Noise Trading in a Laboratory Financial Market: A Maximum Likelihood Approach

Abstract

We study the extent to which, in a laboratory financial market, noise trading can stem from subjects' irrationality. We estimate a structural model of sequential trading by using experimental data. In the experiment, subjects receive private information on the value of an asset and trade it in sequence with a market maker. We find that, in the laboratory, the noise due to the irrational use of private information accounts for 35 percent of the decisions. When subjects act as noise traders, they abstain from trading 67 percent of the time. When they trade, the probability that they buy is significantly higher than the probability that they sell.