．單位首頁
．系所簡介
．系主任
．強力師資
．研究所
．課程概況
．各類學程
．期刊論文
．招生資訊
．演講專區
．榮譽榜
．系友天地
．就業訊息
97年6月6日(五) 2:00 ~3:00 P.M.
演講者姓名： Dr. Mike K. P. So
演講者服務單位： Department of Information and Systems Management, the Hong Kong University of Science and Technology, Hong Kong
Dynamic Modeling of Extreme Dependence Structure: Applications to Hong Kong, Chinese and Other Asian Stock Markets
Abstract
The most well known dependence measure in finance is the Pearson correlation coefficient which indicates the strength of the linear relationship between two random variables. However, correlation sometimes gives us an incompleted or even a misleading description of dependence. In recent development of risk management, there has been increasing emphasis on extreme market losses. Our discussion focuses on tail dependence coefficient (TDC) as an extreme dependence measure. TDC is defined as the conditional likelihood that extreme market losses occur simultaneously. As a main construction block of TDC time series modeling, we utilize a Multivariate Generalized Autoregressive Conditional Heteroscedasticity (MGARCH) model with time-varying correlation matrix proposed by Tse and Tsui (2002). Our methodology is applied to Asian financial markets where we provide evidences that the extreme dependence structures in these markets have generally become stronger in recent years.
國科會數學推動中心補助
