| The global economy has been experiencing the fragile and fluctuated recovery until today. According to the latest'World Economic Outlook' by IFM, the world economy is facing a significant downside risk which will inevitably affect the future of world economy. Due to the impact of global economic integration, increasing free capital flowing across borders, as well as increasing international trade, the linkage between financial markets worldwide is strengthening. As a result, the uncertainty in one market exerts impact on the other markets and enlarges the volatility in these markets as well. Therefore, whether this dynamic interdependence can be described precisely is of paramount importance in leading capital flow across countries, making and implementing related financial policies, making portfolio investment strategies, managing portfolio risk and so forth.Similar to international stock market, the focus of studies on the linkage between international real estate markets also lies in how to model the asymmetric, time varying and fat tail dependence structures. Traditionally, the Pearson liner correlation coefficient is the most popular tool to describe dependence between assets or markets. However, this method has received much blame both from practitioners and academic researchers recently since it is based on the assumption of Normal distribution and not capable of capturing non-liner dependence structure. Nevertheless, Copula method is able to overcome all drawbacks of liner correlation coefficient. Therefore, this study adopts this method, associated with GARCH model, to check the dynamical and asymmetric dependence structure and tail dependence in global real estate markets.This study adopts various kinds of test methods to confirm the dependence structure between price indices of public real estate markets are time-varying, asymmetric and tail correlated. Based on these test results, we choose dynamic rotated Gumbel Copula model, which is able to accommodate these characteristics, to capture the dependence of three securitized RE markets. The empirical results show that the linkage between US and UK RE markets is the strongest. Based on the previous results from Copula model, we further estimate the Copula-VaR (ES) to calculate the value-at-risk (expected shortfalls) of real estate-only portfolios. By comparing the estimation results from t and rotated Gumbel Copulas, we find that the value-at-risk and expected shortfalls estimated from the symmetric t Copula are much smaller. Therefore, the conclude is that the ignorance of asymmetric dependence will cause improper or even false portfolio risk management strategy.We then focus on the tail dependence which is used to describe the probability of joint crash between markets, especially during stressful times. Through examining the tail dependence of returns in international real estate markets, we analyze how the interdependence of international securitized real estate markets has changed since the Global Financial Crisis. The empirical results confirm that, in most cases, the relationship between countries has changed from tail-independent to tail-dependent since 2007. Strong tail dependence persists throughout during crisis and post-crisis periods. The effect of Global Financial Crisis on the interdependence of global real estate securities markets is profound and long term.Finally, we analyze the dynamic tail dependence between China's RE-Stock markets based on the time-varying SJC Copula-GARCH model. Our main finding is that the tail dependences between these two markets remain at a very high level all the observation time, and do not affected by the GFC and European Debt Crisis. This is quite different from the findings got from other countries, indicating that the extreme events from the global market only exert limited impact on China's market. |
PDF Full Text Request