Working Paper: CEPR ID: DP10868
Authors: Philippe Jehiel
Abstract: Prospective investors of new projects consider the returns of implemented projects with similar (observed) attributes and invest if the empirical mean return exceeds the cost. The steady states of such economies result in suboptimal investment decisions due to the selection bias in the sampling procedure. Assuming higher attributes are associated with higher returns, there is systematic overinvestment as compared with the Bayesian benchmark, thereby illustrating that selection bias may explain entrepreneurial overconfidence. Various extensions are considered to illustrate the negative externality that rational investors exert on other investors, the effect of correlation between the attributes considered by various investors, and how trading may be affected by the sampling procedure.
Keywords: investment strategy; overconfidence; selection bias
JEL Codes: C70; D82; D83; D84
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| selection bias (C24) | perceived expected returns (G17) |
| perceived expected returns (G17) | overconfidence (G41) |
| overconfidence (G41) | excessive investment (E22) |
| selection bias (C24) | overconfidence (G41) |
| rational investors (G40) | sampling process of non-rational investors (G41) |
| attribute correlation (C10) | investment behavior (G11) |