Breakout entries

From 09:45 to 11:45 ET (21:45–23:45 ICT, winter) ES sits under the prior high at 6,010 and pokes above it four times. At 11:55 ET a 5-minute candle closes at 6,014; the next pushes to 6,022 with the widest range of the morning. Is this the breakout to buy, or the fifth fake-out? This lesson answers with a rule written before the probe, not with hope, and it is honest about the part nobody can answer yet: whether the rule has an edge.

Key terms

A breakout entry (vào lệnh phá vỡ) buys or sells beyond a defined level (prior high or low, initial-balance extreme, balance edge) once the auction shows it accepts the new price Definition. Not on the first tick through it. One tick through a level is a probe (thăm dò); acceptance (chấp nhận) is the auction continuing to do business beyond the level, measured by checks you wrote down in advance (M4.01).

How it works

The acceptance checklist

This course’s breakout rule is a house rule House rule: one required check and at least one supporting check, all written before the probe and read on the timeframe you trade.

• Required: a full candle close beyond the level, at least 2 ticks beyond it.
• Plus at least one of: the break candle’s range is at least 1.5× the average of the last 10 candles; volume traded beyond the level is at least 1.5× the usual for that time of day; a retest of the level that holds.

The numbers are conventions that make the check scorable, not measured cut-offs. The window has to match the holding time: a 5-minute close supports a trade held for minutes, not one held all day (M4.01 explains why). M4.09 writes the same idea as a full 30-minute rule with a delta check and a scorecard column, for traders who hold longer.

Where the stop goes

The stop sits back inside the broken structure: below the breakout level on a long, usually at the midpoint of the break candle or the prior balance edge House rule. Logic: if price returns inside, the acceptance thesis is dead. Exit, no debate. A close stop keeps each attempt small; it says nothing about how often the stop is hit.

Two failure modes

Failure mode 1: the false break (phá vỡ giả). Price pokes through and snaps back. Contained by the checklist, by keeping standard size (never “extra conviction” size on a breakout, whose edge is unproven), and by honouring the stop instantly House rule.

Failure mode 2: the late break. Entering after the move has already extended. The rule is mechanical: if the distance from the level to your entry exceeds your planned risk before you enter, the breakout entry is cancelled House rule. Switch to the pullback family (M6.03) or stand down. Chasing (đuổi theo giá) is not a breakout entry; it is failure mode 2 with a story.

Myth / hypothesis: “Breakouts work in initiative regimes (open drives, trend days, post-news acceptance) and bleed in rotational balance, where every break gets faded.” Lore It is a sensible picture, but it has not been shown to help on NQ. In a public practitioner study, initiative activity did not continue more often than baseline (+1.1 percentage points, interval −0.5 to +2.7, n = 15,851), lunch-hour breakouts did not fail measurably more than morning ones, and the day type could not be forecast by 11:00 ET from the open and the initial balance. A trend day is a label you can give only at the close. Treat the regime idea as a hypothesis to log, not a filter you already know works.

Why it is still worth knowing

• Shared vocabulary. “Probe, no close beyond, no entry” replaces “it looked strong”.
• Framing risk. The broken level gives you an objective invalidation close to the entry, and the late-break rule caps how much you pay for the move.
• A testable rule. A required close, one supporting check and a fixed stop can be scored on replay and in your journal. That is the only honest route from “unproven” to “measured”.

What the data says

The popular promise of the breakout is that once a level breaks, price keeps going. On NQ that promise has been tested in several forms by one public practitioner study, and none of them was supported.

Outside NQ futures, the closest public result is on the Nasdaq-100 itself: a working paper on QQQ (Zarattini and Aziz, 2023) and one public replication found opening-range breakouts in the direction of the first 5-minute candle earned about +0.13R before costs and about zero after them PlausibleNasdaq index. NQ futures are extrapolated, and our lab has not replicated it yet.

So the honest status of this entry: a rule-based way to join a move with a defined stop, whose edge is unproven. You can trade it small while you score it. You cannot call it high odds.

Worked example: the 18,010 break

Synthetic session, NQ (0.25-point ticks, $5 a tick, $20 a point); it shows the reading, not a signal.

