Picture a Friday at 21:40 GMT, and a spread that is not what it was at noon
Posted: Sat Oct 31, 2026 11:18 am
The setup looks fine on the chart. Price is drifting in a small range, a clean level sits two pips away, and the temptation is to take one more scalp before the weekend.
What the chart does not show is the state of the book. New York is winding down, many desks are flat or heading home, and liquidity providers are thinning their quotes because holding risk over a weekend has a price. Spreads commonly widen through the last couple of hours of the week, and the final stretch before the close can be the widest of all.
Two details make that stretch worse than it looks. First, the move needed to break even is bigger, because the cost of getting in and out is bigger. A target that makes sense at 14:00 can be a coin flip at 21:40 once the spread has doubled. Second, thin books let small orders push price further, so the rare sharp jolt becomes more likely, and a stop on the wrong side can be filled well off its level.
There is also a platform angle. The last tradable minute differs across brokers, and quotes can freeze or go stale in the closing moments. A position left open can sit through a weekend gap that no stop is able to handle.
None of this makes the final hour untradable. It does mean the numbers feeding the decision, expected move against spread cost, are different from the numbers that looked reasonable earlier in the day.
How late into Friday does the spread stay acceptable on the pairs traded here, and where does it start to feel unworkable?
What the chart does not show is the state of the book. New York is winding down, many desks are flat or heading home, and liquidity providers are thinning their quotes because holding risk over a weekend has a price. Spreads commonly widen through the last couple of hours of the week, and the final stretch before the close can be the widest of all.
Two details make that stretch worse than it looks. First, the move needed to break even is bigger, because the cost of getting in and out is bigger. A target that makes sense at 14:00 can be a coin flip at 21:40 once the spread has doubled. Second, thin books let small orders push price further, so the rare sharp jolt becomes more likely, and a stop on the wrong side can be filled well off its level.
There is also a platform angle. The last tradable minute differs across brokers, and quotes can freeze or go stale in the closing moments. A position left open can sit through a weekend gap that no stop is able to handle.
None of this makes the final hour untradable. It does mean the numbers feeding the decision, expected move against spread cost, are different from the numbers that looked reasonable earlier in the day.
How late into Friday does the spread stay acceptable on the pairs traded here, and where does it start to feel unworkable?