Weak-form Efficiency Analysis For Hong Kong Stock Market Based On Hang Seng Index

The aim of this paper is to test the ongoing weak-form efficiency in Hong Kong stock market. According to EMH, the efficiency of stock market can be tested by the prediction of returns. Taking study background into consideration, we chose the weekly closing price of Hang Seng Index from 21/2/1986 to 30/4/2014, and also to make the period long enough for building time series models. We did Auto-correlation test and Unit Root test for the whole sample data series. However the result of the two tests leads to totally different conclusion. This result raises a conjecture that maybe the state of efficiency is changing over the whole period. In order to study the evolving dynamic change of the market efficiency, we applied the rolling sample(rolling window is 100) to fit the AR(1)-GARCH(1,1) model. From the graph of coefficient of AR term, we can conclude that Hong Kong stock market was not so efficient in the first period of time, and then became efficient till now. With regard to imprecision of rolling AR(1)-GARCH(1,1), we applied Kalman Filter method to estimate the time-varying coefficient. Based on the result of Kalman Filter estimation, we drawn the graph of coefficient of AR term. From the graph we can know that Hong Kong market was not efficient in the time period 21/2/1986 to 14/12/1990 and became weak-form efficient from 14/12/1990 to 30/4/2014, which is similar to the result of rolling sample AR(1)-GARCH(1,1).