An Indigenous Analysis of the Gradual Information Diffusion Model

Abstract

The main task of this work is to find out-of-sample evidence of the proposed predictions of Hong and Stein (1999) in the Chinese stock market – the Shang-Hai Stock Exchange (SSE) and the Shen-Zhen Stock Exchange (SZSE). Specifically, in this paper I explicitly test the three sets of implications derived from the gradual information flow model of Hong and Stein (1999). I found some supporting evidence for the gradual information flow model of Hong and Stein (1999). Using firm size as a proxy for information diffusion rate, I confirmed that momentum profits follow a term structure consistent with the prediction of the model. However, I did not find ubiquitous evidence of firms whose information travels slower should exhibit greater long-term reversal, following more pronounced overshooting. I also arrived at a contradictory observation in which stocks with longer trading horizon show virtually muted positive autocorrelations.