GARCH Type Volatility Models Augmented with News Intensity Data
It is well-known that financial markets and investors react nervously to important news, economic crises, wars, political disorders or natural disasters. In such periods financial markets behave chaotically, prices of financial assets may fluctuate very much and volatility changes over time. Understanding the nature of such time dependence is very important for many macroeconomic and financial applications. In this paper we analyze the impact of extraneous sources of information on stock volatility by considering some augmented Generalized Autoregressive Conditional Heteroscedasticity (GARCH) models. We will consider the ‘daily number of press releases on a stock’ (news intensity) as the most appropriate explanatory variable in the basic equation of GARCH model. The results of the likelihood ratio test indicate that the GARCH(1,1) model augmented with the news intensity performs better than the ‘pure’ GARCH model.
KeywordsStock Return Trading Volume GARCH Model Earning Announcement Brit Amer Tobacco
We would like to express our gratitude to Prof. Gautam Mitra, director of CARISMA, for the kindly provided opportunity to use Raven Pack news analytics data, and to Prof. Brendan McCabe and Keming Yu for helpful comments and remarks.
- Hansen P, Lunde A (2001) A forecast comparison of volatility models: does anything beat a GARCH(1,1)? Working paper, Department of Economics, Brown UniversityGoogle Scholar
- Kalev PS, Liu W-M, Pham PK, Jarnecic E (2004) Public information arrival and volatility of intraday stock returns. J Bank Finance 280(6):1447–1467Google Scholar
- Lamoureax CG, Lastrapes WD (1990) Heteroskedasticity in stock return data: volume versus GARCH effects. J Bus Econ Stat 2:253–260Google Scholar
- Mitra G, Mitra L (eds) (2001) The handbook of news analytics in finance. John Wiley & Sons, Ltd., Chichester, West Sussex, UK.Google Scholar