Agent-based Experimental Study On Technical Rules And Abnormal Return

Along with the development of securities activity, investment has already became the phrase which in the life appears frequently. Each investor must carry on the analysis to the stock market firstly, especially the stock price. In the early, Random Walk Hypothesis had always occupied the mainstream position. If the securities price subjects to the random walk process, then the price change is unpredictable. Subsequently Efficient Market Hypothesis was proposed. If the stock market is weak-form efficient, the technical analysis will be the futile. While the recent studies investigate the predictability and profitability of the technical analysis by utilizing and extending some advanced econometric methods and statistics tool. So the argument has also continued.With the development of the method of Agent-based Computational Finance, the paper builds a artificial financial market in which has some kind of investors, in order to carry on some simple researches to predictability and profitability of the technical analysis. The objective of the study is to help investors to choose suitable investment strategy in turbulent stock market, and test the effectiveness of technical trading rules from the new perspective of Agent-based Computational Finance.By building a artificial financial market with several different investors, we simulate different investors'strategy and dynamic behavior, and carry on many group of different parameters of computational experiments, in order to study the absolute return of different technical trading rules. Through statistical analysis of simulation data, result proves that moving average rule and trading range break rule can gain excess returns at a certain extent, and conclusions support the effectiveness of technical analysis finally.