# Journal of Time Series Analysis

## ON THE STABILITY OF A HETEROSCEDASTIC PROCESS

### Journal Article

Abstract. In this paper we derive the stability conditions in a time series regression model with a particular form of conditional heteroscedasticity. The variables affecting the variance include lagged errors, lagged dependent variables and a forecast variable. In addition to the usual stability conditions in a dynamic model, a condition on the parameters of the heteroscedasticity equation is also required.

View all

View all