S/F/28. Masters Dissertation. A study of dynamic linkages among major international stock markets: analysis from pre- and postglobalisation period

This dissertation aims to investigate dynamic interrelationship among major international stock markets using monthly data from 1970:1 to 1979:12 corresponding to preglobalisation and from 1984:1 to 2004:6 corresponding to postglobalisation. To facilitate analysis, Johansens multivariate cointegration technique is used to test for the number of cointegrating vectors. Granger causality tests in the VECM framework and forecast error variance decompositions are employed to explore causal relationship between our markets and partition variance in a market due to the shocks in other markets, respectively. The study is extended to explore time-varying nature of stock markets interrelationship. In this part, rolling causality tests with 5 year time window will be applied to the monthly return series starting from 1980:01 to 2004:06. Our empirical investigation indicates one and two cointegrating vectors over pre- and postglobalisation periods, respectively. As suggested by our formulation of VECM and VDCs, only the US market seems to maintain its dominance as the most influential market over both samples. At the same time, US market did not allow any market to explain its own shocks to significant extent. Based on causality tests and VDCs, France and Canadian markets seem to become relatively exogenous while German market became relatively endogenous over the postglobalisation period. The analysis from rolling causality tests showed evolving pattern of interdependence between markets. Evidence was found on the increased causal relationship during the second half of 1990th, which corresponds to bullish market and the least interdependence at start of new millennium, representative to bearish market. This analysis further confirmed exogeneity of US market and strengthened position of French market over the period under consideration.