Who it's for — Anyone who changes size, rules, or strategy after the last «amazing» or «terrible» week — forgetting the real statistical sample.
Recency bias makes the last trade, session, or streak seem more representative of the system than it really is. After three wins you increase risk; after three losses you abandon the plan — the market does not «have to» continue the last sequence.
In plain terms — «Everything works lately» or «I keep losing lately» — and you act as if that were the new normal, ignoring hundreds of past trades.
Typical effects
| Recent streak | Distorted behaviour |
|---|---|
| Win series | Overconfidence, excessive size |
| Loss series | Abandon system, revenge trading |
| One mega trade | Generalize untested strategy |
| Recent sideways market | Forget prior trend regimes |
Contrast with sample size and expectancy: process decisions should rest on sufficient samples, not the latest block.
Common mistake — Disabling rules after 5 consecutive losses — without checking if they fit expected system variance.
Example — Three wins at 3R: you double size «because the market is easy». Fourth trade — full stop at double size: one loss wipes three wins.
Summary card
- What it is: overweight on recent events.
- Antidote: journal, rolling metrics, fixed size rules.
- Related: Overconfidence.