구조형 벡터 자기회귀 모형을 이용한 뉴스 분해: 범중국 주식시장

We decompose unexpected excess market returns into permanent cash-flow news, temporary cash-flow news, discount-rate news, and non-fundamental news using our log-linear SVAR model. We investigate the relative importance of news conditioned on market status in pan-Chinese stock markets (China, Hong Kong, and Taiwan). The major empirical findings are as follows. First, most excess market return variations are the result of discount-rate and non-fundamental innovations for China and Hong Kong, while, for Taiwan, they are explained by their own innovations. Second, contrary to the evidence of the U.S., discount-rate news is more important than cash-flow news in driving stock price movements in pan-Chinese stock markets. Third, non-fundamental (i.e., investor sentiments) also plays an important role in Hong Kong and Taiwan but not for China. Finally, news premiums are higher in a down market than in an up market. The contribution of this paper is two fold. First, we develop a new pricing test method for the log-linear model. Second, we examine whether investor sentiment can be a risk factor in pan-Chinese stock markets.