透過您的圖書館登入
IP:18.221.187.121
  • 學位論文

極值理論與關聯結構在國際股票市場間分散化的應用

The Application of Extreme Value Theory and Copula Approach in the International Diversification

指導教授 : 鍾麗英
若您是本文的作者,可授權文章由華藝線上圖書館中協助推廣。

摘要


本文研究G5工業國、東亞、拉丁美洲等十四個國家的 MSCI 大盤指數報酬率在金融風暴前後,左尾極值是否有明顯變化,並比較傳統相關、關聯結構與極值相關等是否有所不同。極值理論 (Extreme value theory) 著重尾端極端事件的發生情形,可以用來描述非對稱分配的報酬,更精確地捕捉厚尾現象。Poon, Rockinger及Tawn (2004) 提出雙變量極值模型,藉由兩個非參數統計量,分別估計尾部收斂速度及尾部相依程度;本研究以 Poon 等人於2004年所提出的雙變量極值模型作為本文研究之方法,並且設定各種分配之關聯結構 (Copulas),對各國股票報酬率間之關聯性進行建模,觀察各國股市之間相關結構是否產生變化。 研究發現,不論是普通的相關係數或是極值相關都會隨著時間增加,且在三大區域中,拉丁美洲有最低的左尾風險,導致分散化的效果最佳,而極值相關的衡量在東亞較不穩定,代表相關具有複雜性,會降低執行分散風險的效果。金融風暴前與金融風暴後,有最高的平均報酬區域分別為拉丁美洲及東亞,且跨區後左右尾的相關強度比跨區前小,代表經過跨區投資能有效分散風險。以變異係數來看,拉丁美洲獲得一單位報酬所承受的標準差相對於其他兩區域小,因此投資拉丁美洲能獲得較大的單位利益。

並列摘要


In this paper, we study the dependence structure of asset returns among three regions; G5, East Asia, and Latin America in MSCI stock markets before and after financial crisis to investigate international diversification. Both Correlations and extreme dependence were examined in this study. The tail behavior of extreme events in extreme value theory can be used to describe an asymmetric distribution of asset returns. In this study, we used two non-parametric statistics in bivariate extreme value model proposed by Poon, Rockinger and Tawn (2004) to estimate tail converge rate and tail dependence separately. There are three main findings for dependences. First, both correlations and extreme dependence increase over time and Latin America has the lowest downside risk (best diversification) among regions. Second, the highest average return occurred in Latin America before financial crises and East Asia after crises. Third, Latin America has the lowest coefficient of variation among regions.

參考文獻


黃維誠 (2009), 在非對稱條件波動下風險值跳躍擴散模型之研究, 國立成功大學統計學系碩博士班碩士學位論文。
Hartmann, P., Straetmans, S., de Vries, C. (2004). "Asset market linkages in crisis periods", The Review of Economics and Statistics 86 (1), 313–326.
Ang, A. and Bekaert, G. (2002). "International asset allocation with regime shifts", Review of Financial Studies 15 (4), 1137–1187.
Ang, A. and J. Chen. (2002). "Asymmetric Correlations of Equity Portfolios", Journal of Financial Economics, 63, 443–494.
Bollerslev T. ((1986). "Generalized autoregressive conditional Heteroskedasticity", Journal of Econometrics, 31, 307-327.

延伸閱讀