Study On Limited Arbitrage Of Chinese Equity Investment Funds Based On Agent-based Computational Method

As one of the two cornerstones of behavior finance theory, limited arbitrage theory can reasonably explain the anomalies in the secunrity market which asset price deviate from its base value in a long term. Correctly knowning the reasons of arbitrage risk and limits will help us to optimize the investment strategy and improve market efficiency. At present, most of the studies about the limits of arbitrage concentrate on introducing theory and analysis based on mathematical models.This paper summarizes the theories related to limits of arbitrage. Based on the DSSW model, it build an agency model of limited arbitrage, contains three kinds of traders:arbitrageur, noise trader, investor who invest on the arbitrageur. By using methods of computational experiment finance and based on the Netlogo platform, this paper build the Artificial Stock Market and simulate the optimal strategies of different investors and the trading mechanism in the market, do analysis on the behavior of the investment fund, to find the reasons of limits of arbitrage from a microscopic view. First, it is showed thatBy exerting experiments, we find many meaningful conclution. For example, the noise traders can live for long periods in the market, and gain the stable revenue. And there are some characteristics of the Performance-Based Aarbitrage under the agent model, including:First, it is showed that arbitrageur gain a less revenue, but the noise gain more. Second, the asset price diverge more far from the fundamental value, and shows more volatility. Third, the more the shock of the noise trader, the invalider the asset price, at this time, The action of arbitrage become effective. Fourth, the higher sensitivity of the investors to the arbitrages’ performance, the more constraint the arbitrage have, the more difficult the process which the asset price revert to the fundamental value.