First pullback rules on GBPJPY after impulse

[LondonScalper]

GBPJPY impulses are big and fast, and the first pullback after one is where I've found most of my edge on the pair. These are my rules, written for myself and refined over the last couple of years.

What counts as an impulse. A move of at least 40 pips within 30 minutes during London hours, with no M5 candle against the direction larger than a third of the biggest M5 candle in the move. If it's choppy, it's not an impulse.

What counts as the first pullback. Price retraces between 30% and 60% of the impulse, measured on M5 closes. Less than 30% isn't a pullback, more than 60% suggests the impulse may be finished.

Entry trigger. An M5 candle that closes back in the impulse direction from within that retracement zone. I enter on the next candle open.

Stop. Below the pullback low (for a long) plus 3 pips for spread and noise. On GBPJPY that's often 15 to 25 pips, so size comes down accordingly.

Target. The impulse high for the first half of the position, and a move equal to the impulse for the rest, with the stop moved to break-even once the first target is hit.

Over 70 logged trades, it's my most profitable setup per trade, and my least frequent.

[PTScalper]

GBPJPY impulses are big and fast, and the first pullback after one is where I've found most of my edge on the pair. These are my rules, written for myself and refined over the last couple of years.

What counts as an impulse. A move of at least 40 pips within 30 minutes during London hours, with no M5 candle against the direction larger than a third of the biggest M5 candle in the move. If it's choppy, it's not an impulse.

What counts as the first pullback. Price retraces between 30% and 60% of the impulse, measured on M5 closes. Less than 30% isn't a pullback, more than 60% suggests the impulse may be finished.

Entry trigger. An M5 candle that closes back in the impulse direction from within that retracement zone. I enter on the next candle open.

Stop. Below the pullback low (for a long) plus 3 pips for spread and noise. On GBPJPY that's often 15 to 25 pips, so size comes down accordingly.

Target. The impulse high for the first half of the position, and a move equal to the impulse for the rest, with the stop moved to break-even once the first target is hit.

Over 70 logged trades, it's my most profitable setup per trade, and my least frequent.

Hi LondonScalper,

First off, massive respect for actually logging your trades, refining the data, and having the discipline to sit on your hands while waiting for this specific setup. Trading GBPJPY (the "Dragon") on lower timeframes is a widow-maker for people who lack strict mechanical rules, but your parameters here are incredibly sharp. Filtering for a clean 40-pip sprint in 30 minutes specifically during London hours is a brilliant way to ensure genuine institutional volume is driving the move, rather than algorithmic chop.

I approach Guppy slightly differently, but reading your breakdown got me thinking about how our styles could intersect. My bread and butter is a strictly top-down approach: I rely entirely on the H4 and D1 charts to map out my macro Support and Resistance zones, and then I drop to the M15 for the actual execution.

[PTScalper]

Here is how I view your M5 impulse strategy through the lens of my HTF (Higher Time Frame) + M15 approach, and a few tweaks I'd consider if we blended the two:

Context is King (The D1/H4 Filter)

Your edge is clearly in momentum continuation, but I'd be curious to see what happens to your win rate if you overlay those 70 trades onto H4/D1 key levels. GBPJPY loves to run, but it also loves to aggressively fake-out at major daily inflection points.

If your M5 40-pip impulse fires off bouncing away from a major H4 support zone, that is likely an A+ high-probability setup.

Conversely, if that 40-pip impulse is driving directly into a D1 resistance wall, the 30-60% pullback you are buying might actually be the start of a massive HTF reversal, dragging you out by your stop loss. Mapping those D1/H4 zones could act as a filter to keep you out of the few trades that fail.

[PTScalper]

M15 vs. M5 for Execution

I completely agree with waiting for the 30% to 60% retracement. Anything shallower is just a pause; anything deeper on this pair usually means the momentum is dead.

However, I prefer the M15 chart for execution because it filters out the intra-candle noise that can happen on the M5. When I wait for an M15 candle to close back in the direction of the impulse (like a strong M15 engulfing or pin bar rejecting the 50% level), I find it carries more weight. The trade-off, of course, is that my M15 entry is usually a bit later than your M5 entry, which means my stop loss has to be slightly wider to clear that pullback low safely. But for my psychology, the reduction in whipsaws is worth the slight hit to the Risk:Reward ratio.

[PTScalper]

Trade Management

Your exit parameters are spot on. Taking 50% off the table at the impulse high and moving the stop to break-even is the holy grail of stress-free trading psychology. Leaving a runner for the 100% measured move makes total sense on GBPJPY, given its tendency to trend heavily once it picks a direction for the day.

Since you mention this setup is highly profitable but infrequent, sticking to it is exactly how you survive this game long-term. You've essentially built a sniper strategy. If you ever feel like backtesting a variation, try drawing your major zones on the H4/D1, wait for price to enter that zone, and then look for your exact M15/M5 impulse rules to trigger the reversal away from it.

Great stuff. See you in the charts!

[FXS Prop Desk]

The idea of overlaying H4/D1 levels on the 70 trades is the right next test. To make it fair, the levels should be defined before looking at results, e.g. previous day high/low and weekly open only, rather than drawn by eye afterwards.

Then split the trades into "impulse away from level" vs "impulse into level" and compare the average R. If the difference holds on 70 trades, it's worth collecting another 70 to confirm. GBPJPY's volatility makes small samples especially noisy.