Breakout–retest continuation

It is 11:20 ET (23:20 ICT). NQ spent four sessions inside 24,860.00–25,060.00, broke above at 10:10 ET, and by 11:00 the auction had accepted the higher prices. Now price is sliding back toward the old top. One trader bought at 11:00, near the high, and watches the open loss grow. Another is flat with a written plan: if the pullback holds above 25,060.00 on two 5-minute closes, go long with the stop under the level. This lesson is about that plan: what the retest adds, what it costs, and what the data says about breakouts from balance.

Key terms

Breakout–retest (phá vỡ rồi kiểm định lại): price is accepted beyond a level, then comes back to it and holds on your timeframe before moving on Definition.

Retest (kiểm định lại): the first pullback to a broken level after acceptance. It is an event you wait for, not one you predict.

Old range (vùng cũ): the balance the breakout left; for this setup, ideally a multi-day balance (M1.02, M2.10).

Term you need: acceptance means time, volume and closes beyond the level, scored with checks written before the probe (M4.01). Without it there is no breakout to retest, only a probe.

Chart first: break, accept, retest

How it works

Two questions, asked in order

The breakout–retest asks the auction two questions, and the order matters.

1. Was the break accepted? Time, volume and closes beyond the level, scored with your M4.01 checks. If not, this is not the setup. A probe that snaps back is the opposite verdict, the failed breakout of M5.01: same level, different answer.
2. Does the first pullback hold? This course’s trigger is a house rule House rule: price trades within 5 points (20 ticks) of the broken level, then two consecutive 5-minute bars close beyond it. A wick, a pause or delta turning your way are things to journal; the two closes are what you score.

The anatomy

• Context: an accepted break of a level mapped before the session, ideally the edge of a multi-day balance, the version the public study tested.
• Trigger: the retest rule above. No pullback to the level, no trigger.
• Invalidation: acceptance back inside the old range, written the same way as the original verdict: a 30-minute close back below a broken high (above a broken low) House rule. A wick into the old range that recovers is a deeper test, not a failure. Your hard stop sits a few points beyond the level so a fast move cannot run past the rule.
• Target: this course writes targets in R (multiples of your risk) or at the next level you mapped (M4.07) House rule. The classic measured move (the height of the balance projected from the breakout) is a ruler for how far the old auction spanned; whether price reaches it more often than any other distance is not yet measured Lore.

Why entry location changes the arithmetic, and what it does not change

Reward-to-risk is arithmetic Definition. A long near the level with a stop just under it risks a few points; the same long bought at the high of the run, with the stop at the same invalidation, risks several times more for the same target. Chasing without a retest means entering at the extreme with a wide stop.

What the arithmetic does not tell you is the win rate. The retest entry is a filter, and filters change who gets in. You never buy the breakouts that run without looking back, and a pullback to the level is also what a breakout about to fail looks like at first. So retest trades are a different sample from “all accepted breakouts”, with a different, unknown hit rate. A 2R trade breaks even at a 33% win rate before costs; a 0.5R trade needs 67% (M6.08). Which entry earns more on NQ over many trades is not yet measured.

Why it is still worth knowing

• Shared vocabulary. “Accepted above, retest held on two 5-minute closes” can be checked after the session; “breakout confirmed” cannot.
• Framing risk. The retest puts the invalidation close to your entry, which keeps the loss per trade small.
• A testable hypothesis. Every part is written as a rule, so you can score it on your own sessions (see the drill).

Myth / hypothesis: “A broken level flips: old resistance becomes new support.” Lore Role reversal has not been tested as stated by our lab or the public study we use. The closest measurement points the other way: in our lab’s recent NQ and ES futures check, prior-day highs, lows and closes held less often than random prices at the same distance Measured: no edgeNQ (details in M4.01). The retest trigger above does not assume the level is special; it asks the auction and waits for the answer.

What the data says

The closest tested claim comes from one public practitioner study on NQ: a breakout from a multi-day balance, kept only if it did not close back inside within 15 minutes. The retest entry taught here is narrower and has not been tested.

Worked example: the long side

Synthetic session (no real date), NQ ticks of 0.25 points, priced for 1 MNQ ($0.50 a tick, $2 a point). It shows the reading, not a signal.

• Map, before the open. Four sessions balanced inside 24,860.00–25,060.00, a 200-point range. Written checks from M4.01: 30 minutes beyond, 1.5× the usual hourly volume beyond, two 30-minute closes beyond. Retest rule: within 5 points of the level, then two 5-minute closes beyond.
• 10:10 ET (22:10 ICT), probe. First trade above 25,060.00. No verdict.
• 11:00 ET, verdict. Closes of 25,081.50 (10:30) and 25,101.50 (11:00), about 50 minutes above the level, volume about 1.6× the usual. Accepted. No trade yet: the price is 41.5 points above the level.
• 11:05–11:25 ET, pullback. High 25,104.75, then a slide. The 11:25 bar dips to 25,062.50, 2.5 points above the level, and closes at 25,071.25 with positive delta. The 11:30 bar closes at 25,076.00. Two closes above after a touch inside 5 points: trigger met.
• The plan. Long 1 MNQ at 25,076.00. Hard stop 25,051.00, 9 points under the level: 25 points = 100 ticks × $0.50 = $50 = 1R. Rule exit: a 30-minute close below 25,060.00 means flat at market, even if the stop has not traded. Target 2R = 25,126.00 (+50 points, $100).
• The ruler. The measured move projects 25,060.00 + 200 = 25,260.00, 184 points from entry, over 7R. That describes the balance’s size; it is not a target Lore.

