Herd Mentality And Crowd Psychology

Popularity says nothing about accuracy

When a great many people are doing the same thing, it is hard not to read that as information. Herd mentality is acting on the reading: following what the crowd appears to be doing, on the assumption that a lot of people moving together must know something. In markets that assumption is unreliable, because the crowd is not an independent panel of judges. It is a group of participants mostly reacting to each other.

The reason it feels safe is social rather than financial. A trade that goes wrong while everyone else was doing the same thing feels less personal than one you reasoned your way into alone. The comfort is real and the protection is imaginary: a position loses exactly the same amount whether or not it was fashionable.

Why a crowded trade becomes the fuel for its own reversal

This is the part worth understanding mechanically, because it explains why the sharpest reversals happen when a view is most widely held.

Suppose sentiment on a pair reaches the point where the great majority of open retail positions are long. Two things follow. First, most of the people who wanted to buy have already bought, so the marginal buyer who would push price higher is largely used up. Second, all those positions carry stops, and stops below a crowded long position are sell orders sitting in a cluster.

Price does not need bad news to fall from there. It needs an absence of new buyers plus one push into the cluster. The first stops trigger, those sell orders push price lower, which triggers the next band of stops, and the move accelerates for reasons that have nothing to do with the original reasoning. The crowd supplied both the lack of demand and the supply. How forex prices move covers the order-flow side of this in more detail.

Where it shows up in ordinary decisions

• Entering because the chat is excited. By the time an idea is circulating widely, the entry that made it attractive has usually been and gone, which is why crowd-sourced trades so often come with a terrible price relative to the invalidation level.
• Selling into panic because visible commentary has turned negative, rather than because your own invalidation level was reached.
• Reading attention as confirmation. Heavy volume and heavy discussion tell you participation is high. They do not tell you which direction the participation is about to resolve in.
• Copying a position without copying the exit. A trade taken from someone else without their stop, target and sizing is not their trade. It is a different trade wearing their entry price.

Social media adds a reporting problem on top of the crowding one. Winning trades get posted and losing ones quietly do not, so any feed you follow shows a heavily filtered sample. The crowd therefore looks more successful than it is, which makes following it feel more reasonable than it is. You are not observing a group of participants, you are observing the part of their activity they chose to publish.