中文摘要：

文章利用时变相关的多元波动率模型，对美元汇率、上证指数以及地产指数的逐日数据建立三元时变相关的广义自回归条件方差（TV—GARCH）模型，从而得出三者的波动性特征以及其显著的时变相关系数，可知地产价格的收益率波动最大。汇率与地产指数、上证指数的相关关系复杂，可存在正相关或负相关关系，而股票市场与房地产市场间存在正向相关关系，且相关系数的大小随时间变化。

英文摘要：

This paper studies the multivariate volatility model with time-varying correlation coefficient, and then proposes a Triple-Volatility Model among the exchange rate, stock price and real estate price based on the daily data, and then get Triple-Volatility characteristic and the significant time-varying correlation coefficient. The study shows that volatility of the real estate price is the biggest among them, there is complicated correlation between the exchange rate and stock price, either positive or negative correlation is existent. Nevertheless, the stock price and real estate price have a positive correlation which is time-varying.