05.31.18

Probability Weighting and Household Portfolio Choice: Empirical Evidence

Research Report
Insights Report

Conventional economic models assume that investors confronted with risky choices maximize expected utility; yet in the real world, people are prone to making predictable errors.

Summary

Prior research has confirmed two anomalies in household portfolio choice: non-participation in the stock market and under-diversification among those who do participate. This study shows how a common behavioral bias known as probability weighting – i.e., overweighting small probabilities and underweighting large probabilities – influences household investment decisions. The results have important implications for how financial products are designed and communicated to savers and retirees.

Key Insights

  • Probability weighting explains financial decisions that seemingly violate utility-maximizing behavior.
  • Most people overweight small probabilities and underweight higher probabilities to some degree.
  • Investors who overweight the probability of realizing extreme gains through equity exposure may prefer to hold a few individual stocks that could rise rapidly, versus investing in a broadly diversified portfolio.
  • Risk-averse investors who underweight the likelihood of generating moderate positive returns through equity investments might prefer to avoid stocks altogether.

Portfolio under-diversification, despite offering poor risk-adjusted returns, can appear attractive to some investors.

Methodology

To elicit individuals' probability weighting preferences, the researchers designed an Internet survey module and fielded it in a nationally representative sample of several thousand participants in the American Life Panel. The module elicited certainty equivalents for a series of binary lotteries. The probabilities of winning the lotteries varied from small to large, allowing the researchers to measure each respondent's probability weighting behavior. All respondents received a fixed participation fee and had the opportunity to obtain real monetary incentives based on their choices.

Authors

Stephen G. Dimmock

Nanyang Technological University

Roy Kouwenberg

Mahidol University and Erasmus University Rotterdam

Olivia S. Mitchell

University of Pennsylvania

Kim Peijnenburg

HEC Paris and CEPR

Sign up for the TIAA Institute newsletter

Get the latest research and insights straight in your inbox

We are sorry.

The service that receives your request is unavailable at the moment. Please try again.

Related Publications