You buy a stock on Monday. By Wednesday you have read four articles, two broker notes and a dozen tweets about it, and every single one says you were right. That feels like research. Very often it is confirmation bias in trading: the habit of hunting for information that agrees with the position you already hold, and quietly skipping whatever doesn’t.
We have written about anchoring bias and outcome bias before. This one is sneakier than both, because it doesn’t look like a mistake while you’re making it. It looks like diligence.
In this article
- What confirmation bias in trading actually means
- Where the idea comes from: two experiments worth knowing
- Why committing out loud makes it worse
- A worked example: one trade, two readers
- Confirmation bias costs money, not just pride
- Practical ways to fight confirmation bias in trading
- A quick self-check before your next trade
What confirmation bias in trading actually means
Psychologists describe confirmation bias as the tendency to search for, interpret, favour and recall information in a way that supports what you already believe. The English psychologist Peter Wason refined the idea as a preference for information that fits a hypothesis over information that opposes it. Other names for it include myside bias and congeniality bias.
Translate that to a trading screen and it shows up in three ways:
- Biased search. You long a stock, then type its name plus “target price” or “upside” into a search bar. Nobody searches “why my stock will fall”.
- Biased reading. A results update with mixed numbers gets read as “strong core business” if you are long and “margin pressure” if you are short. Same document, two stories.
- Biased memory. You remember the three times your favourite setup worked and forget the five times it didn’t.
None of this needs bad faith. It is how attention works when you have money riding on one answer.
Where the idea comes from: two experiments worth knowing
Wason’s 2-4-6 game
In an experiment published in 1960, Wason told participants that the number triple (2, 4, 6) followed a hidden rule. They could propose their own triples, and he would say whether each one fitted. The real rule was simply “any ascending sequence”. People struggled to find it and often announced far narrower rules, such as “the middle number is the average of the first and last”.
The reason is the interesting bit. If someone thought the rule was “each number is two more than the last”, they tested triples like (11, 13, 15), which would pass under their rule. They rarely tried something like (11, 12, 19), which could have shown them they were wrong. Wason read this as a preference for confirming over falsifying, and coined the term “confirmation bias”.
Later researchers, Klayman and Ha in a 1987 paper, argued that this was less a bias towards confirmation and more a “positive test strategy”, where people run tests that fit their working idea. For a trader the distinction hardly matters. Either way, you are only running the tests your thesis is likely to pass.
If every check you run is one your trade is likely to pass, you haven’t tested the trade. You’ve just rehearsed it.
The Stanford death-penalty study
A Stanford team took people who felt strongly about capital punishment, half for and half against, and gave them descriptions of two studies on whether the death penalty deters murder. The studies were fictional, and the conclusions were swapped between groups. After reading detailed accounts, participants on both sides largely returned to their original view and rated the study that agreed with them as better designed. They set a higher bar of evidence for the study that went against them.
The Wikipedia summary of that experiment notes that 23% of participants reported their views had become more extreme. But later comparisons of before-and-after attitudes in similar experiments found no significant change, which suggests the self-reported shifts may not have been real. Researchers Deanna Kuhn and Joseph Lao concluded that polarisation is real but far from inevitable, and happens in only a small minority of cases.
Swap “death penalty” for “this PSU bank is a re-rating story” and you have most stock forums on a busy day.
Why committing out loud makes it worse
One experiment in the same body of research is especially relevant to anyone who posts their trades. Participants watched coloured balls drawn from one of two hidden baskets. One basket held 60% black and 40% red balls; the other held the reverse. The draws alternated colours, so the evidence favoured neither basket.
People who had to say their probability estimate out loud after every draw grew more confident in whichever basket they had picked first. A second group, who gave an estimate only at the end, did not show that polarisation. The takeaway the researchers drew is that the effect seems to kick in when people openly commit to a position, not just when they hold one.
Think about what that means for a trader who announces “buying here, target 20% higher” in a Telegram group or on X. Public commitment is a good thing for accountability. It is a bad thing for changing your mind.
A worked example: one trade, two readers
The numbers below are a made-up illustration, not data on any real stock.
Say Ravi buys 200 shares of a mid-cap auto parts maker at Rs 500, so Rs 1,00,000 in the trade. His thesis: export orders will pick up. He keeps a stop at Rs 460, which risks Rs 8,000, or 8% of the position.
