Fear of disappointment can make us fearful of new information, says TSE researcher Marianne Andries, and this has important implications for our understanding of investor behavior. In a new working paper, she proposes a theory of inattention based on the idea of information aversion. Sometimes, her results suggest, withholding information may serve us better than transparency.
When information costs are due to technological limitations, finding ways to provide more information can greatly improve market efficiency. But Marianne’s research suggests such policies are not always desirable. “Experiencing the roller coaster of life can be stressful,” she says. “A natural way to avoid this stress is to close your eyes for the ride.”
This idea that people might want to stay away from information draws a very different picture to that provided by standard models of inattention. In her recent working paper, titled ‘Information Aversion’, Marianne’s theory of inattention is solely based on fear of information, rather than on cognitive limitations or the external costs of acquiring information. In her model, disappointment-averse agents optimally decide to stay away from some sources of information. This framework has rich implications reflecting key observations on information and risk-taking behavior in the lab and in the field, among participants in financial markets.
Together with co-author Valentin Haddad (UCLA), Marianne starts by characterizing the endogenous information costs implied by disappointment aversion, and finds them to differ fundamentally from both the cognitive constraints and the exogenous costs commonly used in the economics literature on inattention.
Under disappointment aversion, agents inflate the probabilities of outcomes that disappoint. As information arrives, each piece of news creates scope for disappointment. The agent therefore prefers to receive less fragmented information and observe simultaneous bundles of news in which good news can cancel out bad, disappointing, news. Such information aversion is a direct consequence of the agent’s attitude towards risk.
To analyze how agents cope with their fear of information, Marianne looks at how the frequency of information observations impacts the valuation of risky lotteries. She finds her model justifies the experimental evidence that shows agents’ valuations of risky assets decrease when they are given more frequent and more detailed information.
She also studies how agents balance the cost of paying attention to the economic environment with the benefits of making better informed decisions. In an illustrative example, Marianne considers an information-averse investor who manages his or her wealth to finance consumption over time, allocating savings between a risk-free asset, and a risky asset yielding higher average returns.
Marianne shows that this investor optimally decides to observe the value of the risky portfolio at equally spaced points in time. In between observations, the investor consumes deterministically from a risk-less portfolio, and allocates any remaining wealth to the risky asset. The marginal cost of infrequent observation is due to the loss in expected returns when more wealth is placed in the risk-free asset. Unique to Marianne’s model, the marginal benefit comes from a relief from the stress of receiving information.
More risk-averse investors are also more inattentive, Marianne finds. Attention decreases in periods of high volatility, even when higher expected returns keep the difference between risk-adjusted returns and the risk-free rate constant. This prediction reflects an increase in the marginal cost of information as risk increases and is in line with recent empirical evidence.
How can suppliers of information best serve the needs of information-averse investors? Marianne’s model provides the basis for a theory of optimal information. Her results suggest financial institutions can foster more investment by providing “distress” signals following sharp market downturns. While an in-depth treatment of this area is left for future research, she outlines a few implications.
One way to help information-averse agents is to lump news together in bundles delivered at precise points in time. Such behavior is consistent with company disclosure policies organized around scheduled earnings announcements. Similarly, monetary policy and other macroeconomic announcements, such as employment numbers or quarterly growth, are disclosed at precise points in time, and mostly scheduled in advance.
Agents do not want to receive information too often. However, when they do observe information, they want it to be as precise and “transparent” as possible. In Marianne’s framework, it can be beneficial if suppliers sometimes refrain from releasing information; but the release of partial or distorted information is not beneficial.
However, Marianne warns that this form of information withholding generates asymmetric information, and agency problems are likely to arise (for example, between an investor and her wealth manager). To account for information aversion, optimal compensation contracts need to provide the necessary incentives, while minimizing the information needed to enforce them.