Would the bond market be a safer choice for Chinese investors when the stock markets are in turmoil? In order to provide advice for the investors attempting to enter into bond market, this paper employs Copula functions to analyze the nonlinear dependence structure between stock and bond markets, as well as the non-linear dependence structures between different bond types. The Markov switching technique is then used to explore the possible time-variety in these dependence structures. The empirical results show that, for daily returns, the relationship between stock and bond markets is rather weak. While for weekly and monthly returns, the two markets display a negative relationship, and the tail dependence exists in their dependence structure. In addition, on the bond market, there are obviously positive relationships between different bond types. Their dependence degrees tend to be higher for the returns with lower frequency, and their dependence structures are non-linear and time-varying. Our comprehensive analysis in this paper is quite valuable, especially when investors consider introducing bonds into their portfolios.
bond marketnon-linear dependence structureCopula function
国家科技期刊平台
NSTL
万方数据
维普
