| Purpose of this dissertation is to identify features of momentum effect and to discover reasons why it takes effect in China's stock market through empirical analysis.The dissertation starts with a review of development from traditional finance theory to behavioral finance theory. Successively the review covers Efficient Market Hypothesis, Capital Asset Pricing Model, abnormal phenomena discovered continuously, proposal of new capital asset pricing models, explanations to these abnormal phenomena by the academia, and finally behavioral finance theory.Then with relevant data from CSMAR of 1140 A-share stocks of all listed companies in Shenzhen or Shanghai stock market, the dissertation presents a detailed analysis to momentum effect in China's stock market as well as exploration in the following two folds. The first fold is analysis on income characters of momentum strategy under the circumstance of intercross combination of developing period and holding period with methods of overlapping and non-overlapping sampling respectively. Analysis also unfolds income characters of equally weighted deciles portfolios when developing period coincides with holding period with method of overlapping sampling. Both overlapping and non-overlapping sampling analyses indicate obvious effect of momentum effect in China's stock market.In the combination of 25 periods with overlapping sampling, the highest profit of the zero-cost momentum portfolio appears in 1x1 strategy with an annual return rate of 11.72%. The conclusion here is not completely in line with that of Jegadeesh & Titman(1993),Weimen Liu.Norman Strong, Xinzhong Xu(1999) and Rouwenhorst (1998) , who after study on sample data from America, Britain and multi-national market, drew identical conclusion that 12 3 strategy is the most profitable with return rates of 16.9%, 23.3%, 17.5% respectively. Compared with their conclusions, the return rate of momentumportfolio in China's stock market is apparently lower.In order to verify the above conclusion, and to probe further into the relations between momentum effect and its terms, the dissertation goes on with the rate of equally weighted deciles portfolios under the condition of identical sorting and holding period. If both periods of sorting and holding are one month, the results show that though not strictly increasing continually, it goes up generally. The average annual rate of benefit of combination of "Winners-Losers" is 11.72%, being distinctly positive under 10%. If the two periods are three month, the results remain the same while the average annual benefit rate of the combination is 8.66% and again, is outstandingly positive under 10%. When each of the periods is six month, it shows that the tendency is gradually increasing, but begins to decrease after the eighth combination. The average annual benefit rate of "Winners-Losers" is 8.17% with a notable plus under 5%. However, when the periods are nine months respectively, a reverse tendency starts to show. The above-mentioned rate is -6.1%, being negative under 10%. And when the periods are twelve months each, the reverse tendency becomes distinct. The above-mentioned rate is -12.83% and negative obviously under 1%.All in all, the dissertation draws its conclusion after the above discussion as follows: Momentum effect phenomena does take effect in China's stock market, but it lasts only for half a year or so, shorter than those in western developed markets. After that, benefit reversal phenomena begin to show themselves gradually.At last, the dissertation inquires into reasons of momentum effect phenomena in China's stock market from perspectives of risk, scale, B/M, proportion of liquid shares and turn-over rate. In the end, all the above factors are found to help explain why momentum effect takes effect in China's stock market.
|
PDF Full Text Request