Rumor and prediction: Making sense (but Losing Dollars) in the stock market

It is proposed that by making sense of unpredictable price movements, rumors spawn anti-regressive predictions and adversely affect trading decisions despite investor denigration of rumors. Two experimental stock market simulations investigated these ideas. Subjects were presented with news (Study 1) and published and unpublished rumors (Study 2) while participating in a computerized investment game. As compared with controls, all manipulations caused departures from a profitable buy-low-sell-high (tracking) trading strategy despite strong differences in rated credibility between types of information source. Subjects claimed that rumor sources were non-credible and that they were not influenced by rumors in trading decisions; nevertheless they traded on rumors as though they were news. Results integrate rumor theory with the psychology of prediction, extend the corpus of rumor literature to behavior, and highlight the sense-making function of rumor in situations filled with uncertainty.