Nobody sets out to fool themselves. You form a view, you research it, and the research supports it. The problem is that the second step is rarely neutral. Once you have a view, you search for, interpret and remember information in ways that favour it — and you do it without noticing.
In trading this is expensive, because the market doesn't care how well-researched your view felt. This guide covers what confirmation bias is, the specific signs it leaves in your notes, and the habits that reduce it.
What is confirmation bias?
Confirmation bias is the tendency to favour information that supports what you already believe. Psychologists have studied it for decades, and it's one of the most consistently observed biases in human judgement. It works in three ways:
- Search. You look for evidence that supports your view and spend less time on evidence that might not.
- Interpretation. You judge ambiguous evidence as supportive, and hold contrary evidence to a higher standard.
- Memory. You recall the facts and past calls that fit your view more easily than the ones that don't.
A classic experiment illustrates the second point: when people with opposing views on a contested issue were shown the same mixed evidence, both sides rated the studies supporting their position as more convincing — and came away more certain than before. The same evidence pushed them further apart.
Researchers have found the same pattern among investors. A study of stock message boards found that investors paid more attention to posts agreeing with their existing view, and that those who showed the strongest bias tended to be more confident, traded more — and earned lower returns.
Eight signs of confirmation bias in your trading
You can't feel confirmation bias directly. You can see its traces in how you research and what you write down.
- Your reading list agrees with you. Most of the sources you consulted support the trade. Algorithmic feeds make this worse: engage with bullish posts and you'll be shown more of them.
- You scrutinise bad news harder than good news. A supportive data point gets accepted at face value; a hostile one gets questioned, adjusted or called noise.
- The same data is bullish either way. A rate cut is good because money gets cheaper; no rate cut is good because the economy must be strong. If no outcome could hurt your view, it isn't a view.
- Your trade gains reasons after entry. The original reason weakens, but new ones appear to replace it. The position stays; the justification changes.
- You dismiss the bear case by its author. "They're always bearish" or "they're talking their book" replaces engaging with the argument.
- Your invalidation point moves. The level or event that would have made you wrong gets pushed further away as it approaches. (How to set one that holds.)
- You average down because it's "even cheaper". Sometimes justified — but often a way to avoid asking why the market disagrees.
- You remember your good calls best. Your mental track record is better than your actual one, because the misses fade.
Why it gets worse once you're in a trade
Before entry, confirmation bias shapes which ideas you pursue. After entry, it has money behind it. Several other forces join in:
- Commitment. Having acted on a view, changing your mind means admitting the action was a mistake.
- Ownership. People tend to value what they own more highly than the same thing they don't own.
- Sunk cost. The time spent researching and the loss already taken both make it harder to walk away.
- Community. Online groups built around a trade reward agreement and punish doubt, which turns individual bias into a shared one.
This is why a losing trade can look exactly like an early one. From inside the position, every fact has a supportive reading available — and that's the one you'll find.
Catch it in your own notes: a 15-minute audit
Take the notes from your last three trades — or your current open positions — and check them for these patterns.
| Check | What to look for | What it suggests |
|---|---|---|
| Source balance | Count the sources that support the trade versus those that challenge it | A heavy tilt means you searched for agreement |
| Primary sources | How many claims trace back to a filing, statement or official data release | Commentary-heavy notes inherit other people's bias |
| Assumptions | Claims written as facts that are really forecasts | Beliefs that were never tested |
| Opposing case | Whether a strong counter-argument is written down at all | If not, it was never seriously considered |
| Invalidation | Whether it was written before entry, and whether it has changed | A moving exit is a sign of motivated reasoning |
| Reasons over time | Whether the reasons now differ from the reasons at entry | Trade creep — the position outlived its logic |
Don't aim for a perfect score. Aim to find the one trade where the pattern is clearest, and ask whether you'd open it today, from flat, at this price.
How to reduce confirmation bias
Knowing about confirmation bias doesn't stop it. Research on the "bias blind spot" suggests people readily see biases in others while believing themselves less affected. What works is process — habits that force the other side into view whether you feel biased or not.
