Volatility Transmission and Correlation Analysis between the USA and Asia: The Impact of the Global Financial Crisis

Abstract

This paper examines volatility transmission and conditional correlations behaviour between the US and the Asian stock markets considering the effect of the Global Financial crisis. One Asian mature market and 10 emerging markets are included in the sample. To carry out the analysis, we use a multivariate asymmetric GARCH model. Results show that there exists volatility transmission between the US and the Asian markets. Moreover, it is found that, after the crisis, volatility transmission patterns have barely changed. Finally, results suggest that the lower the country‘s level of development, the lower the correlation with the USA.

Keywords:

• Volatility spillovers
• GARCH
• correlation
• international
• financial markets

Keywords:

• C32
• G01
• G15

Acknowledgements

We are grateful for the financial support of the Spanish Ministry of Education and Technology and FEDER under the project ECO2009-14457-C04-04.

Notes

1. Results (not reported) do not change when we introduce a dummy variable which is equal to 1 from the Lehman Brothers bankruptcy until the end of the sample period and 0 otherwise. These results are available from the authors upon request.

2. When the ADF test is applied to the first difference of individual time series, the null of unit root process is strongly rejected in all cases.

3. Asymmetric volatility refers to the empirical evidence according to which a negative shock increases volatility more than a positive shock of the same size. In the financial literature, two explanations of the asymmetries in equity markets have been put forward: the leverage effect and the volatility feedback effect. Which of the two effects is the main determinant of asymmetric volatility remains an open question.

4. A range of starting values was used to ensure that the estimation procedure converged to a global maximum. The estimations were repeated with random restarts of the starting values. None of the estimation results indicated any local maximum. The results also seem robust to alternating convergence criterions and optimizing methods. Consequently, we are confident that we have found a global maximum.

5. Results for the VAR model (not reported) are available from the authors upon request.

6. The coefficient for Philippines is significant at the 10% level.

7. The coefficient for Taiwan is significant at the 10% level.

8. This fact could be due to the Korean Government's announcement of an economic stimulus plan. After this announcement, Kospi Index returns raised while S&P500 Index returns decreased.