Determinant Technical Indicators of Different Stock Index: by a Data-driven Approach

Jun Qian ； Qiu-Qin He

《International Journal of Intelligent Technologies and Applied Statistics》 7卷3期 (2014/09) Pp. 229-240

https://doi.org/10.6148/IJITAS.2014.0703.03

並列摘要

This study investigates the determinant technical indicators which affect the trend of stock index by a data-driven approach. Applying the research results got by Li and Racine [10], we find the indicator mavg, adx, MA_30 and CMF are relevant to both SSEC and NASDAQ while the other relevant variables are different. The difference indicates that in the short term the determinant moving average indicator's length of NASDAQ index is longer than Shanghai index's. And investors in China care about the tendency while in America they care about buy and sell points. To test the validity, a varying-coefficient partially linear model is constructed. We find that the model which contains all nonlinear indicators has a better fitting effect than the benchmark model. It means the result of variable selection is effective. Then we construct path model and the result indicates that the fitting effect of the path model is not better than the benchmark model's. This evidence supports that stock technical indicators affect the trend and fluctuation of stock price by joint effort.

並列關鍵字

Variable selection ； Stock technical indicator ； Data-driven approach ； Nonparametric econometrics ； Path model