I have been a close follower of behavioral economics research. This broad research is insightful and has caused me to think deeper about how to make better decisions. It has certainly reinforced my belief that using algorithms to make decisions is better than discretionary judgment. However, I have read a series of recent papers that have caused me to take a closer look at some of the core behavioral beliefs that have been established in this area. See the work of Dan Gal and Derek Rucker in the Journal of Consumer Psychology and the recent article in the Observation Section of https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3049660
A core behavioral economics foundation concerning risky choices is the concept that loss aversion has a strong influence on behavior. As simply put by the original authors, Dan Kahneman and Amos Tversky “Losses loom larger than gains”. The relative pain from a loss is greater than the gain from winners, and this asymmetric view is more than just a function of risk aversion.
The disposition effect, driven by loss aversion, states that investors will hold losers and sell winners. It has become a bedrock behavioral view about investors and is a core reason why many managers have stop-losses in models to counter-act this bias.
Now we have research that calls into question loss aversion as the core reason for these effects both from a theoretical and empirical point of view. There is clear evidence that contradicts loss aversion, but it has either been dismissed or ignored. Loss aversion is a description of behavior and not an explanation of behavior. This research is not offering an alternative to loss aversion but rather a commentary on its usefulness and explanatory power.
This new research may not completely change minds on the importance of loss aversion, but it does tell us that loss aversion is a subtle concept and should be employed with more care. How we evaluate decision outcomes is very sensitive to a reference point that is often the status quo. The set-up of the problem influences results that suggests that any general conclusions concerning loss aversion may be suspect.
According to Gal and Rucker,
“In general, it can be stated that the name “loss aversion” represents exceptional branding from the perspective of enhancing the idea’s intuitive appeal as everyone is essentially averse to losses (just as everyone is attracted to gains). This good branding might have led researchers to identify phenomena as being supportive of loss aversion even though the phenomena, while involving losses, do not involve comparisons of the impact of losses relative to equivalent gains. As discussed in the previous section, examples include the sunk cost effect, the disposition effect, and others.”
This research is subtle and may not change an investor’s decision-making, but it is a testimony to careful thinking about problems. The obvious may not always be correct and a simple narrative is not always applicable to a wide set of problems.
Do I worry about the pain from trading loses? Yes. Should I take extra steps to reduce downside “pain” more than what would be the case given my level risk aversion (specifically account for loss aversion)? I am less sure. Accepting conceptual uncertainty may make for decision-making.