Recency bias in trading is the habit of giving the last few trades, or the last few days of price action, far more weight than they deserve. A short losing run starts to feel like proof that a method is broken, and a short winning run feels like proof that you have cracked it.
Quick answer: what is recency bias in trading?
It is judging your method, your sizing or the market by what happened most recently instead of by the full record. Typical signs are dropping a tested method after a few losses or raising risk after a few wins. The fix is to judge results over a large, fixed sample and change rules only at planned reviews.
What recency bias in trading looks like
The bias is not unique to traders. A 2018 paper in the Journal of Vision opens by noting that human bias towards recent events is common and well studied, and that it persists even when past events contain no information about the future. Our attention leans towards the latest thing whether it is useful or not.
In trading, it tends to show up in a few familiar ways:
- Abandoning a method after a short losing run, even though the full record is positive.
- Raising position size after a few winners, because it suddenly feels easy.
- Chasing the market that moved most this week and ignoring the ones you prepared for.
- Letting the last trade decide the next one: skipping a valid setup because the previous one lost.
Recent price action vs recent results
Not everything recent should be ignored. The latest swing highs and lows, today’s news and the current range are genuinely relevant to the next trade. Reading current structure is the job.
The trap is treating your own recent results the same way. Ten trades is a tiny sample of a method’s behaviour. The Nobel committee’s summary of Daniel Kahneman’s work describes people who assign the same probabilities to small and large samples and so seem to adhere to a law of small numbers. A trader who rewrites the rules after a bad fortnight is doing exactly that.
| Bias | The thought | The error |
|---|---|---|
| Recency bias | “The last five trades show what this method is” | Overweights a small, recent sample |
| Gambler’s fallacy | “After five losses, a win is due” | Expects the streak to correct itself |
| Hot hand | “I’m on a run, size up” | Expects the streak to continue |
| Anchoring | “It was 200 points higher last week” | Fixes on one remembered number |
The gambler’s fallacy in trading and recency bias often appear together after a losing streak, and anchoring bias in trading is its cousin: one fixes on a recent run, the other on a single number.
Worked example: ten trades vs one hundred
Example (hypothetical): a trader on a £5,000 account risks £50 per trade, so 1R is £50. Over 100 trades taken by the same rules, the method wins 40 times at 2R and loses 60 times at 1R.
The full record is 80R won minus 60R lost: +20R, or +£1,000, which is an average of +0.2R (£10) per trade. Over the last ten trades, though, there were two winners and eight losers: +4R minus 8R, or -4R (-£200). Looked at alone, the last ten trades say the method is broken; the hundred say it is doing what it did before.
How normal is a losing streak?
For a method that loses 60 percent of the time, any given six trades all lose with a probability of 0.6 multiplied by itself six times, about 4.7 percent. But 100 trades contain many overlapping stretches of six. Assuming each trade is independent, our calculation puts the chance of at least one run of six or more losses in 100 trades at roughly 87 percent.
In other words, the streak that tempts the trader to quit is the expected experience, not a warning. The real question is whether the long-run average has changed, which is what measuring expectancy in trading over a full sample tells you.
The winning side costs money too
Now reverse it. After five winners in a row, the same trader feels sharp and doubles risk to £100. The next two trades lose: -£200, the same damage as four losses at normal size. Spread bets and CFDs are leveraged, so a bigger stake also means a gap or fast move can cost more than the planned amount.
A big win can do the same; our piece on what to do after a big winning trade covers that moment. A run of wins can also tip into overconfidence bias in trading.
How to reduce recency bias in trading
You will still notice the last few trades; the aim is to stop them making decisions.
- Fix risk per trade in advance. A streak, winning or losing, is not a reason to change size. Change it only at a planned review.
- Judge the method on a set sample. Decide in advance, for example, that the method is reviewed every 50 trades, and not before.
- Keep the full record in front of you. Your trading journal for beginners should show the running total in R next to the last ten trades, so the short view never stands alone.
- Compare against the backtest. If your backtest of the trading strategy showed streaks of six or seven losses, a live streak of six is within the range you already accepted.
- Review weekly, decide monthly. A weekly trading review is for checking whether rules were followed; rule changes wait for a full sample.
- Prepare the watchlist before the week starts. It stops last week’s biggest mover pulling you into a market you have not studied.
Common mistakes
- Switching methods after every bad week. Each switch resets the sample, so you never learn whether anything works.
- Calling a losing streak a broken edge. Check it against your own drawdown history first; our guide to handling a trading drawdown explains how.
- Trying to win it back. Recent losses feeding bigger trades is the path into revenge trading.
- Chasing last week’s mover. That pull is closely linked to FOMO in trading.
- Ignoring recent price structure. The opposite error: current highs, lows and ranges do matter.
Frequently asked questions
What is an example of recency bias in trading?
Dropping a tested method after five losses in a row, even though its 100-trade record is positive, is the classic case. Doubling position size after a winning run is the same bias pointing the other way.
Is recency bias the same as the gambler’s fallacy?
No. Recency bias assumes recent results show what will continue, while the gambler’s fallacy assumes a streak must reverse. Both treat a short run as more meaningful than it is.
Does a losing streak mean my strategy has stopped working?
Not on its own. Compare the streak with the longest losing runs in your backtest and journal, and judge expectancy over a full sample before changing anything.
Should recent price action matter more than older data?
For reading structure, often yes: the latest swings and ranges are what the next trade depends on. For judging your method, no: that needs the whole record.
Next steps
Add a running R total to your journal today and set the trade count for your next method review. For the wider picture, read our guide to the five psychological hurdles traders face.
The five key psychological hurdles and a rules-based risk model are both covered in our price action trading course, and the basics section is free to start.
This article is for education only and is not financial advice. Trading leveraged products such as CFDs and spread bets carries a high risk of losing money.
