Estimating variability in the Italian stock market: An ARCH approach

The variability of financial markets has become the focus of considerable interest, especially over the past decade. In this study, ARCH models are applied to the Italian stock market, at both general and sectoral levels, to identify the processes generating variances and to test whether the variances are explainable by an autoregressive equation. The predictive power of the estimated equations has been evaluated by comparing them with forecasts obtained from alternative estimation techniques. The outcome supports the idea of an autoregressive structure for the variances and a hyperreactive behavior of the Italian stock market to the arrival of destabilizing news. Copyright Kluwer Academic Publishers 1993ARCH, maximum likelihood, variance,