The impulse-pullback-continue setup has threads for EURUSD, EURGBP, EURJPY and USDJPY, plus a couple of general ones on pullback depth. They describe the same trade with different filters, and the filters are the useful part. Here's the combined picture.
Define the impulse
The EURGBP thread is clear that an impulse means a decisive displacement away from a level, not three overlapping M5 bars that lean one way. The first pullback card thread puts a number on it: at least 1.5 times the average five-minute range, in one direction, through the level, with normal spread. It also requires the level to exist before the open, such as Asia's high or low, the previous day's extreme, or a round number that mattered overnight. Not something drawn after the move starts.
Define the pullback
This is where people invent trades, and the threads each have a rule against it:
- Two-candle rule on EURUSD M5: the pullback must print two candles in a row that can't push to a new extreme against the move, with the second has to close back in the impulse direction. One doji isn't permission.
- Pullback into something: the USDJPY thread only takes pullbacks that land in prior micro-structure, like a just-broken level, a small shelf, or a sweep pocket. "Pullbacks into nothing are hopes. Pullbacks into something are trades."
- The half retrace thread waits for price to come back towards the middle of a 20 to 40 minute London impulse, using a zone around 40 to 60 percent rather than an exact line. A reply there likes it when price drifts into the zone on small overlapping candles rather than diving into it.
The pullback depth thread suggests measuring it on your own pair: take 30 or more impulses, record each pullback as a percentage of the impulse, and look at the distribution. Many shallow pullbacks retrace 30 to 50 percent. Replies add two refinements. Some pairs habitually pull back shallower or deeper than textbook levels, so measure your own. And use the median, because one or two very deep pullbacks drag the average down and have you waiting for entries that rarely come.
Abort rules matter as much as entry rules
The EURJPY abort thread is the one I'd print. The idea is dead if any one of these happens: the pullback extends beyond half of the impulse with acceptance rather than a wick; a lower timeframe break of the pullback structure goes against you and holds; or time runs out with no continuation. Any one is enough. A reply allows one re-entry, but only once price makes a fresh high or low in the impulse direction. If it doesn't, that idea is finished for the morning.
The EURUSD two-candle thread has the same spirit: a deep wipe through the impulse midpoint cancels the idea for that drive, and the stop doesn't get widened.
Costs on crosses
The EURGBP thread asks whether the setup still pays after costs, and the reply's numbers explain why it's tight: 0.5 to 0.8 pips spread at the London open plus about half a pip commission, on a cross with modest targets. The rule there is to skip when the projected reward is too small a multiple of that.
The common thread
All of these threads agree the setup is simple. The difference between a trade and a guess is having the impulse, the pullback and the abort written down before the open.
What's your rule for a pullback that's gone too deep? Half the impulse, a structure break, a time limit?
Threads this pulls from
- Impulse-pullback-continue on EURUSD M5: the two-candle rule I use
- EURGBP and the impulse pullback: does it still work after costs?
- When impulse pullback fails on EURJPY: my abort criteria
- First pullback after London impulse: entry criteria
- Half retrace entries after the London impulse on EURUSD, no indicators, just the 50 percent line
- USDJPY pullback scalps into prior micro-structure, not into thin air
- Pullback Depth Statistics: How Far Do Retracements Usually Go?
- Pullback Trading, Buying Dips and Selling Rallies Properly
