This work experimentally examines forecasting and trading behavior. Subjects play the role of both analyst and shareholder over the course of experiments consisting of a series of repeated games with or absent conflicts of interest. In a stylized trading setting, I test whether standard equilibrium, normative behavior, or limited strategic reasoning best predicts behavior. In the presence of conflicts of interest a substantial proportion of subjects’ behavior appears non-skeptical in the role of shareholder, though the same subject is deceptive in the role of analyst. Absent conflicts of interest, subjects behavior in the role of shareholder is nearer a best response to the same subject’s behavior as analyst. The results are consistent with limited strategic reasoning and suggest that simply disclosing conflicts of interest does not evoke skepticism of forecasting, nor does the elimination of conflicts of interest in itself induce honesty.
Shields, T.W. (2010). Do analysts tell the truth? Do shareholders listen? An experimental study of analysts' forecasts and shareholder reaction. ESI Working Paper 10-05. Retrieved from http://digitalcommons.chapman.edu/esi_working_papers/126