Expectations driven price volatility in an experimental coweb economy

The purpose of this paper is to investigate expectation formation in a controlled experimental environment. Subjects are asked to predict next periods price in an unknown, unstable nonlinear cobweb economy with expectational feedback. Subjects have no information about underlying market equilibrium equations, but can learn by observing past price realizations and predictions. We conduct both single-agent and multi-agent experiments. In the multi-agent experiments forecasting errors were lower and average earnings were higher than in the single-agent experiments. In both the single and multi-agent experiments in most cases convergence in the mean to rational expectations (RE) occurs. Typically, the null hypothesis that the sample mean of realized prices is different from the RE steady state cannot be rejected. However, prices do not converge to the RE steady state but keep fluctuating irregularly. In fact, almost all experiments exhibit strongly significant excess price volatility driven by endogenous expectations.