A Study On The Effect Of Transaction Tax On A Social Network Based Artificial Stock Market

As the most convenient and stable channel for enterprise financing, the financial market is playing an increasingly important role. However, the frequent financial crises remind people to re-examine the safety and regulation of financial markets. In recent years, a growing number of economists have been involved in areas of research like policy making and risk prevention, trying to find out an effective way to make up for the lack of financial regulation. Among those, financial transaction tax is considered to be an effective tool to punish speculation, so as to make the market more stable.A study on the effects of a transaction tax on one closed market is presented in this paper, using a continuous double auction market based on heterogeneous agents. The micro structure of the market is composed of three kinds of traders:random traders, chartist traders and fundamentalist traders. It is assumed that the resources allocated to them are limited and follow a Pareto Law. Agents in the market put their orders by considering of their expectations of the spot return and imitating their neighbors’expectation. The well known stylized facts such as volatility clustering and fat tails are analyzed in this paper. Different from previous models, in this paper, a social network of different agents is constructed. Agents then communicate and learn from each other based on this social network. In this way, the model is much closer to the real market. Furthermore, this paper try to figure out whether and how the introduction of transaction tax can affect the financial market’s volatility. Computational experiments are performed and it is found that tax increased volatility, while social network weakens this effect, making the market more stable.