ABSTRACT
There is evidence that people form nonstandard beliefs. Overconfidence reveals that beliefs and judgments are not always accurate. While the gambler’s fallacy, extrapolation, and base-rate neglect suggest that new information does not change beliefs according to Bayes’ rule. This chapter considers implications and modeling of such beliefs. It begins with three puzzling behaviors from business and finance — entrepreneurship, mergers, and investor trading — each of which is consistent with overconfidence. The following section introduces two non-Bayesian models of belief updating: a belief in the law of small numbers and base-rate neglect. These models can explain the evidence of gambler’s fallacy, extrapolative beliefs, and base-rate neglect. The chapter concludes with an application of extrapolation to explain asset price movements in financial markets (i.e., medium-term momentum, long-term reversal, excess volatility, and bubbles).
