Impact of news announcements on stock market returns and volatility spillover across north America, European Union and Pacific Asia

Abstract

Purpose – This thesis has two objectives. The first one is to test empirically the level of interdependence across major stock markets returns, namely, US, EU, and Asia in terms of return and volatility spillover. The second one is to evaluate the impact of news announcements on their stock market volatility. Design/methodology/approach – To model the volatility of the three above mentioned stock markets we apply a battery of univariate time series models from the GARCH family. Dickey Fuller unit root test is used in order to ensure that the three return series are stationary. The mean equation in the GARCH is assumed to follow an ARMA process and the residuals of the GARCH are tested for heteroscedasticity and autocorrelation using the Ljung-Box test. A comparative approach is considered and the best GARCH model is the one that has homoscedastic and not autocorrelated errors. To test for volatility spillover we consider a simultaneous equation model estimated using a three stage least square approach that tackles the problem of endogeneity. To sum up, this study applies a combination of econometric tools –ARMA-EGARCH and 3SLS technique to spot both the instantaneous and delayed volatility spillovers among major stock market returns and to examine the impact of news surprises. Findings – Empirical results show that news announcements significantly affect stock market returns in US and Asia. Furthermore, news announcements affect the transmission of volatility between US and Asian stock markets, however; no volatility spillover was found between EU and US in terms of news announcements. We also found significant evidence of bidirectional volatility transmission between US and Asia stock market returns and between EU and Asia stock market returns. Furthermore, Negative shocks are found to have more impact on all stock returns under study than positive shocks. Practical implications – Our empirical findings contribute to the literature of interdependence among major stock markets returns. It provides insight on the impact of news announcements on stock market returns and volatility spillover. It also provides gaudiness for investors and portfolio managers to effectively implement diversification and hedging strategies. Originality/value – Most studies consider economic surveys such as the consumer price index, the targeted federal funds rate, the unemployment rate, and non-farm payroll to capture the effect news announcements on volatility spillover. This thesis uses a one comprehensive indicator for news announcements, the Bloomberg Economic Surprise Index (BESI). BESI encompasses all the above and takes into consideration changes in 39macroeconomic and financial indicators. Second, this thesis employs comprehensive data covering recent period and includes major stock markets around the world, namely, US, EU, and Pacific Asia countries.