Over the next two weeks, four pieces of information arrive:
| News | How Ravi reads it | A neutral reading |
|---|---|---|
| Management says export enquiries are “healthy” | Thesis confirmed | Vague; no order numbers given |
| Quarterly export revenue is flat year on year | “One quarter is noise” | The core thesis has no evidence yet |
| A large customer delays a model launch | Skips the article | Direct risk to next year’s orders |
| A broker raises its target price | Shares it with three friends | One opinion, with no new facts |
The stock drifts to Rs 465. Ravi, now thinking the market “hasn’t understood the export story”, moves his stop down to Rs 430. If it hits, the loss is Rs 14,000 instead of Rs 8,000. That is 75% more than he planned to lose, and the only thing that changed was how much he wanted to be right.
Notice that Ravi never lied to himself about any single fact. He just weighted the facts by whether they agreed with him. That is exactly the pattern the Stanford participants showed.
Confirmation bias costs money, not just pride
The research summary on confirmation bias in finance says it can make investors overconfident and lead them to ignore evidence that their strategies will lose money. In studies of political prediction markets, traders who resisted bias made more profit. For instance, those who read a candidate’s debate performance in a neutral rather than partisan way were more likely to come out ahead.
In our own markets, the damage tends to show up in a few familiar places:
- Moving stop-losses further away because the “story is intact”, as in Ravi’s case. Our risk management guide covers why a stop set before the trade should stay where it was set.
- Averaging down on a losing position after reading only the bullish notes.
- Echo-chamber feeds, where the accounts you follow on X or YouTube all hold the same view you do.
- Backtests built to pass: tweaking parameters until a strategy looks good, then calling that proof.
Practical ways to fight confirmation bias in trading
1. Write the exit condition before the entry
Before you buy, write one sentence: “I am wrong if ___.” Make it something you can check, like a price level, a missed order number or a specific result. Wason’s subjects failed because they never looked for the triple that would break their rule. Writing the exit condition first is you naming that triple in advance.
2. Run a pre-mortem
The confirmation bias research suggests investors imagine their investment has already collapsed and ask why. Project managers call a version of this a pre-mortem: the team assumes the project has failed and works backwards to find what caused it. A 2007 Harvard Business Review article, as quoted on that page, put the difference neatly: a normal review asks what might go wrong, while a pre-mortem assumes the patient has died and asks what did go wrong.
For a trade, give yourself five minutes. “It’s three months from now and this position is down 15%. What happened?” You will usually come up with two or three risks you hadn’t priced in.
Ask what would have to happen for this trade to fail, then go and look for exactly that.
3. Search for the bear case on purpose
For every bullish note you read on a stock you hold, find one bearish one. If you can’t find any, that tells you something too: either the trade is crowded or you haven’t looked hard enough. The finance research above suggests taking a contrary viewpoint “for the sake of argument”, and this is the simplest way to do it.
4. Keep a decision journal, not a P&L journal
Log what you expected, what evidence you used and what evidence you dismissed, at the time you dismissed it. A month later, read the dismissed column. Biased memory can’t edit a written record.
5. Commit privately, review publicly
Given the ball-drawing result, it may help to keep live positions to yourself and share trades only after you exit. You still get accountability, without locking your ego into a stance mid-trade.
6. Let rules move the stop, not feelings
If your stop moves, it should only move in the direction that reduces risk, such as a trailing stop that follows price up on a long. A stop that moves further away after the stock goes against you is almost always confirmation bias at work.
A quick self-check before your next trade
- Have I written down what would prove me wrong?
- Have I read at least one serious argument against this trade?
- Would I take this same trade if I didn’t already own the stock?
- Am I sharing this position publicly before it is closed?
- Has my stop moved further away since entry?
If two or more answers worry you, cut the size or skip the trade. Missing a trade costs nothing. Holding onto a broken thesis because every article you chose to read agreed with you can cost quite a lot.
Disclaimer: This article is for educational purposes only and is not investment advice. The trade example uses made-up numbers to explain a concept. Trading in equities and derivatives involves risk of loss; please consult a SEBI-registered investment adviser before making investment decisions.





Leave a Reply
Join the conversation
Commenting is open to registered readers. Sign up with just your email — no password needed, and it takes under a minute.
Sign up to comment Log in