An experimental asset market is used to test the effect of news concerning the underlying value of an asset on its trading price. Statistical support is found for the hypothesis that investors underreact to news on asset valuation. The results are consistent with the viewpoint that price and valuation history have a significant effect on trader behavior. Two sets of experiments involve a single asset with a single payout at the end of the experiment. Payout conditions are updated at the midpoint of the market trading period. The two sets of experiments have different payout expectations during the first half of the experiment but the payout expectations are identical after the midpoint. Although the expected payouts are identical after the midpoint of the market trading periods, the trading prices for the two sets of experiments differ significantly even after their payout expectations coincide. This provides support for underreaction and indicates that decision makers tend to “anchor” their price expectations to pre-existing prices and/or valuations.
Read this paper at Chapman.edu.