Study Of Herding And The Relationship With Market Efficiency

Investors’ non-rational behavior has been one of important study objects ofbehavioral financial. Investors’ non-rational behavior has not only increased thevolatility of stock market, but also affected the stock market efficiency. In this paper,we study the new test method of herd behavior, based on the research of theory andtest method of herd behavior, and testing the herd behavior in stock market,and thenstudy the relationship between market herding behavior and the market efficiency, themain content as follows:1. The Improve of herding test methods about stock market and the test ofherding behavior of inventors in Shenzhen stock market(1) We propose an improved model of herding, V-CCK test model, based onintegrating advantages of CCK, I-CCK, and T-CCK model, and considering themarket effects of heteroscedasticity;(2) We test the herding in the Shenzhen stock market, on the base of theimproved V-CCK model, and compare test results with existing models, the testresults of V-CCK model show that the proposed V-CCK model is better;(3) The static results show that there was a significant herding in the aggregateChinese stock market, and the herding behavior can be found both in up and downmarkets, but the herding in downside market is stronger than the herding in upsidemarket; the dynamic results show that the herding shows the weakened trend overtime in China’s Shenzhen stock market, and the herding in downside market is stillstronger than the herding in upside market.2. The related research of herding and effectiveness in Shenzhen stock marketWe propose indicators for herding(HE) and the indicators of effectiveness(EFF)based on V-CCK test model and MF-DMA(multi-fractal detrending moving average)method, and calculate the indicator series ofHEt andEEFt based on slidingwindow technique.(1) Correlation test shows that there is a positive correlation between HE andEFF; from the distribution maps of the two sequences, we find that the two sequenceshave a clear downward trend over time, the market becomes more and more effective;(2) Linear Granger causality test results show, HE is not linear Granger cause ofEFF, and EFF is linear Granger cause of HE; (3) The results of non-linear tests show, there exists nonlinear dynamic trendbetween HE and EFF, it shows that the conclusions by linear Granger causality test isnot accurate. Using VAR model, we filter out the linear relationship between HE andEFF based on the non-linear test, and then study the relationship between HE and EFFbased on non-linear Granger causality test. The results show that there are existingreciprocal non-linear Granger cause between HE and EFF, it means that the herdingaffects the level of market efficiency in Shenzhen stock market; in turn, the marketefficiency also affects the level of herding.