中文摘要：

危机传染效应以及股市间日趋紧密的联动性大大增加了各国政府跨市场风险监管和协调的难度。基于全球21个代表性股市在1995年9月-2015年5月间的走势,实证分析各国股市泡沫演化过程及其相依结构特征,结果显示：（1）21个股市均存在多次泡沫,且泡沫严重时期主要集中在1997年、2006年-2007年间;中国沪深股市泡沫程度相对更为严重,美国股市泡沫程度最小;各股市泡沫之间的相依结构分布具有明显的区域聚集特征。（2）各股市泡沫之间普遍存在较高的相依性,且呈现出对称或不对称的上下尾相依结构特征;新加坡股市、巴西股市、德国股市和美国股市起到枢纽中心地位的作用。（3）R-Vine Copula方法的建模效果优于C-Vine Copula和D-Vine Copula方法。

英文摘要：

Crisis contagion effect and increasingly close linkage among stock markets have greatly increased the difficulty of cross-market risk supervision and coordination for all governments. This paper makes an empirical research on price bubbles in global stock markets which include 21 representative stock markets based on backward sup ADF test method （BSADF） from September, 1995 to May, 2015. It finds that there are some significant bubbles among those stock markets in the entire period. Most serious bubble occurred in 1997 and the period from 2006 to 2007. The bubble of Shanghai and Shenzhen stock markets in China is more serious than that of other stock markets. The bubble of US stock market is the smallest. In the entire period, there is a high dependence among those stock markets＇ bubbles, which shows a dependency structure of symmetry or asymmetry upper and lower tail. Singapore and Brazil have played a central role in the dependence structure of Asian stock bubbles, and at the same time, UK and USA have played a central role in the dependence structure among European and American stock bubbles. R-Vine copula model is more suitable for measuring the dependency structure of stock bubbles among global stock markets compared with D- Vine copula and C-Vine copula model.