On log-symmetric duration models applied to high frequency financial data

This paper deals with a new generalization of autoregressive conditional duration (ACD) models. In special, wepropose a new family of ACD models based on a class of log-symmetric distributions. In this new class, it is possibleto model both median and skewness of the duration time distribution. We discuss maximum likelihood estimation ofthe model parameters. For illustrative purposes, we analyze a high frequency financial data set from the German DAXin 2016.