Behavioral Finance Bias Deep Dive: Loss Aversion

This begins a series of deep dive articles on the most important behavioral biases that we confront as investors. Now before you dismiss these articles with a cavalier “I already know this stuff,” I want to point out that much like Shakespeare, most people who quote behavioral finance and its bias precepts have never actually read the source material. My point in this series is to summarize the literature for you. Including the naysaying research which is almost never quoted. My ultimate goal is to have this series culminate in a Theory of Behavioral Finance.

A Helpful Mnemonic Device: LOCHAARM

Though there are over a hundred biases formally identified I believe that there are less than ten that deserve the attention of investment professionals. It isn’t that the others are unimportant, it is just that many of them are corollaries of the major biases. These major biases are:

  • Loss aversion
  • Overconfidence
  • Confirmation
  • Herding
  • Anchoring
  • Availability
  • Representativeness
  • Mental accounting

A helpful mnemonic device for remembering these biases is LOC HAARM, brain lock that harms investment performance. First up this week is: loss aversion.

Loss Aversion: Origins

Loss aversion[1] was first formally described in 1979 by Daniel Kahneman and Amos Tversky in their seminal paper, “Prospect Theory: An Analysis of Decision Under Risk.”[2] It may surprise you, though, that they did not use the “loss aversion” nomenclature in this, their most famous work. It was not until an address given at a meeting of the American Psychological Association that they used this most well-known of behavioral finance terms. These proceedings were captured in a work called “Choices, Values, and Frames.”[3]

“Prospect Theory” was the culmination of many different thinkers’ and researchers’ work over the centuries in trying to understand and model human decision-making. Kahneman and Tversky specifically sought to criticize the prevailing theory at the time: expected utility theory. Chances are that if you studied economics you learned all about expected utility, as developed by thinkers like Daniel Bernoulli in the 18th century up through to von Neumann and Oskar Morgenstern in the 20th century.

Loss Aversion: The Findings

The most widely quoted finding of research done on loss aversion, sometimes known as myopic loss aversion, is that people generally experience the pain of loss more severely than they do the pleasure of gain. Simply stated, most of us would prefer to avoid a loss of €100 than we would to gain €100. Consequently, most people try and avoid personal harm or loss, first, and to seek pleasure, second.

Here is a simple illustration of loss aversion: A game pays $1.10 if a fair coin comes up heads and costs $1 if it comes up tails. You can participate in this game as often as you wish, as long as you hold up a dollar prior to each flip to ensure you are good for the potential loss.

Would you participate? In answering this question, it might help to know that the expected return on each round is $0.05 or 5% of your one dollar “investment.”[4] Does this change your decision regarding whether or not to participate?

Of course, you should play this game forever, as the expected return in each round is $0.05 and you will become hugely wealthy over time. But it turns out that very few of us would participate. This is because most of us experience far greater pain from losing one dollar than is the joy of winning $1.10. What is more, the fact that the game can be played multiple times seems to have a limited impact on the decision. It appears we view each round as a separate event.

How large must the return to heads be in order to attract the typical individual?  Research has shown that the amount is somewhere around $2, which means that the ratio of gain to loss is approximately 2-to-1. That is, short-term loss aversion is twice as strong a feeling as the positive feelings regarding a comparable sized gain. These results are verified as proven by much research with humans, as well as, and very importantly in other animal subjects.

This 2:1 loss aversion ratio has a pervasive effect in the world of investing. It appears that we are myopic even when the investment horizon is supposed to be long term, such as when saving for retirement. Rather than viewing such situations as 20, 30, or 40-year time periods, we view them as a series of days, months, quarters, or years. We evaluate performance in each short time period and apply 2-to-1 loss aversion to our gains and losses in each period. Thus, if an investment loses money during our short measurement period, we may decide to get out of that investment and into another. This is true even if high short-term volatility is offset by high long-term returns.

The choice of time period over which to apply the 2:1 loss aversion is arbitrary. There are no guidelines for deciding on this time horizon and so different investors choose different lengths, with no rhyme or reason. While there are analyses that can lead to buying and selling a particular investment, there is rarely a time horizon that makes logical sense as part of this analysis.

