Confirmation Bias in Trading: Seeing Only What You Want to See

Once you have a position, your brain starts working as its lawyer. Bullish news looks important, bearish news looks like noise. This tendency to seek out and favor information that supports what you already believe is called confirmation bias, and it affects every trader.

How it shows up

• Searching for support. After buying, you read bullish posts and skip bearish ones.
• Explaining away warning signs. A break of support becomes "a shakeout".
• Holding losers. Each new piece of bad news is discounted, so the exit keeps getting delayed. This combines with loss aversion.
• Overrating a setup. You remember the times it worked and forget the times it failed.

Why journals are the antidote

Memory is biased. A journal is not. When you think "this setup always works", your journal can tell you it worked 48% of the time with an average win of 1.1R. That is the most direct way to correct a belief.

Habits that counter confirmation bias

1. Write the invalidation before you enter. "This idea is wrong if price closes below X." When X is hit, there is nothing to debate.
2. State the bear case. For every trade idea, write one sentence about what could go wrong.
3. Use stops that do not depend on your opinion. A stop at a structural level or a multiple of ATR is decided by price, not by your mood.
4. Measure setups, do not remember them. Look at results by setup tag rather than relying on recall. See tagging trades.
5. Review losing trades first in your weekly review. They have the most to teach and are the easiest to skip.

Beware of curated information

Social media feeds and chat rooms tend to show you more of what you already engage with. If you only follow traders who share your view, you will rarely see a strong counterargument. Diversify what you read, or read less during market hours.

In TradeThink

Use setup tags and the reports to check what each setup really earns, and use notes on each trade to record the invalidation level you wrote before entry. When the data disagrees with your memory, trust the data.