Agent-based Experimental Study On Price Limits

Price Limits is a kind of regulatory measures used by regulators for curbing excessive fluctuations and maintaining the capital market stabilizing. But for it's great affect to the financial markets, it's not changed easily. Empirical study is traditionally used for Price Limits, but it's hard to remove other interference factors and only can be done after the implementation of the policy. We can't forecast the effects if the policy is carried out. Agent-based Computational Finance (ACF) can make up for lack of empirical analysis.In this paper, Redhat Linux 9 and Swarm 2.2 stable release are used as software platform; the study is based on the Artificial Stock Market (ASM) built by Professor Zhang Wei's Research Institute. The ASM is composed of heterogynous investors,BSV investors,RE investors and passive investors, and Price Limits is added in the paper. The author aim to find out the Price Limits'effect on the price of stock and try to explain theories and hypotheses about the policy. If the simulation can be used in actual policy-making, it can be very meaningful.The main conclusions are as follows:(1) The smaller the Price Limits are, the smaller the variance of the stock price, the weaker the volatility clustering, the stronger the correlation between price .This means Price Limits is efficacious, and prove the Price Stabilization Effect.(2) The smaller the Price Limits are, the larger the mean of traded volume. That means the turnover increases, and mobility is enhanced, and the Price Limits enhances the stock liquidity.(3)Whether Price Limits is added, leverage effect exists. The smaller the Price Limits are, the weaker the leverage effect. Bad news has a positive impact on Logarithm of conditional variance, and good news has a negative impact on Logarithm of conditional variance.