Can industries predict the stock market in China?

Abstract: In the thesis we investigate the return predictabilities of ten industry portfolios on the Chinese stock market as well as intra and inter industry predictabilities. We find a number of industry returns, including energy, telecommunication services and financials are able to predict the market up to three days. Gradual information diffusion theory is applied to understand our results. An industry's ability to predict the market is not correlated with its propensity to forecast industrial production growth. To detect inter and intra industry prediction power and linkages, autoregressions and vector autoregressions of the industry portfolios are further conducted.