Parametric Estimation of Monthly Volatility Using Autoregressive Conditional Duration Models
Abstract
This paper employs a method to estimate monthly volatility by integrating the conditional return variance over a month using the autoregressive conditional duration (ACD) models. The ACD models fit the daily data surprisingly well. Maximum likelihood Estimation (MLE) method is used to estimate the conditional expected duration equation, which is assumed to follow the augmented ACD models. The estimated monthly stock volatility are adopted to investigate, if any, the link between macroeconomic variability and the stock market volatility. We find that, for the period 1944/01-1975/06, PPI inflation, monetary base growth and industrial production predict stock market volatility very well, which are estimated by ACD methods; the monthly stock volatility, estimated from ACD models, also helps predict the macroeconomic volatility in the period 1975/07-2008/12.