Mean reversion works until the day it does not
Posted: Fri Oct 30, 2026 12:16 pm
A pair has stayed inside a 20 pip range for six hours. Buying near the bottom and selling near the top has worked eleven times in a row. The strategy feels solid.
Then a central bank headline lands and price leaves the range, keeps going, and never returns. Each of the previous eleven wins was 4 pips. The loss is 40.
This asymmetry is the defining risk of range-fading scalps. Small, frequent wins are paid for by rare, large losses, and a strategy can look flawless for months before one event takes back most of the gains.
The protection is not a smarter entry. It is a hard stop that is never moved, size that makes the worst day survivable and a rule to stand aside around scheduled events.
Win rate tells half the story. The other half is the size of the worst loss relative to the average win. A 90 percent win rate with a 10:1 loss ratio breaks even at best.
How is the worst-case loss accounted for in range strategies here?
Then a central bank headline lands and price leaves the range, keeps going, and never returns. Each of the previous eleven wins was 4 pips. The loss is 40.
This asymmetry is the defining risk of range-fading scalps. Small, frequent wins are paid for by rare, large losses, and a strategy can look flawless for months before one event takes back most of the gains.
The protection is not a smarter entry. It is a hard stop that is never moved, size that makes the worst day survivable and a rule to stand aside around scheduled events.
Win rate tells half the story. The other half is the size of the worst loss relative to the average win. A 90 percent win rate with a 10:1 loss ratio breaks even at best.
How is the worst-case loss accounted for in range strategies here?