Investor sentiment-styled index in index futures market

In this paper, we define and analyze the sentiment-styled index for the CSI 300 index futures in the Chinese futures market. Our sentiment-styled index for the CSI 300 index futures from April 16, 2010 to April 30, 2019 is constructed by the first and second principal component analyses, rather than only by the first principal component analysis used in the Baker and Wurgler (Journal of Finance 61(4): 1645-1680, 2006) method. The sentiment-styled index explains 78.38% of the sample variance. The vector error correction model is adapted to study the dynamics of cointegration of the sentiment-styled index and the logarithmic futures price. We use the GARCH-DCC model to illustrate the spillover effect between the sentiment-styled index and the Chinese futures market. We show that this investor sentiment-styled index does have the price discovery from the Granger causality and common factor weights and the hedging function from the Baba-Engle-Kraft-Kroner model empirically; furthermore, we use the curvature term of the sentiment-styled index to determine the multiple unit roots. More empirical results for the sentiment-styled index of the Chinese stock market, the sentiment-styled index of the CSI 300 index futures, and the return of the CSI 300 index futures market are studied in this paper.