Tobin tax and the the information efficiency of financial markets: A structural estimation

We study the effect of a transaction tax on the trading of a security. In our model there are informed traders, who receive private information on the value of a security, and noise traders who trade for liquidity reasons. Through a structural estimation, we estimate the proportion of informed and noise traders. Moreover we estimate the precision of the private information traders receive and the price elasticity of noise traders demand. We then introduce in the model a transaction tax. Theoretically the tax has a positive effect, since it reduces "bad volatility" due to noise traders, and a negative effect, since it reduces the incentive of informed traders to trade using their own information. Indeed, the tax may even lead to an information cascade in the market, in which no information at all is aggregated. To evaluate the importance of these effects, we simulate the model with transaction tax, using the parameters obtained in our structural estimation. We then compare the simulated price with the one we observe in the market and with the one we would observe if the market were perfectly efficient. Our preliminary results show that the negative effect prevails on the positive one.