Risky Shifts

Female Underwear (Not)

No doubt some will be disappointed to learn that a risky shift is not a slightly daring article of female underwear, but an outcome of the counterintuitive process of group polarization. This says that if you take a group of roughly like-minded people they will gravitate to a more extreme position than that of the average team member. 

Mostly people would expect the opposite, that the average would hold sway, but the fact that this doesn’t seem to happen is a concern in many spheres, some of them more important than investing.  However, as financial markets often seem to be populated by like-minded groups of wildly overoptimistic people, they’re excellent breeding groups for risky shifts.

Group Polarization

Way back in 1959 James Arthur Finch Stoner of Antioch College submitted his Masters thesis, entitled A Comparison of Individual and Group Decisions Involving Risk.  He started by remarking, rather mildly:

“The belief that groups of people are more cautious than individuals is held by a number of people, and possibly by a majority”

But  then went on to observe:

“The analyses show that when the subjects reached decisions as members of a group they tended to advocate significantly more risky courses of action than they had chosen when they reached decisions as individuals”.

Even better, or worse, according to your viewpoint, people who agreed with the group decision became more confident about the decision, while those who disagreed became less confident about their own opinion.  The group opinions polarized to a more extreme position.

Extremism

There’s since been a lot of research into group polarization, because it’s clearly of interest across a lot of different arenas, particularly in a world seeming determined to run itself by committee.  Perhaps the key additional finding is that polarization is more likely or more powerful in groups with a shared identity or shared beliefs.  Here’s Cass Sunstein in Deliberative Trouble? Why Groups Go To Extremes:

“Notably, groups consisting of individuals with extremist tendencies are more likely to shift, and likely to shift more; the same is true for groups with some kind of salient shared identity (like Republicans, Democrats, and lawyers, but unlike jurors and experimental subjects). When like-minded people are participating in “ iterated polarization games”—when they meet regularly, without sustained exposure to competing views—extreme movements are all the more likely.”

Yep, political meetings, terror cells and investment clubs are all hotbeds of polarizing opinion.

There are all sorts of complicating factors associated with polarization: one factor appears to be that individuals with extreme positions tend to express them with great confidence, which in turn encourages people to shift towards them.  But more important seems to be the power of group identity – the more similar people are the more likely it is that the group will experience a risky shift.

Polarized Bubbles

Stephen Utkas, writing for Vanguard in Market bubbles and investor psychology, has proposed a neat four stage model of bubble formation in which polarization is strongly implicated.  Firstly he suggests that the representative heuristic, in which we form opinions based on concepts we can easily bring to mind, biases us into chasing short-term success stories: it’s an example of the hot-hand effect where we mistake a short-term run of luck for skill (see: Recency: Hot Hands and the Gambler's Fallacy). 

Unfortunately “short-term” in markets may actually be most of an investor’s active lifetimes.  For instance, it’s entirely possible that this problem lies behind beliefs such as “stocks always outperform bonds”, or that “buy-and-hold is a strategy that always works”.  These concepts aren’t permanent features of markets but did work for quite long periods of time in the 1980’s and 1990’s – long enough for people to lose the true statistical relationships in trends that dominated their investing lifetimes.

Secondly, Utkas suggests that overconfidence then kicks in and triggers extrapolation of the trends seen in the first stage (see: Overconfidence and Over Optimism).  Forecasts of future returns become increasingly optimistic, especially as memories of past problems recede.  The scene is set for a risky shift.

Mass Mania

The issue with this kind of model is the problem of transmission: how do individual biases turn into mass manias?  The argument here is that it’s caused by a combination of groupthink and group polarization.  As optimism grows and disasters recede into distant memory large groups of like-mindedly overoptimistic investors undergo a risky shift.  It’s a beguiling concept, and one that with modern communication technology seems all too possible, even if it’s harder to find similar mechanisms to explain pre-internet bubbles.

Finally Utkas suggests that the inevitable crisis that occurs when a bubble bursts causes another group polarization effect, but this time in the opposite direction of risk.  Suddenly, faced with the nasty reality of markets, groups shift to a risk adverse position, and this re-polarization – or a conservative shift – contributes to the undermining of the market position.

It’s a neat use of a number of individual biases affected at the group level to explain market phenomena, and would seem to applicable to sentiment changes in markets such as we’ve experienced quite frequently over the past decade, as well as larger bubble and bust phenomena.  The problem is that it’s hard to test and, as the whole experience of group polarization shows, commonsense intuition is as likely to guide us wrong as right when it comes to human behavior in these situations.

Filter Bubbles

Nonetheless it’s an analysis that seems particularly relevant to the heavily connected world we live in.  The celebration of the vast realms of data and analysis that the internet offers us (see: Twits, Butter and the Superbowl Effect) needs to be tempered with the realization that greater access to the opinions of like-minded others is not necessarily an unmitigated benefit.  In particular the internet is increasingly likely to create personal bubbles of experiences that correspond to what the avid searcher already believes in.  

This is the Filter Bubble experience that Eli Pariser is so passionate about, where profiles of individuals are used to customize the content they’re presented with, can operate to prevent us being brought into contact with people or ideas that we don’t agree with.  It’s hard, for instance, to believe that there are many people reading this who don’t broadly agree with the benefits of corporate capitalism; despite this there are good arguments that it’s not a panacea for the world’s ills.

Served and Owned

If people are served only that which they already agree with, and are brought into contact only with people whose views they share, then polarization is likely.  This has implications for society at large, of course, but is a particular problem for investors who, by and large, tend to have quite a lot of opinions in common.

So, oddly enough, exposing ourselves to opinions we don’t like and don’t agree with may actually help us become better investors.  Of course, some people will never be open to changing their minds short of radical brain surgery with a can opener and a spoon.  Generally such people aren’t useful role models for anyone with a mind of their own and the patience to use it.

Related articles:

• Backfiring Investment Theories
• Behavioral Finance's Smoking Gun
• Sharpshooting the Investment Gurus
• Overconfidence and Over Optimism
• Twits, Butter and the Superbowl Effect
• Recency: Hot Hands and the Gambler's Fallacy

Group Polarization, Risky Shift and Internet Filter Bubble added to The Big List of Behavioral Biases.