Almost all of the time horizons chosen in investment analysis are arbitrary and the length decided upon is based on emotion rather than on logic. This means that many of us are unable to reap the benefits of time diversification, where the volatility of the investment diminishes as period to period movements offset one another over time.

Loss Aversion: An Obscure Wrinkle

Sadly, most of the investment community’s understanding of myopic loss aversion stops here, with the memorized heuristic that the pain experienced by loss is approximately twice as great as the pleasure of gain. Yet, Kahneman and Tversky’s loss aversion is quite a bit more interesting.

Specifically, they also point out that their research, along with that of others, shows that people tend to show risk taking behavior under certain circumstances, too, and not just loss aversion! An example straight from “Choices, Values, and Frames” is that when given a choice between, a) an 85% chance to lose $1,000 and with a 15% chance to lose nothing, versus, b) a sure loss of $800, that a large majority of people choose option a).

What is going on here? Kahneman and Tversky point out that traditional expected utility is an incomplete description of human decision-making and outline three important characteristics of Prospect Theory to better describe decision making than in expected utility theory.

The following chart, Figure 3 taken directly from Prospect Theory[5] helps to explain the schizophrenic nature of people, who, under some circumstances are loss averse – the 2:1 ratio – but in other very similar situations they are risk-takers!!

Relativity

First, people do not evaluate choices that stand to affect small amounts of money relative to their total wealth, but instead in terms of gains, losses, and importantly, the maintenance of their status quo. Also, people typically evaluate risky decisions as changes in our wealth right now, as opposed to changes to our total wealth forever. By contrast in traditional economics expected utility theory the assumption is that people evaluate all decisions, including for inconsequential amounts, relative to their total wealth. The evidence says that people just do not operate this way.

Concavity and Convexity

Second, both gains and losses are considered relatively, not absolutely. In other words, a gain of £100 is more meaningful to a person who begins with £50 than it is for someone beginning with £1,000. In expected utility theory the assumption is that there is an indifference with regard to absolute gains. In other words, if we map out the perceived value of gains they are concave, with the same absolute increase in wealth, say £100, eventually peaking out. This is what the figure above demonstrates.

However, traditional expected utility theory also says that a loss of £100 for someone beginning with £200 is more meaningful than the same loss applied to a beginning value of £1,000. In other words, concavity is assumed in losses as well. However, as we said earlier, research has shown that risk taking actually increases as the chance for losses increases. This means that the value function for losses is actually convex.

Steepness of Losses

Third, and this is the very definition of loss aversion, while gains and losses affect people in a non-linear, either concave or convex way, losses are experienced as having a much greater effect on wealth than the actual changes in wealth would imply.

Anxiety about loss, of course, finds a rich partner in investing which has many possible sources of anxiety. These include: volatile securities markets, changes in monetary policy, leverage that may exacerbate losses, fiscal policy gridlock, short-termism, bewildered clients, large legal liability, et al. Put another way, many investors are constantly on the prowl for things that may lose them money. Grrrrowl!

Loss aversion is something that has been understood by investors for decades and well before its specific definition by Kahneman and Tversky. For example, anecdotally, there are investment homilies, such as, “stocks climb a wall of worry” that illustrate an understanding that loss aversion plays a role in the functioning of financial markets. However, Kahneman and Tversky’s loss aversion goes beyond a mere sociological jargon exercise (i.e. giving a formal name to something of which we are all familiar) to outline how gains and risks are treated differently, and unexpectedly by most people.

Loss Aversion: Criticism

Despite the storied success of Prospect Theory and its loss aversion, starting in 2006 critics began questioning its sweeping conclusions. For example, David Gal states that, “The absence of an accepted psychological theory to account for loss aversion has led to a paradoxical situation: loss aversion is cited as an explanation for phenomena associated with loss/gain tradeoffs (e.g., the endowment effect, status-quo bias, risky bet premium) and, circuitously, the same phenomena are cited as evidence for the existence of loss aversion.”[6] Gal offers up his own theory to explain loss aversion: people tend to prefer stability and the status quo. Later work by this same researcher finds that:

“[T]he studies of low-stakes wagers that supposedly establish loss aversion typically frame the choice to take the bet as a change to the status quo. As a result, researchers have confused simple inertia, the tendency to stick with the status quo in the absence of a meaningful incentive to change, with loss aversion.

