The Limits of Noise Trading: An Experimental Analysis

Abstract

We report the results of a laboratory market experiment that allows us to determine not only how noise traders fare in a competitive asset market with other traders, but also how the equilibrium changes if a securities transaction tax ("Tobin tax") is imposed. We find that noise traders lose money on average: they do not engage in extensive liquidity provision, and their attempt to make money by trend chasing is unsuccessful as they lose most in securities whose prices experience large moves. Noise traders adversely affect the informational efficiency of the market by driving prices away from fundamental values. With a securities transaction tax, noise traders submit fewer orders and lose less money in those securities that exhibit large price movements. The tax is associated with a decrease in market trading volume, but informational efficiency remains essentially unchanged and liquidity (as measured by the price impact of trades) actually improves. We find no significant effect, however, on market volatility, suggesting that at least this rationale for a securities transaction tax is not supported by our data.

Keywords: behavioral finance, noise, limits to arbitrage, day trading, experiments, Tobin tax, securities transactions tax, informational efficiency, liquidity, volatility

JEL Classification: G10, C92