Identification of stock market forces in the system adaptation framework

Based on the system adaptation framework which has been proposed in our previous work, this paper focuses on the input selection of this framework to identify crucial market influential factors. We first carry out an empirical research to preselect influential factors from economic and sentimental aspects. The causal relationship between each of them and the internal residue of the market is then tested. Lastly, a multicollinearity test is applied to those factors that show significant causality to the internal residue of the market to exclude the redundant indicators. As the causal relationship plays an essential role in this method, both linear time-varying and nonlinear causality tests are employed based on the predictive ability of our framework. This double selection method is applied to the US and China stock markets, and it is shown to be efficient in identifying market influential factors. We also find that these influential factors are market-dependent and frequency-dependent. Some well-tested factors in the developed market and literature may not work in the emerging market.