5 Cognitive Biases That Are Costing You Money (And How to Stop Them)

The Five Biases That Show Up Most In Trader Journals

Cognitive biases are not character flaws. They are evolved shortcuts that helped a different brain in a different environment survive a different set of problems. The same shortcuts, run at a trading desk, are reliably expensive.

Across journal reviews, five biases show up more than the others. The intervention is not awareness. Awareness is necessary and insufficient. The intervention is a written structural change at the layer where the bias actually operates.

1. Confirmation Bias

What it is: The brain weights information that supports the current position and discounts information that contradicts it.

What it looks like: Long on a name, you find yourself reading only the bull case. The bear thesis is dismissed as noise before it is read carefully.

The structural fix: Before any entry, write three reasons the trade could fail. Three. In writing. If you cannot generate three, the work to take the trade is not done.

2. Anchoring

What it is: The first number encountered disproportionately influences subsequent judgment.

What it looks like: A stock at one hundred dollars drops to eighty. Eighty feels cheap relative to one hundred, even when fundamentals justify sixty. You buy because of the anchor, not because of the level.

The structural fix: Value the trade against a written set of levels generated independently of the recent high. If the level was not on the chart before today, it does not count as a reference.

3. Loss Aversion

What it is: Losses are felt approximately twice as intensely as equivalent gains. This is one of the most replicated findings in behavioral finance, going back to Kahneman and Tversky in 1979.

What it looks like: Winners cut early to lock in the gain. Losers held past the stop to avoid realizing the loss. Same chart, opposite handling, because the body is making the call.

The structural fix: Stops and targets entered as bracket orders before the trade is opened. The in-trade decision is removed entirely.

4. Recency

What it is: Recent events are weighted more heavily than older events of equal importance.

What it looks like: Three winners in a row, and the next position size drifts up. Three losers in a row, and the next valid setup is refused.

The structural fix: Position sizing is a written percentage of equity, recomputed weekly, not after every trade. The number does not respond to last Friday.

5. Overconfidence

What it is: Skill is over-estimated, especially in domains with noisy feedback. Trading is one of those domains.

What it looks like: Trade frequency rises. Sizing rises. The risk plan that worked is treated as a beginner constraint rather than a load-bearing wall.

The structural fix: Track win rate and expectancy across at least one hundred trades before drawing any conclusion about edge. The honest number almost always reads lower than the felt number.

The Pattern Underneath All Five

The biases are not the problem. The biases are constants. The problem is the absence of a written structural layer that intervenes before the bias becomes a click. The journal, the bracket order, the pre-trade pause, the weekly review. These are not optional add-ons. They are the load-bearing walls that make the rest of the system survivable.