We evaluate Eugene Fama’s claim that stock prices do not exhibit price bubbles. Based on US industry returns 1926-2014 and international sector returns 1985-2014, we present four findings: (1) Fama is correct in that a sharp price increase of an industry portfolio does not, on average, predict unusually low returns going forward; (2) such sharp price increases predict a substantially heightened probability of a crash; (3) attributes of the price run-up, including volatility, turnover, issuance, and the price path of the run-up can all help forecast an eventual crash and future returns; and (4) some of these characteristics can help investors earn superior returns by timing the bubble. Results hold similarly in US and international samples.
Building on a textbook description of associative memory (Kahana 2012), we present a model of choice in which a choice cues recall of similar past experiences. Memory shapes valuation and decision in two ways. First, recalled experiences form a norm, which serves as an initial anchor for valuation. Second, salient quality and price surprises relative to the norm lead to large adjustments in valuation. The model provides a unified account of many well documented choice puzzles including attribution and projection biases, inattention to hidden attributes, background contrast effects, and context-dependent willingness to pay. Unifying these puzzles on the basis of fundamental psychological processes – memory and attention to surprise – yields new predictions relative to existing accounts of the same phenomena.
We revisit La Porta's finding that returns on stocks with the most optimistic analyst long‐term earnings growth forecasts are lower than those on stocks with the most pessimistic forecasts. We document the joint dynamics of fundamentals, expectations, and returns of these portfolios, and explain the facts using a model of belief formation based on the representativeness heuristic. Analysts forecast fundamentals from observed earnings growth, but overreact to news by exaggerating the probability of states that have become more likely. We find support for the model's predictions. A quantitative estimation of the model accounts for the key patterns in the data.