Measurement and prediction on periodicity of Shanghai Composite Index fluctuation

摘要

Forecasting stock market index is significant for government macro-manipulation, investing risk aversion, and corporate operating profits. This paper applies three methods, which are linear trend method, H-P filtering method and Band2-Pass filtering method, to analyze periodicity and predictability of its fluctuation, and contrast predicted value with actual value, based on the data of Shanghai Composite Index from 2007–2009. We find that each of these three methods can give a prediction on the Shanghai Composite Index within a certain range. In terms of accuracy, H-P filtering method and Band2-Pass filtering method have a higher accuracy with a better scope of application and reveal the predictability of stock index.