The Determinants And Policy Effect Of Stock Index

With the rapid development of capital market and computer technology, there is a dramatic rise in trade data. Against this background, This study proposes a data-driven approach to investigate the dominant factors （mainly the technical indicators） affecting the trend of different stock index. Through nonparametric method in blended data we eliminate variables and find the key technical variation that affect the tendency of NASDAQ index and Shanghai stock index. Then we construct complete model and path model to analyse the policy effect. At last, we construct forecast model and design a trading strategy to find the advantages and disadvantages of the model framework. The empirical results show that:First, most of the determinant technical indicators are the same but the mechanism is different. The same determinant variables in SSEC and NASDAQ are mavg, adx, MA₃0and CMF. In terms of the different determinant variables, the variables MA₁0and CMO are relevant to SSEC but not to NASDAQ, and the variables MA₅and SAR are relevant to NASDAQ but not to SSEC. With regard to the property of the variables, mavg and CMO are linear in SSEC, and MA₅is linear in NASDAQ; there exist the same nonlinear variables such as adx, MA₃0, CMF in SSEC and NASDAQ, and different nonlinear variables like mavg and SAR in NASDAQ, and MA₁0in SSEC. Considering the mechanism, we find that the linear part of the model affects the tendency of the stock index and the nonlinear part affects the volatility of the stock index. It is also known by calculation that the most significant change on the volatility of NASDAQ is caused by SAR but that of SSEC is MA₃0; the most significant positive impact on the tendency of NASDAQ is induced by adx but that of SSEC is by CMF.By comparing impact of the same variable on the different stock index, we find that CMF and adx have the positive impact on the SSEC and NASDAQ index, while MA₃0has the negative impact on the SSEC and NASDAQ index. In other word, CMF, adx and MA₃0have the same positive and negative effects on the SSEC and NASDAQ index. We also find that the impact strength of the three indicators on the SSEC and NASDAQ is different. CMF has stronger impact strength on SSEC than NASDAQ, while adx has stronger impact strength on NASDAQ than SSEC. Therefore, it is concluded that the differences show the Shanghai Stock Index is influenced by cash flow and NASDAQ index is influenced by the Directional Index.These results imply that the investment differences of the Chinese investors’concerning tendency and the American investors’concerning buy&sell points. In China, the investor’s focusing on trends more easily lead to chasing the market; while in USA, the investor’s focusing on volatility more easily lead to a phenomenon that when the river rises, the boat floats high.At last, we design a trading strategy and find the profit of the strategy is more than the index and the max drawdown of the strategy is less than the index. It means that the performance of the trategy is better than the index. However, when there is a trend reversal the prediction is not good enough. The reason is only the linear part be extrapolated and the nonlinear part is the same as last one.