Choose the pair by range per session, not by habit
Posted: Sat Oct 31, 2026 8:40 am
A lot of scalpers trade the same pair all day because it is the one they know. The problem is that average true range changes a lot between sessions for the same pair, while costs stay nearly flat. A pair that moves 40 pips an hour at the London open and 12 pips an hour in the Asian window has a completely different cost-to-range ratio at each time.
The useful number is the one-minute or five-minute average true range divided by the all-in cost per trade. If the typical five-minute range is 4 pips and the round trip costs 1.2, that is a poor environment for a 5 pip target. If the range is 9 pips, the same cost is much less of a drag.
Build a small table: for each pair you trade, compute the average five-minute range by hour of day over the last few weeks, then divide by your real cost. Highlight the hours where the ratio drops under some threshold you choose, and simply avoid those pair-hour combinations.
Results will differ by broker, because spreads and commissions differ, so use your own fills, not a published average. Averages also hide event days, so look at medians too.
Which pair and session combination gave you the best range-to-cost ratio when you actually tabulated it?
The useful number is the one-minute or five-minute average true range divided by the all-in cost per trade. If the typical five-minute range is 4 pips and the round trip costs 1.2, that is a poor environment for a 5 pip target. If the range is 9 pips, the same cost is much less of a drag.
Build a small table: for each pair you trade, compute the average five-minute range by hour of day over the last few weeks, then divide by your real cost. Highlight the hours where the ratio drops under some threshold you choose, and simply avoid those pair-hour combinations.
Results will differ by broker, because spreads and commissions differ, so use your own fills, not a published average. Averages also hide event days, so look at medians too.
Which pair and session combination gave you the best range-to-cost ratio when you actually tabulated it?