forex-scalping.com

The first pullback after a CHoCH is where people get trapped

You see a clean change of character on the lower timeframe, price breaks the last higher low, and you immediately look for a short on the first pullback. That is the most popular setup in this style, and also one of the most crowded. The issue is that the first pullback often retraces into the area where the old trend's orders still sit, and many times price takes it out and continues the previous trend. I have noticed the better entries usually come from the second reaction or from a pullback that is shallow and rejects a clear zone. A rule I follow: if the pullback after CHoCH retraces beyond the 50 percent of the impulse leg, I stop treating it as a clean reversal and demote the setup. Also check whether the CHoCH took out meaningful liquidity first. A CHoCH from nowhere is weaker than a CHoCH after a sweep of a previous day high. The signal is not the break itself. It is the story that surrounds it.

The 50 percent rule is a good start, but I'd also look at how the pullback moves, not only how far. A pullback that drifts up in small overlapping candles is usually just a pause before the next leg down. One that comes back in a few big bullish candles with closes near their highs is telling you buyers are still around. Same retracement depth, very different message. I'd also be careful on news days, because a CHoCH printed just before a release can be reversed by one candle and the structure read stops meaning much until the dust settles.

Question for people who trade this: do you wait for the first pullback to fail and then sell the second one, or do you need a fresh lower high to form before you're interested? The second option means you give up some of the move, but you also get a clearer stop, above that new lower high instead of somewhere vague above the CHoCH area. I lean toward the second, mostly because it removes the urge to guess where the pullback will end. Curious whether people find the later entry actually costs them much in R once they track it properly.