Why cross mismatches vanish before you can click
Posted: Sat Oct 31, 2026 11:49 am
The myth: if EUR/USD, GBP/USD and EUR/GBP are slightly out of line on the screen, that gap is free money for anyone quick enough to hit all three legs.
Mismatches do appear on retail platforms, usually a few tenths of a pip, and the arithmetic looks tempting. The problem is who is actually in a position to close them. Banks and professional firms run triangular arbitrage with co-located servers, direct market access, and execution measured in microseconds. They close the gap long before a retail platform finishes drawing the tick.
What retail traders see is typically one of three things. A quote timing difference, where one leg updated and another has not yet refreshed on the platform. A spread effect, where the apparent gap is smaller than the combined cost of crossing three spreads. Or a stale price that would be rejected on a real order, since the quote shown is not what is actually dealable.
So the gap is often an artifact, not an opportunity. Even when it is real, three legs mean three spreads, three sets of slippage risk, and a chance one leg fills while another does not, leaving an unintended position.
That said, the idea is not useless. Watching for a cross that lags its constituents is a decent way to notice a feed running slow, and a slow feed is something worth knowing before entering.
Has anyone ever seen a triangular mismatch that survived the spread on a live account, or is it always gone when the orders hit?
Mismatches do appear on retail platforms, usually a few tenths of a pip, and the arithmetic looks tempting. The problem is who is actually in a position to close them. Banks and professional firms run triangular arbitrage with co-located servers, direct market access, and execution measured in microseconds. They close the gap long before a retail platform finishes drawing the tick.
What retail traders see is typically one of three things. A quote timing difference, where one leg updated and another has not yet refreshed on the platform. A spread effect, where the apparent gap is smaller than the combined cost of crossing three spreads. Or a stale price that would be rejected on a real order, since the quote shown is not what is actually dealable.
So the gap is often an artifact, not an opportunity. Even when it is real, three legs mean three spreads, three sets of slippage risk, and a chance one leg fills while another does not, leaving an unintended position.
That said, the idea is not useless. Watching for a cross that lags its constituents is a decent way to notice a feed running slow, and a slow feed is something worth knowing before entering.
Has anyone ever seen a triangular mismatch that survived the spread on a live account, or is it always gone when the orders hit?