Confirmation Bias

Confirmation bias is the tendency to search for, interpret and remember information in a way that supports what you already believe. Once a trader is long a stock, bullish news looks convincing and bearish news looks like noise. Once a trader believes a strategy works, every winning example is proof and every losing example is an exception. It is one of the most pervasive biases in human thinking, and markets give it endless material to work with.

How it shows up in trading#

Why it is dangerous#

• Late exits: contrary evidence is ignored until losses are large.
• Overconfidence: a one sided view makes trades feel more certain than they are. See Overconfidence.
• Bad strategy decisions: strategies are kept or adopted based on cherry picked examples.
• Echo chambers: online communities reinforce views without testing them.

Habits that counter confirmation bias#

1. Write the bear case for every long trade (and the bull case for every short). What would have to happen for you to be wrong? Make that your stop. See Stop Loss Strategies.
2. Define invalidation before entry. If price breaks the level that disproves your idea, exit regardless of what the news says.
3. Use a checklist that requires the same evidence for every trade, rather than choosing evidence after deciding. See Pre-Trade Checklist.
4. Test strategies on full data, including all losing trades, and on data you did not use to design them. See In-Sample vs Out-of-Sample Testing.
5. Seek out disagreement: read analysis that argues the opposite of your view.
6. Keep a complete journal, so your memory of results is checked against the record. See Trading Journal.

Confirmation bias in research#

Researchers and system developers face the same problem. Testing many variations and reporting the one that worked, or stopping research once results look good, are forms of confirmation bias that produce strategies that fail live. See P-Hacking and Multiple Testing and Overfitting and Curve Fitting.

A useful question#

Before entering or holding a trade, ask: "If I had no position, would this evidence make me enter this trade in this direction today?" If the honest answer is no, your position may be shaping your analysis rather than the other way round.

Common mistakes#

• Adding indicators until one agrees with you.
• Following only like minded traders.
• Explaining away every loss without checking whether the idea was flawed.

Frequently asked questions#

What is confirmation bias in trading?#

The tendency to favour information that supports your existing view or position and to discount information that contradicts it.

How can I avoid confirmation bias?#

Write the opposite case for every trade, define invalidation in advance, use a consistent checklist and keep a complete journal of all results.

Does confirmation bias affect strategy testing?#

Yes. Focusing on winning examples or testing many variations until one looks good can make weak strategies look strong.

Next, learn about the bias that often follows success: Overconfidence.