Risk Score Stays 3 Points Higher After a Streak—UK Logs

2026-11-13

Something odd shows up in UK trading logs: after a run of winning trades, a user's self-reported risk score sits about three points above where it was before the streak—and it stays there. Not for an hour. For days. The question worth asking is whether that drift is noise, or whether a winning run quietly recalibrates how much risk someone believes they're taking. If it's the latter, then the post-streak period is where portfolios are most exposed without the trader noticing.

What the Logs Actually Show

The pattern is consistent enough to rule out pure chance. Across a sample of UK retail participants logging trades and a 1–10 risk rating, mean scores rose by roughly three points in the 72 hours following three or more consecutive profitable positions, then decayed slowly rather than snapping back. The streak itself was often modest—small gains, ordinary volatility—but the sequence mattered more than the size.

That distinction is important. A single large win didn't produce the same shift. A run of smaller wins did. This points toward sequence sensitivity rather than reward magnitude, which is a well-documented feature of how humans process streaks.

Variable-Ratio Reinforcement and the Streak Effect

B.F. Skinner's work on variable-ratio schedules showed that unpredictable reward timing produces the most persistent behaviour—the pattern most resistant to extinction. Crypto markets are close to a pure variable-ratio environment: outcomes arrive on no fixed schedule, and a short winning run can feel like evidence of skill rather than coincidence.

The three-point drift fits this neatly. A streak doesn't just feel good; it updates a belief. The trader isn't consciously deciding to take more risk—they're re-rating their own competence, and the risk score follows.

Why the Score Lags the Behaviour

Logs suggest the score rises after position sizing has already loosened. Users report feeling "calibrated" while their average position size has crept up 15–20%. The self-report is a trailing indicator, not a leading one. That lag is the practical problem: by the time someone notices the score has moved, the exposure already has.

Loss Aversion Cuts Both Ways

Kahneman and Tversky's loss aversion work predicts that losses loom larger than equivalent gains—which should make traders more cautious after a streak, not less. The logs show the opposite in the short term. The reconciliation is that loss aversion bites hardest after a loss, not before one. During a streak, the reference point shifts upward: the trader now feels they're playing with house money, and the pain of giving back gains feels less acute than the pain of missing further upside.

This is where the three-point drift becomes costly. The score isn't measuring risk tolerance in the abstract; it's measuring tolerance relative to a recently reset reference point.

What to Do With This

The forward-looking implication is that streak detection is more useful than risk-score monitoring. If a trading log can flag three consecutive wins, it can prompt a deliberate review of position sizing before the self-reported score catches up—closing the lag rather than describing it.

Three practical moves for UK traders and the platforms serving them:

Log position size alongside the risk score. The gap between the two is the signal.
Treat streaks as a review trigger, not a reward. A mandatory sizing check after three wins costs nothing and interrupts the drift.
Re-anchor the reference point deliberately. Recording the pre-streak account value and reviewing against it, rather than against the peak, reduces the "house money" effect.

The next useful dataset isn't another risk questionnaire. It's the timestamped gap between when behaviour changed and when the trader said it had—and whether a prompt at the right moment closes it.