Re: Avoiding EURUSD during Tokyo morning when spreads exceed my filter
Posted: Thu Oct 15, 2026 6:29 pm
Agreed that one global cap fails in both directions. A simple way to build session-specific caps: collect a few weeks of spread samples, compute the median for each hour of the day, and set the cap at about 1.5-2x that hourly median. In MQL that's a 24-element array per symbol; in Pine a lookup by hour.
The nice side effect is that unusual conditions stand out automatically. At 02:00 GMT a 1.2-pip EURUSD spread might be normal, while at 09:00 it would be blocked, without any manual session switching. Re-computing the table quarterly keeps it in line with the broker's current pricing.
The nice side effect is that unusual conditions stand out automatically. At 02:00 GMT a 1.2-pip EURUSD spread might be normal, while at 09:00 it would be blocked, without any manual session switching. Re-computing the table quarterly keeps it in line with the broker's current pricing.