Trading the second move instead of the spike
Posted: Sat Oct 31, 2026 8:13 am
The spike belongs to algorithms and to anyone willing to pay 4 pips of friction for the privilege. Chasing it by hand is a losing contest, because the fill lands after the move is mostly done.
The second move is a different game. After the initial burst, price usually gives back a chunk, spread normalises, and then it either resumes or fails. That phase has tighter spreads, readable structure and a natural stop location behind the pullback low or high.
A workable set of conditions, to be tested rather than trusted:
- Spread back within roughly 1.5 times normal
- A pullback that holds somewhere around 38 to 50 percent of the spike
- A one-minute structure break in the direction of the original move
The cost is obvious: the first 15 to 30 pips are gone and some releases never pull back at all. The gain is a stop that can be 5 pips instead of 12, with fills that resemble what the chart shows.
To check whether this suits a given pair, mark the last 20 releases and note how long the first pullback took, how deep it went, and whether the spike extreme was retested. Patterns differ between pairs and between types of release, so the sample matters more than any rule of thumb.
Does the first pullback or a break of the pullback's high have the better record in your logs?
The second move is a different game. After the initial burst, price usually gives back a chunk, spread normalises, and then it either resumes or fails. That phase has tighter spreads, readable structure and a natural stop location behind the pullback low or high.
A workable set of conditions, to be tested rather than trusted:
- Spread back within roughly 1.5 times normal
- A pullback that holds somewhere around 38 to 50 percent of the spike
- A one-minute structure break in the direction of the original move
The cost is obvious: the first 15 to 30 pips are gone and some releases never pull back at all. The gain is a stop that can be 5 pips instead of 12, with fills that resemble what the chart shows.
To check whether this suits a given pair, mark the last 20 releases and note how long the first pullback took, how deep it went, and whether the spike extreme was retested. Patterns differ between pairs and between types of release, so the sample matters more than any rule of thumb.
Does the first pullback or a break of the pullback's high have the better record in your logs?