Term Spread의 중국경기예측력에 관한 연구

Abstract

In a time when the influence of China's economy has been growing day by day, it is significantly important for Korea to develop variables that could predict economic fluctuations of China and the changes in its real economy since Korea's economic dependency on China is huge. Predictability of Term Spread has widely been examined in the West and is universally recognized. There could be two contradictory opinions whether the Term Spread can effectively predict changes in China's economy. First, as it relates to whether Term Spread can reflect market expectation, expectation on the central bank would actively react to the economy. And when it comes to China, the central government’s will is what is most important. Considering that the central bank is enforcing the will of the central government, instead of being independent, Term Spread in China will have high predictability. Second, when Term Spread considers the yield curves of bonds different only in maturities while fixing other differences, and considering that the systems in bond market should mature and the transaction should be active to form a right expectation on future economy, a question arises whether the current bond market in China can play this role. Thus, this study examined that Term Spread can predict real economy of China and its economic fluctuations through empirical study. First, among terms spreads consisting of the yield differences between 10-year bond and 1-year bond, 10-year bond and 2-year bond, and 5-year bond and 1-year bond, term spreads between 10-year and 2-year bonds are found to be best in predicting economic fluctuations and changes in GDP. In particular, if the difference is 2 to 4, Term Spread showed a high predictability and explanatory power in economic fluctuation and GDP. After conducting an empirical study on the predictabilities between leading composite index and Chinese PMI index through probit model, it has been found that leading composite index and PMI index independently have predictability, but form three composite index models with Term Spread, which was found to be the most superior in terms of predictability. As it relates to Term Spread, it showed statistically significant predictability from one month to 25 months whereas composite index and PMI were found to have statistically significance only in one month. In other words, composite index and PMI can be explained by Term Spread.