“Indeed, when decisions about losses and gains are decoupled from a choice between change and the status quo, there is no evidence for loss aversion. For example, asked to select between receiving $0 or accepting a bet with 50% odds of either losing or winning $10, about half the test subjects choose to take the bet. In other words, if the status quo option is presented as an active choice — to “receive $0” rather than “not accept the bet” — the preference for safety vanishes.”[7]

Additional critics of loss aversion have found, for example, that “potential losses never loomed larger than gains for low magnitudes…”[8] In the preceding study they were using the results of actual games involving gambles where the monetary stakes were low, rather than the rarefied environment of a research lab. Furthermore, the researchers also found that, “loss-aversion disappears even for higher monetary values, if contextually an even larger anchor is provided [Note: I will be discussing anchoring in a future post]. The results imply that Prospect Theory’s value function is contextually dependent on magnitudes.”

Loss Aversion: Our Industry Worships It

As a brief reminder it is my opinion that most of the ecosystem of the investment industry is structured to kowtow unnecessarily before the loss aversion altar. That is, I recognize that it is sometimes no fun to stand before our usually overly emotional clients who scream “SELL!”

Our clients’ distaste for losses creates an anxiety for the investment management firm that the customer may leave. The unfortunate side effect is an industry that has become short-term in its orientation, and with ever greater numbers of closet indexing asset managers. Sadly, this means that there is a diminished reputation of active management, and most importantly, much lower returns for average investors over the long-term.

Here are some of the examples of kowtowing to customer anxieties about losses:

  • Misguided conversations about risk;
  • Short-termism;
  • The philosophy of asset allocation; and,
  • Massive overproduction of investment products designed for customers.

This extremely important point is true regardless of what you believe is the actual magnitude of loss aversion, or whether or not asymmetry between potential gains and losses is a real phenomenon as discussed in relation to Prospect Theory’s Figure 3. For the record, I believe that loss aversion is a source contributing to the emotions driving equity prices to the downside, and is therefore a rich source of alpha.

A closely related cognitive error to loss aversion is known as the fallacy of composition. Here people believe that in order to do well over the long-run, they must do well in each and every period. We then set about trying to accomplish this and end up doing poorly over the long run. In fact, this is the very point of Joachim Klement’s “Career Risk” research [see it out].

Interestingly, the so called “absolute return” funds seem to play into this fallacy, which is to a large extent driven by loss aversion. The only investment of which we are aware that produces positive returns every period is cash and we all know how well that works for building long-term wealth even if it miraculously keeps pace with inflation. 

Conclusion

I have spent so much time with loss aversion because it is, by far, the most important of the biases because of its effect on the entirety of how investment management is conducted. Further, most behavioral finance aficionados can only quote the 2:1 heuristic, rather than understanding its full complexity.

Related: MBA Thinking Can Ruin Businesses

[1] Sometimes this is referred to as “myopic loss aversion” and was first coined by Benartzi and Thaler in Myopic Loss Aversion and the Equity Premium Puzzle (1995)

[2] Kahneman, Daniel, and Amos Tversky. “Prospect Theory: An Analysis of Decision Under Risk.” Econometrica, Vol. 47, No. 2 (March 1979): pp. 263-292

[3] Kahneman, Daniel, and Amos Tversky. “Choices, Values, and Frames.” American Psychologist (April 1984): p. 342

[4] Mathematically this is: $0.05 = [(0.5 · $1.10) + (0.5 · $1.00) – $1.00]

[5] Kahneman, Daniel, and Amos Tversky. “Prospect Theory: An Analysis of Decision Under Risk.” Econometrica, Vol. 47, No. 2 (March 1979): p. 279

[6] Gal, David. “A psychological law of inertia and the illusion of loss aversion.” Judgement and Decision Making, Vol. 1, No. 1 (2006): pp. 23-32

[7] From https://blogs.cfainstitute.org/investor/2018/06/05/what-does-loss-aversion-mean-for-investors-not-much/ accessed 30 June 2018

[8] Mukherjee, Sumitava, Arvind Sahay, V.S. Chandrasekhar Pammi, and Narayanan Srinivasan. “Is loss-aversion magnitude-dependent? Measuring prospective affective judgments regarding gains and losses.” Judgment and Decision Making, Vol. 12, No. 1 (2017): pp. 81-89