Impact of information arrival on volatility of intraday stock returns

Abstract

In this empirical study we have considered the impact of information flow on the volatility of a particular stock using high frequency return and news data on the Eurostoxx 50 market. In addition to using volume as a proxy for information flow, we have included company specific announcements, to the conditional variance of the Generalized Autoregressive Conditional Heteroscedastic model (GARCH). For this purpose we have constructed five measures of the impact of public information flow in the market transforming commonly available news scores through different techniques such as linear and exponential decreasing weight, impact function etc. We have analyzed the behaviour of volatility, estimated by squared returns for the next 4 hours after arrival of a non overlapping news, having a significant impact on the firm’s stock return. A significant impact of the information flow accessed by the news score coefficient is observed for majority of in our analysis. Furthermore, the inclusion of the news scores variable improves the overall model in the sense that it increases the likelihood value of the model. However we do not observe any significant change in the volatility persistence due to inclusion of our news variable.