The chaser. The trader who bought the 11:00 close at 25,101.50 has the same invalidation, so the same stop at 25,051.00: 50.5 points of risk ($101). To 25,126.00 is 24.5 points, about 0.49R. To earn 2R from there needs 25,202.50. Same direction, same level, same idea: the entry location alone changed the arithmetic. It did not change the unknown win rate.

The mirror and the failure

Short side. Turn the session upside down. The balance bottom is 24,860.00; price breaks below at 10:10 ET and is accepted by 11:00. The pullback high is 24,857.50, then two 5-minute closes below the level. Short 1 MNQ at 24,844.00, hard stop 24,869.00 (25 points, $50), target 2R at 24,794.00, flat on a 30-minute close back above 24,860.00.

The failure: the retest held, the breakout did not. Back to the long. At 11:45 ET price drops to 25,054.25, and the 12:00 30-minute period closes at 25,052.75, below 25,060.00. The hard stop did not trade, but the rule exit did: flat at about 25,052.75, a loss of 23.25 points ($46.50, about 0.93R). Nothing was misread: the retest described 11:25 to 11:30, and the close back inside was the auction changing its answer. Exit; do not turn a continuation into a hope trade. If the old range then accepts the move down, M5.01’s failed-breakout plan takes over.

The real October 2026 NQ balance in M2.10 shows the same thing on a daily scale: the first test after the break held above the old range, value built higher for three sessions, and by 8 October price was back inside.

Prop accounts and resting orders. In a prop-firm evaluation your fills are simulated: a fill at the level there does not prove a live order would have filled (MKT2.03). Firms also ban strategies that exploit simulated fills, such as orders or brackets placed to profit from how the simulator fills them. Read the firm’s prohibited-conduct rules before you automate this setup (PROP1.04).

Common mistake: Anticipating the retest: resting a limit at the broken level in advance, on the theory that “breakouts come back”. Sometimes they do not, and you miss the move; sometimes the limit fills on the way down into a breakout that is failing, with no trigger at all. The retest is a trigger, not a prophecy. Wait for the pullback and the two closes House rule.

Common mistake: Treating a better reward-to-risk as an edge. 2R against 0.49R is arithmetic. Whether retest entries win often enough to beat entering on acceptance, or not trading at all, is not yet measured on NQ, and the multi-day balance breakout itself is only a lead. Size it as an unproven setup.

Common mistake: Moving the invalidation after the retest “because it held once”. The retest describes ten minutes; a 30-minute close back inside is new information. Keep the exit you wrote before the entry.

Drill: score twenty breakout–retests

In market replay (PLAT2.06 shows the ATAS steps), work through recent NQ sessions:

1. Before each session, mark any balance of three to five sessions and write the acceptance checks, the retest rule, the invalidation and a 2R target.
2. For every accepted break, log: did a retest within 5 points happen, and when? Did the two closes come? What came first, 2R or the invalidation?
3. Also log every accepted break that never retested, and where price was two hours later.
4. Stop at twenty accepted breaks. Count: retests that triggered, their 2R/invalidation tally, and breaks that never came back.

What we don’t know yet

• Whether the retest entry beats entering on acceptance, or not trading, on NQ after costs: not yet measured.
• How often accepted NQ breakouts come back within a few points of the level before moving on: not yet measured.
• Whether the multi-day balance breakout lead survives a forward test. Our lab’s rule is frozen for NQ data after August 2026; at about 50 events a year, expect “lead” on this page for a long time.
• Whether measured-move projections are reached more often than random distances: not yet measured.

Recap

• A breakout–retest needs acceptance first (M4.01); without it you are looking at a probe, or at M5.01’s failed breakout.
• The trigger is a pullback that holds on two 5-minute closes beyond the level, a house rule you can score; the invalidation is acceptance back inside the old range.
• Entry location changes reward-to-risk arithmetic; it does not tell you the win rate, and the retest filter skips the breaks that never return.
• The closest tested claim, the multi-day balance breakout, is a lead for a forward test on NQ, not an edge; “broken levels flip into support” is lore.

Quiz

Further reading

• Auction Market Theory: Understanding of market and orderflow dynamics: acceptance and balance in auction-theory terms.
• How To Trade A Market Profile Ledge: trading formed structure with context.
• Order Flow, Measured: all claims: the public practitioner study behind the evidence box (claims AP-03 and AP-04).
• Our lab: the multi-day balance breakout rule is fixed in advance for a forward test on NQ sessions after August 2026; prior-day levels held less often than random prices in a recent NQ and ES futures check (46 sessions each, 2026).

Prerequisites: M4.01 Acceptance, M4.09 Order-flow confirmation, EV3.01 Sample size and power. Next: M5.08 Pullback to VWAP or value, the other continuation entry that waits for a pullback. Connects to: M5.01 (the mirror verdict) · M5.10 (the intraday version) · M5.14 (entry patterns with R:R).

Educational content only, not financial advice. Futures trading involves substantial risk of loss. Synthetic examples are illustrations, not trade calls.