Forecasting Chinese Stock Market with Extreme Values

Abstract

By decomposing stock returns with high-low extreme values, this paper investigates the predictability of Chinese stock market with a vector autoregressive model. Empirical studies, both in-sample and out-of-sample, performed on the Shanghai Stock Exchange Composite Index (SSEC) show that the SSEC is highly predictable in both statistical and economic sense. The findings obtained in this paper are based on history price information, which hints that the Chinese stock market might not be weakly efficient.

Keywords: Predictability, Chinese Stock Market, Extreme Values, Vector Autoregressive Model

JEL Classification: C32, C53, G14