Extracting global stochastic trend from non-synchronous data

Abstract

We use a Kalman filter type model of financial markets to extract a global stochastic trend from the discrete non-synchronous data on daily stock market index returns of different stock exchanges. The model is tested for robustness. In addition, we derive "most important" hours of world financial market and estimate the relative importance of local versus global news for different stock markets. The model generates results that are consistent with intuition. Key words: emerging stock markets, transition economies, financial market integration, stock market returns, global stochastic trend, state space model, Kalman filter, non-synchronous data. JEL codes: C49, C58, G10, G15, F36, F65