There is almost endless evidence showing that your average investor acts emotionally (selling into a bear market, buying in the wake of a rally) hurting their own long-term returns. This is usually attributed to untrained investors getting caught up in the media hype one way or the other, but research into the effect that anchoring has on estimation shows that even people who should know better can get caught up in the alleged wisdom of crowds.


A well-known concept in psychological circles, anchoring is when someone relies on the reasoning of other people instead of making his own informed choice, either as a way to fit in with the group or simply because the cognitive load is too high and using the group consensus is more practical.

So Center for European Governance and Economic Development researchers Lukas Meub and Till Proeger decided to put people in a situation very much like the one facing your everyday retail investor (h/t TurnKey Analyst).

Additional, unimportant information impacted estimates instead of being discarded

First, people were tasked with estimating the solution to an algebraic equation that had one random term tacked on, and were rewarded based on how close they got, and then repeated for fifteen rounds. Meub and Proeger then repeated the experiment with more complicated equations to increase the cognitive load.

Rationally, the optimal strategy is to simply solve the equation since the random term could be positive or negative and had an expected value of zero, regardless of what happened last round.

One group simply went through their paces, predicting the outcome round after round. But the other group was shown the average solution from the previous round. What’s important is that this additional information doesn’t serve any purpose. The best strategy is to solve the equation, ignoring the random term.

“The additional, albeit useless, information shown in the treatment groups creates a general noise in subjects’ estimations. The social anchor values have an overall significant and relevant impact, especially when cognitive load is high,” write Meub and Proeger. “On average, subjects are unable to ignore the averages determined by the predictions of all players, as the rational strategy would suggest… the anchoring bias is not only driven by subjects performing poorly,” (emphasis mine).

In other words, it’s not just the people who struggled to solve the equations who are affected by social anchoring. People who are perfectly capable of figuring out the rational strategy don’t follow it when exposed to the random information of last round’s average solution.

Social anchoring may also cloud investors’ judgment

The analogies between this experiment and the markets are straightforward. People are faced with a challenging task that can be approached rationally – determine the net worth of companies and then compare that estimate to the current stock price. This isn’t an easy task by any means, and it isn’t deterministic, but it can be approached methodically. Just like in the experiment, a methodical approach isn’t what we see, and anchoring may be part of the reason why.