1. Write the trade and the exit before you research further
Put your trade and your invalidation conditions on paper early. Once they're written, it's much harder to quietly move them later. The pre-trade checklist is a good template.
2. Look for the opposite first
Studies have found that simply asking people to consider the opposite — how they'd judge the evidence if it pointed the other way — reduces biased interpretation more than asking them to be fair. Before reading anything supportive, spend time on the best argument against.
3. Use one standard of evidence
For every source, ask: would I accept this if it said the opposite? If a bearish note from the same analyst would be dismissed as unreliable, the bullish one shouldn't count as proof.
4. Build the bear case yourself
A bear case written by someone else is easy to dismiss. One written by you, from your own evidence, is not. See bull vs bear case examples for how to structure it.
5. Change your inputs
Deliberately follow a few thoughtful people who usually disagree with you. Read primary sources, which haven't been filtered through anyone's view. And check positioning — if everyone you read agrees, you may be in a crowded trade.
6. Re-underwrite from flat
At regular intervals, ask: if I had no position, would I open this one today, at this price, for these reasons? If the answer is no, the position is being held by bias, not by the trade.
7. Keep a decision journal
Record what you believed, how confident you were and why, at the time of the decision. Memory rewrites itself to fit outcomes; a journal doesn't.
8. Get an outside read
Ask someone with no position — and no stake in agreeing with you — to find the weakest part of your trade. Ask them to attack it, not to review it.
Related biases to watch
| Bias | What it is | How it feeds confirmation bias |
|---|---|---|
| Anchoring | Over-weighting the first number you saw, such as your entry price | New information gets judged against the anchor rather than on its own |
| Overconfidence | Believing your judgement is more accurate than it is | Makes it feel unnecessary to look for contrary evidence |
| Sunk cost | Letting past investment influence future decisions | Raises the cost of admitting the evidence has changed |
| Narrative fallacy | Preferring a coherent story to messy evidence | Evidence that fits the story is remembered; the rest is dropped |
| Disposition effect | Selling winners too early and holding losers too long | Losing positions attract a search for reasons to hold |
| Herding | Following what others are doing | Agreement from the crowd feels like independent confirmation |
Mistakes when fighting it
- Reading the bear case only to rebut it. That's confirmation bias with extra steps.
- Assuming awareness is enough. Knowing the name of a bias doesn't make you immune to it.
- Overcorrecting into paralysis. The goal isn't to doubt everything. It's to hold every view to the same standard, then act.
- Confusing contrarianism with independence. Automatically disagreeing with the crowd is just a different bias.
An outside read on every trade
Research tools are usually built to confirm, not challenge. Blindspot is built to disagree with you: it builds the strongest case against your trade as carefully as the case for it, checks the claims your trade rests on with a web search that prefers primary sources, and flags unexamined assumptions and contradictions in your reasoning. No signals, no price targets — it scores the reasoning, never the trade.
Frequently asked questions
What is confirmation bias in trading?
It is the tendency to seek out, interpret and remember information in ways that support a trade you already favour or hold, while discounting evidence against it.
What is an example of confirmation bias in trading?
A trader who is long a stock reads a rate cut as bullish because money gets cheaper, and no rate cut as bullish because the economy must be strong. When every possible outcome supports the position, the view can no longer be proven wrong.
How can traders avoid confirmation bias?
It can't be eliminated, but process reduces it: write the trade and invalidation conditions before entry, look for the opposing argument first, apply the same standard of evidence to both sides, build the bear case yourself, and regularly ask whether you would open the trade today from flat.
Why is confirmation bias dangerous for traders?
It makes a trade feel better supported than it is, delays exits when the evidence changes, and turns losing positions into long-term holdings. Traders often discover what they missed only after the loss.
Is knowing about confirmation bias enough to avoid it?
No. People tend to see biases in others more easily than in themselves. Reducing confirmation bias depends on habits and checks that force contrary evidence into view, rather than on awareness alone.
This guide is general education about research process. It is not investment advice or a recommendation to trade any instrument. Trading involves risk of loss.