Three core setups
All three assume the gap passed the validation from the identifying chapter, clean geometry, real impulse behind it, and structural context nearby. Skip any setup where you had to convince yourself the gap counts.
Run them in order. The retest entry is the one to master first, it teaches you what a defended level feels like, and everything else builds on that read. Add the breakout-pullback only once you can sit through a full retest without flinching, and treat the order-block confluence as the setup you take when the market hands it to you, not one you go hunting for. Trying to run all three from day one usually means running none of them well.
Wait for price to return to the gap after the initial breakout. Enter on a limit order at the near edge of the gap (the high for bullish, the low for bearish), confirmed by a rejection candle on the entry bar. Stop goes at the far edge of the gap. Target the next structural level.
Best when: the broader timeframe is trending and the gap sits at a level already watched.
Take the breakout itself, then wait for the pullback to the gap's midpoint (sometimes called the “consequent encroachment” level). Enter on a momentum candle with a visible volume lift off that midline. Stop below the gap's far edge. First target: the next liquidity level.
Best when: price is trending hard enough that a full retest to the gap edge doesn't happen.
The highest-probability of the three. An FVG that sits inside an order block, the last opposing candle before the impulse, concentrates two ICT primitives at the same price. Entry logic mirrors Setup 01, but the stop goes at the order-block edge rather than the gap edge. Bigger room, better win rate.
Best when: you have both geometry (the gap) and institutional positioning (the order block) agreeing at the same price.
What a valid entry actually looks like
The three setups tell you where to look; the entry itself lives in the candle that touches the gap. Price tagging the level is not the trade, plenty of gaps get sliced straight through, and a limit order resting at the edge fills you right before the level fails. What you want is evidence the gap is being defended: a candle that wicks into the zone and closes back out of it, ideally on a visible pickup in volume. On a bullish retest that's a lower wick piercing the gap high with a close above it; on a bearish retest, an upper wick into the gap low with a close below.
The cleaner that rejection, the tighter you can work the entry. A decisive rejection candle lets you enter on its close with the stop just past the gap's far edge. A sloppy, overlapping retest, three or four bars grinding sideways inside the zone, is the market telling you the level is being absorbed rather than defended, and those resolve against you more often than not. When in doubt, demand the close outside the gap before committing.
NQ prints a bullish FVG on the 09:45 impulse, gap high 18,242, gap low 18,230. Price runs another forty points, stalls, and rotates back down over the next twenty minutes. At 10:22 it tags 18,241, wicks to 18,236, halfway into the gap, and closes back at 18,243 on the heaviest bar of the pullback. That's the defense.
Entry on the close at 18,243, stop below the gap low at 18,228, fifteen points of risk. First target is the session high at 18,290, roughly 47 points, a shade over 3R before any scaling. The gap held to the tick on the wick; the stop never came close. That asymmetry, small defined risk against a structural target, is the whole reason to wait for the retest instead of chasing the impulse.
All three setups work best aligned with a clean higher-timeframe trend. In chop or rotation, FVG retests fail more often than they hold, because the bigger picture isn't providing the follow-through. If the 1-hour is directionless, consider sitting the session out.
Risk management
The setups don't matter without the structure underneath. Two non-negotiables:
Stop placement
Always beyond the gap's far edge, below the gap low on a bullish entry, above the gap high on bearish. Anywhere tighter and normal market noise takes you out. The gap is either a level or it isn't; the stop location is the bet.
Position sizing
One to two percent of the account per trade. Size for the stop distance, not for a fixed contract count, a wider-stop gap on NQ warrants fewer contracts than a tighter-stop gap on ES.
Managing the trade once you're in
A defined-risk entry is only half the job; how you carry the position decides whether the edge compounds. Once price moves one full R in your favour, move the stop to breakeven, an FVG that's working rarely comes back to your entry, and a gap trade that does is usually failing. From there, scale: bank a portion at the first structural target and let the rest run to the next liquidity level under a stop that trails each higher low. The goal isn't to catch the whole move, it's to lock the high-probability first leg and keep a free runner on the tail.
Resist the urge to widen the stop when price hesitates near the far edge. The stop location was the bet you made at entry; moving it turns fifteen points of risk into thirty and quietly doubles the loss the setup was designed to cap. If the gap is being tested that hard, the thesis is already in question, take the small planned loss and wait for the next one.
Fifteen points of stop on NQ is $300 of risk per contract at $20 a point. On a $50,000 account risking one percent ($500), that funds one contract, round down, never up. The same gap on the MNQ micro risks $30 a contract, so the $500 budget carries sixteen. Drop to a tighter three-point ES gap ($150 a contract) and one percent funds three. Same dollar risk each time; the contract count moves with the stop, not the symbol.
When the retest fails
No setup is a guarantee, and the FVG retest fails in two recognisable ways. The first is the clean slice: price returns to the gap and cuts straight through the far edge without pausing. That's your stop, and it's a good loss, the thesis was tested immediately and answered. Take it and move on.
The second is more dangerous because it's slower. Price stalls inside the gap, coils, and neither rejects nor breaks for several bars. This is the absorption case, resting orders are eating the retest without a decisive reaction. A gap that hasn't produced a rejection candle within three or four bars of entering the zone has lost its edge; the imbalance is being filled in slow motion rather than defended. If you're already in, this is where a breakeven stop earns its keep. If you're still waiting, it's the signal to stand down and let the level resolve before trusting it again.
Failed gaps are also information. A bullish FVG that gets sliced and closes below its low often flips into resistance on the next approach, the same imbalance, read from the other side. The disciplined move after a loss isn't to re-enter the same idea harder; it's to mark the failed level and watch whether it starts rejecting price in the opposite direction.
Conceptual takeaways
- Three setups: retest, breakout-pullback, order-block confluence. Master one before adding the next.
- Wait for the close back out of the gap, a tag is not a trade.
- Stop beyond the far edge, noise takes you out of anything tighter.
- Size for the stop distance, not for a fixed contract count.
- At +1R, move to breakeven; a working gap rarely comes back.
- A gap that stalls inside the zone is being absorbed, stand down.
Next chapter: the psychology of trading an FVG. Every working setup eventually meets a trader who can't follow it, and FVGs are high on the list of setups where execution breaks down first.
The Profit Forecast indicator reads your ATM orders and converts tick distance to dollars and R-multiples live, eliminates the mental math that trips traders up during FVG management.