Overreaction in Expectations under Signal Extraction: Experimental Evidence

We experimentally evaluate three behavioral models of expectation formation that predict overreaction to new information: overconfidence in private signals, misperceptions about the persistence of the data-generating process (DGP), and diagnostic expectations. In our main experiment, participants repeatedly forecast the contemporaneous and one-step-ahead values of a random variable. They are incentivized for accuracy, informed of the exact DGP and its past history, and provided with noisy signals about the unobserved contemporaneous value. One treatment features a persistent AR(1) process, while another has no persistence. We also report on an experiment with no noisy signals. At the individual level, we find systematic overreaction even when the DGP is not persistent and regardless of whether a signal-extraction problem is present. By contrast, consensus (mean) forecasts exhibit underreaciton, consistent with evidence from other studies. Overall, our results indicate that misperceptions about persistence provide the most compelling explanation for the observed patterns of expectation formation.