ACKNOWLEDGEMENTSForemost, I would like to express my sincere gratitude to my supervisor, Lu Liu, for her guide and persistent support. Not only has she taken her time and patience to give me a number of valuable suggestions, but also showed me the most suitable approach to address the research questions. She has always kept track of my thesis to ensure that I was on the right progress.I am also indebted to Tuan-Minh Nguyen at Department of Mathematics, who has always offered me his enthusiastic assistance to better understand various statistical and mathematical issues.
Dinh-Vinh Vo
ABSTRACTThe paper sheds light on financial contagion within the Euro Area and Asia, and contagion from the Euro Area to Asia during two recent crises: the global financial crisis and European sovereign debt crisis. Applying the multinomial logit regression model, the paper investigates how the macro-finance variables affect the coincidence of extreme negative returns (coexceedances). In addition, I apply both original constant threshold i.e. 5%percentile of unconditional distribution of daily stock returns and Value-at-Risk to estimate extreme negative returns. These approaches offer a similar pattern. The empirical findings reveal that, in the Euro Area and Asia, the probability of the occurrence of coexceedances is strongly explained by the idiosyncratic risks: the changes in exchange rates, the regional stock market volatility, and global shocks: the changes in the U.S. long-term interest rates, the TED spread. The global volatility index is only significant to explain the likelihood of coexceedances in the Euro Area, not in Asia. These analyses lead to the conclusion that contagion in Asia is more important than in the Euro Area. Another important finding indicates the existence of contagion from the Euro Area to Asia. That is, the probability of coexceedances in Asia is predictable and depends on the number of joint occurrence of extreme return shocks in the Euro Area.