Do rollover spikes really hunt stops? Log it before believing it

[Amanhasnoname]

The claim comes up in every thread: spread widens at 5pm New York, price tags the obvious stop level, spread normalises, and the move reverses. It feels personal, because the trade that got stopped is the one remembered. The honest answer is that a stop can be triggered by a spread spike without any intent behind it, and a chart showing only bid prices will hide why.

What actually happens mechanically is simple. A short stop executes on the ask. When the ask jumps by 3 pips for a few seconds and the bid barely moves, the stop fires and the chart shows no wick. A long stop executes on the bid, so a widening that comes from a lower bid prints a wick. Both cases are the same event seen from different sides.

To separate conspiracy from mechanics, collect data. For 30 rollovers, note whether any of your stops were hit between 16:58 and 17:05, what the bid and ask were, and whether price reversed within 15 minutes. If stops are hit, and the bid chart never reached the level, that is spread, not a hunt.

Either way, the fix is identical: no stops sitting in that window.

Has anyone logged enough rollovers to say whether the reversal afterwards is real or just selective memory?