Majors vs crosses: comparing cost-adjusted expectancy

[FXS London Scalper]

A setup that works on EURUSD might also appear on EURGBP or AUDNZD. The pattern is the same, but the economics may not be.

A quick framework:

Expectancy per trade (in pips) = win rate × average win − loss rate × average loss − cost

Take a hypothetical setup with a 55% win rate, average win 6 pips, average loss 5 pips:

- EURUSD, all-in cost 0.7 pips: 0.55×6 − 0.45×5 − 0.7 = 3.3 − 2.25 − 0.7 = 0.35 pips.
- EURGBP, all-in cost 1.2 pips: 3.3 − 2.25 − 1.2 = −0.15 pips.
- AUDNZD, all-in cost 2.0 pips: −0.95 pips.

The same pattern goes from slightly positive to clearly negative just from cost. Crosses also often have different ranges, so the 6-pip target may be more or less realistic.

The fair comparison is in R: scale targets and stops to each pair's typical range, then subtract cost as a fraction of the stop.

Many traders find the cross version of their setup only works if the cross has a noticeably larger typical move, which sometimes happens around specific data releases.

Do you trade crosses? Did you check the cost-adjusted numbers?