A Study Of Dynamic Dependence Features And Driving Factors Of International Stock Markets

Estimating the international stock market dependencies accurately, quantitative analyzing and understanding the driving factors, which have great significance for investors to diversify risk, and for policy authorities to make market regulations and prevent financial risks. In order to get comprehensive understanding of stock market dependencies and the influencing mechanism, this dissertation first quantitative analyzed general features and driving factors of the global stock market dependencies systematically and dynamically, then analyzed features of market co-movement and contagion when financial crisis happened, thirdly analyze features and driving factors of dynamic dependencies between China and other markets.First this dissertation analyz ed general features of dynamic dependencies among global stock markets from different aspects. Considering market dependencies would be different in different areas, we divided 36 global stock markets into three regions : America, Asia-Pacific, and Europe, and tested market linkage features in each region. We used VECM model to test long term and short term causal relations, used state space model to test dynamic price causality, and used AG-DCC model and rolling history correlation model to test dynamic correlation. The market dependence feature test of America showed that, Americas markets had financial contagion effects, correlations between markets would increase significantly when negative impact happened. The US and Canada markets had the highest correlation; Mexico and Brazil markets had high correlations with the US and Canada markets; Venezuela market was independent. The market dependence feature test of Asia Pacific region show ed that, the correlations among markets were strengthened by the Asian financial crisis, and weakened by the US subprime mortgage cr isis. Correlations among Southeast Asian emerging markets were most significant; correlations between Singapore and emerging markets were significant; Russian market was independent. The market dependence feature test of Europe region showed that, the correlations between Greece, Portugal, Ireland markets, which facing sovereign debt crisis, and other Europe markets, were decline after debt crisis. The average dynamic correlations among Europe markets were higher than correlations among America markets and correlations among Asia Pacific markets; correlations among Western Europe markets and correlations among southern Europe markets were high, but correlations among central and eastern Europe and correlations among northern Europe were low.Then this dissertation quantitative analyz ed general driving factors of global stock markets dynamic dependencies. We established a analysis system which cover economic and trade factors, financial market characteristics factors, and regional contact and regional economic cooperation factors. First we analyzed theoretical influence path of every factor, then we chose market dynamic correlations as samples, and used panel data model to empirical test which factor is significant. The empirical test on economic and financial connection showed that, the difference of GDP and the difference of real interest rate ha d negative effect on market linkage, relative export dependency and relative import dependence ha d positive effect on market linkage, inflation difference and bilater al exchange rate changes had no significant effect on market linkage. The empirical test on financial market characteristics showed that, the difference of market value, the difference of financialization degree, and the difference of market volatility ha d negative effect on market linkage, the difference of listed company numbers and the difference of capital account openness had no significant effect on market linkage. It also showed that correlations among developed markets were higher than correlations among emerging markets. The empirical test on location and regional economic cooperation showed that, the geographic distance between countries ha d negative effect on market linkage, joining NAFTA and ASEAN would improve market linkage, but joining EU had no significant effect on market linkage.Consider that market dependence would rising rapidly or deviating from long term equilibrium during financial crisis, the dissertation analyzed the special characteristics and influencing mechanism of dynamic market dependencies in the special period of financial crisis. Firstly, we reviewed the development, related stock market performance, and characteristics of risk contagion of Asian financial crisis, US subprime crisis, European debt crisis. Secondly, we empirical tested market co-movement effects in financial crisis: used Granger causality model to test causal relation, used fixed coefficient model, rolling history correlation model and AG-DCC model to test correlation, used time varying t-Copula model to test tail dependence, used panel regression model to test pure contagion. The conclusion showed that three crisis had positive direct effects on market linkage, Subprime crisis had the biggest impact, European debt crisis had the smallest impact. Thirdly, we constructed a financial risk contagion model based on dynamic game to explain market contagion mechanism in financial crisis from a behavioral finance perspective, which focused on diversification strategy of market participants and information asymmetry. The conclusion showed that, crisis in emerging market was difficult to infect developed market, but crisis in emerging market would infect another independent emerging market through some developed markets; when developed market ha d crisis, capital would flow to emerging market which had good economic fundamentals, so crisis in developed market might promote prosperity of emerging market; emerging markets which had similar risk sensitivity would suffer more losses than developed markets when facing international share impact factors such as drastic fluctuation of oil price.Consider that as a typical emerging market which is in the process of economic integration and financial liberalization, China stock market might have special linkage characteristics which other markets, the dissertation then analyzed the special characteristics and driving factors of dynamic dependencies between China and foreign stock markets. First we improved AG-DCC model from mean spillover effect, asymmetric volatility and asymmetric dynamic correlation. Then we used the improved AG-DCC model to empirical test dynamic dependence characteristics between China and foreign stock market s. The test showed that, in terms of correlation degree, the correlation between China and Hong Kong marke t was the highest, the correlations between China and other Asian market were relatively high, the correlations between China and other European and American developed markets were relatively low, Hong Kong stock market was an important medium for China market to connect with foreign market; in terms of dynamic trend, the correlations between China and foreign markets were increasing gradually in 1997 to 2010, the correlations between China and foreign markets except Hong Kong market was declining in 2010 to 2015. Thirdly we empirical test driving factors of linkage between China and US market, the conclusion showed that, besides general significant influencing factors such as China’s exports to US, the difference of inflation, the difference of real interes t rate, and the difference of market value, there were another two important factors could improve the China market’s integration to the world: Chinese companies listed overseas, and China’s financial liberalization. Finally this dissertation proposed policy suggestions to deal with the phenomenon and the characteristics of stock market dependencies on two aspects, including policy to deal with financial liberalization and stock market co-movements, and strategy to prevent financial contagion during the period of crisis.