Return Predictability, Expectations, and Investment: Experimental Evidence

Autor: Milo Bianchi, Marianne Andries, Karen K. Huynh, Sébastien Pouget
Rok vydání: 2020
Předmět:
Zdroj: SSRN Electronic Journal.
ISSN: 1556-5068
DOI: 10.2139/ssrn.3710466
Popis: We design an experiment to study how investors form their expectations and make risky investments under different market conditions. Together with the past realizations of a risky asset, our subjects observe a signal a that, in some rounds, helps predict future returns. When subjects perceive a as useless, they irrationally extrapolate from recent return realizations. When they perceive a as useful, instead, they correctly incorporate it and extrapolate much less. We interpret those findings in a forecast model in which subjects have imperfect ability to detect predictability and face uncertainty about the correlation between signal a and future returns. We also find that the level of risky investment and its elasticity to forecasts are larger when a is perceived as useful, suggesting that subjects recognize that predictability in our setting reduces risk. Yet, the elasticity of investments to forecasts remains low -- a puzzle relative to their high risky investment.
Databáze: OpenAIRE