Recency Bias in Trading: Why the Last Trade Dominates the Next · EI ALGOS

Recency Bias in Trading: Why the Last Trade Dominates the Next

Anantha Krishnan··4 min read
Recency Bias in Trading: Why the Last Trade Dominates the Next

Recency Bias in Trading

Recency bias is the systematic tendency to weight recent information more heavily than older information — even when the older information is statistically stronger. In trading, this shows up as: your last three trades dictating what you think of your strategy, regardless of the 200 trades that came before.

Why it matters

Trading is a variance-heavy activity. A perfectly good 55%-win strategy will regularly produce five-loss streaks. A weak strategy will regularly produce five-win streaks. The last five trades tell you almost nothing.

Recency bias makes traders:

  • Abandon profitable strategies after normal losing streaks.
  • Double down on unprofitable strategies after normal winning streaks.
  • Change position size mid-streak, amplifying the eventual mean reversion.
  • Rewrite plans after a single bad session, throwing away months of process work.

The counter

Two ideas neutralise it.

  1. Trust the sample size, not the streak. If your strategy has produced 200 trades with a positive expectancy, five in a row against you does not change the underlying edge — only your confidence in it. Trust the math.
  2. Backfill your journal with a rolling summary. At the top of every trading session, look at the 90-day summary: win rate, expectancy, average R, plan adherence. That number is what your strategy is; the last five trades are noise around it.

How EI ALGOS surfaces it

The Trader Intelligence Score is a long-window read on trader evolution — 60 to 180 days. It exists specifically because short-window feelings mislead traders about their own performance. When your Emotion Score drops mid-streak but TIS is stable, that gap is recency bias in action.

EI ALGOS is an educational decision-support platform. Nothing in this article is investment advice.

Related reads