• Map, written before the open. Overnight balance 17,980–18,010; the prior day’s high is 18,036. Scenario: if a 5-minute candle closes at least 2 ticks above 18,010 and one supporting check passes, buy; no trade on a probe.
• 09:45 ET (21:45 ICT). A 5-minute candle closes at 18,014: 4 points above the level (required check ✓).
• 09:50 ET. The next candle has the widest range of the morning, 2× the 10-candle average, and trades to 18,022 (supporting check ✓).
• Entry. Buy 18,016. Late-break check: 18,016 − 18,010 = 6 points from the level, less than the 10-point planned risk ✓.
• Stop and target. Stop 18,006, back inside the balance: 10 points = $200 on 1 NQ. Target 18,036, the next mapped reference (the prior day’s high): 20 points = $400. Projecting the balance height (30 points) is a common alternative; how often such projections are reached on NQ has not been shown to beat a baseline.
• Break-even. With $15 of round-turn costs, a win nets +$385 and a loss −$215, so the trade needs to win 215 ÷ 600 ≈ 36% of the time to break even (M6.08). How often this rule reaches its target on NQ is not yet measured, so you do not know whether it clears 36%.
• Outcome on this synthetic day. Price holds above 18,010 and reaches 18,036 in 40 minutes: +$385 after costs. One day is one outcome, not evidence.
• Contrast. At 10:40 ET (22:40 ICT) NQ spikes to 18,060 in one vertical 3-minute move, with no 5-minute close beyond any new level. No entry: that is a probe. Chasing it at 18,058 would put the entry 48 points above the broken level, far more than the planned risk, so the late-break rule cancels it. Same direction, opposite decisions: the written rule decided.

The mirror and the failure

The short side (mirror). Below a mapped support the rule reads upside down: a 5-minute close at least 2 ticks below the level plus one supporting check, a stop back above the level, and the same late-break cancel. Flip the figure to see it. The public tests above covered breaks in both directions.

The failure. Same map, different morning. The 09:45 candle closes at 18,014 and the 09:50 candle expands to 18,022: both checks pass. You buy 18,016. The 10:00 ET candle closes at 18,004, back inside the balance, and the stop at 18,006 fills: −$215 after costs. Nothing was done wrong: the rule was written, followed and exited without debate, and it lost. If the break has now failed, the idea to study is the opposite one, the failed breakout (M5.01), which has no edge shown on NQ either.

Prop-firm note. In a prop evaluation the stop must be a real order placed with the entry. Evaluation accounts usually fill in simulation, and simulated fills are not live fills; tuning orders to exploit simulated fills is treated as prohibited conduct by firms such as Topstep. The Prop Firm course (PROP1.04) covers the rules; check your firm’s help center before you trade.

Common mistake: Buying the first tick through the level and calling it a breakout entry. The first tick often arrives as a sweep, and on NQ sweeps through three or more price levels continued slightly less often than comparable bursts over the next minute PlausibleNQ (forward test still pending; outcomes measured at trade prices, with the bid-ask bounce checked on only 3 days; the effect is smaller than a round trip’s costs, so it is not a trade). The bigger point is simpler: at the first tick no check has run, so there is nothing to call acceptance.

Common mistake: Treating a passed checklist as an edge. The checklist decides when you are allowed to enter and where you are wrong; it does not change the odds you have measured, because nobody has measured them yet. A breakout that is easy to see and easy to name is not a base rate.

Drill: score ten probes on your own chart

Open five recent NQ sessions in replay on 5-minute bars and mark the overnight high, overnight low, prior-day high and prior-day low before 09:30 ET.

1. For the first ten probes across those levels, log: time, level, whether the 5-minute bar closed at least 2 ticks beyond the level, whether at least one supporting check passed, and the distance from the level to the close.
2. Compare that distance with a 10-point planned risk to see whether the late-break rule cancels the entry.
3. Count how many of the ten probes would have qualified as an entry under the written rule, and how many would have been filtered out as probes or late breaks.

Recap

• A breakout entry requires a mapped level, a close beyond it and at least one supporting check, written before the probe; a tick through is a probe, not a trade.
• The stop goes back inside the broken structure; a return inside kills the thesis, so the exit is automatic.
• Two failure modes: false breaks (contained by the checklist, standard size and the stop) and late breaks (cancelled when the level-to-entry distance exceeds planned risk).
• Compute the break-even win rate with costs before the click; on NQ, breakout continuation has no edge shown, and the checklist’s own edge is not yet measured.

Further reading

• Order Flow, Measured: all claims: the public practitioner study behind the evidence box (claims OI-08, AP-03, TE-01).
• Trading Key Market Auction Reversals (Axia Futures): auction logic for when initiative is real and when it fails.
• Our lab’s evidence review: the opening-range breakout on the Nasdaq-100 earns about +0.13R before costs and about zero after them (working paper plus one public replication); our own replication on NQ is on the backlog.

Prerequisites: M6.01 (the four entry families) and M4.01 Acceptance. Next: M6.03, the pullback alternative when you are late.

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