Opening Range Breakout: The Complete Guide

What the opening range is

The opening range is the high and the low of the first 15 minutes of a trading session, measured on 1-minute bars from the fixed session open. That is the entire definition. The window starts when the clock says the session starts, it ends 15 minutes later, and the box it produces never moves again for the rest of the day. There is nothing to draw, nothing to interpret, nothing to redraw. Two people running the same clock get the same box, every day.

An Opening Range Breakout (ORB) is what happens when price leaves that box with force. Not touches it. Not wicks through it. A 1-minute bar closes beyond the range high or the range low, and the volume on that bar confirms real participation. That continuation is the whole strategy: define the range, wait for the break, ride the follow-through.

The core engine is five rules, and it is complete on its own:

• Fixed windows. Four sessions, set by the GMT clock: Sydney 22:00, Asia 00:00, London 07:00, New York 13:30.
• Fixed box. The opening range = high/low of the first 15 minutes of each session. Fixed once the window closes.
• Confirmation. A 1-minute bar close beyond the range (wicks do not count) plus tick volume at least 1.5x the average of the prior 20 bars. Both required.
• Targets. 1R and 2R, where 1R = the distance from the confirming close to the opposite side of the range.
• Invalidation. The opposite side of the opening range. Beyond it, the idea is wrong.

Everything else in this guide — expected move, gamma walls, whale flow — is context layered on top of that engine. Context grades a signal or vetoes a target. It never generates a trade. Keep that separation clean and the method stays testable.

Why opening auctions matter

Every session opens with a burst of orders. Overnight positioning meets fresh flow: hedges get adjusted, stops from the prior session get filled, new directional money arrives at a fixed, known time. For 15 minutes, all of that disagreement gets compressed into one price band. The high of that band is where buyers gave up paying more. The low is where sellers gave up hitting bids. That band is the day's first honest reference.

When price later closes beyond the band on elevated volume, one side has been absorbed and beaten. The trade thesis is simple: force that clears a fresh auction extreme, backed by participation, tends to continue. When price clears the band on thin volume, nobody with size agreed — and those breaks routinely snap back into the box. The method exists to separate the first case from the second, mechanically, once per closed 1-minute bar.

The reason the window is fixed matters as much as the window itself. A range that starts exactly at the session open, every day, produces a sample you can measure. A range you place by feel produces a story. This engine trades the sample, not the story.

The four session windows (GMT)

All session math is done in GMT. The engine converts GMT to broker server time internally and handles daylight saving, so the definition never drifts. These are the canonical windows:

Three practical notes. First, Sydney crosses GMT midnight; the engine handles the rollover, but if you chart it manually, remember the session that opens 22:00 Monday closes 07:00 Tuesday. Second, sessions overlap — at 07:30 GMT both the Asia session (open since 00:00) and the London session (open since 07:00) are live, each with its own box. They are tracked independently. Third, the windows are defined in GMT, not in local or exchange time. When your platform shows a different clock, convert; do not move the definition.

The instruments the engine covers: XAUUSD (Gold), XAGUSD (Silver), and US100 / NQ / USTEC (Nasdaq 100). Same rules on all three. What changes per instrument is typical range size and which sessions carry real volume — covered in section 8.

The box: high, low, mid

At session open + 15 minutes, the box is done. It has four numbers:

• OR high — the highest price traded in the 15-minute window. The long trigger level.
• OR low — the lowest price traded in the window. The short trigger level.
• Mid — (high + low) / 2. A quick control gauge: price holding above mid inside the box leans long; below leans short. The mid is published context, not a trigger.
• Range — high minus low, in points. This number is your future 0.something-R or multiple-R; it drives whether the eventual trade is worth taking at all.

The engine posts each box to Discord the moment the 15-minute window completes: OR high, OR low, and range in points, per instrument, per session. The website levels feed publishes the same per-symbol, per-session high/low/mid with a status field — pending (window not complete), active (box set, no break yet), complete (session played out), or invalidated. Source on every level: ORB Flow Engine.

Range size is a filter you apply with your eyes before any break happens. A Sydney gold box 1.2 points wide is a fakeout trap: the stop is tiny, noise alone can cross it, and one spread widening ruins the R math. A New York US100 box 180 points wide on a news day is the opposite problem: 1R is enormous and the 2R target may sit outside anything price can plausibly reach that day. The box itself is never adjusted for this — the box is the box. What adjusts is your decision to trade it and your size (section 9).

Confirmation: close plus volume

A breakout is confirmed only when two gates pass on the same closed 1-minute bar. The engine evaluates once per closed bar — never intrabar, never on a tick.

Gate 1 — the close. The bar's closing price must be above the OR high (long) or below the OR low (short). The wick is irrelevant. If the OR high is 20,150 and a bar spikes to 20,158 but closes at 20,147, nothing happened. That bar is a rejection, not a breakout, and the engine ignores it. Stop hunts live in wicks; acceptance lives in closes.

Gate 2 — the volume filter. The breakout bar's tick volume must be at least 1.5x the average tick volume of the 20 bars immediately preceding it. Not the session average, not the daily average — the 20 bars right before the break. This is a local test: did participation expand at the exact moment price cleared the level? Every alert reports the actual ratio (e.g. "2.3x average") so you see how emphatic the break was. The rule as taught: no volume confirmation, no trade. A breakout without volume is a rumor, and rumors snap back into the box.

Note the fourth and fifth rows. Each session fires at most one long and one short per day. Once a side breaks, it latches — no re-triggering on every subsequent bar that also closes outside. If a bar closes above the high on 1.1x volume, no signal fires, but the side has not latched either; a later bar that closes above with 1.5x+ volume still confirms. Both gates, same bar, first time it happens: that is the signal.

Targets: the 1R/2R math

R is defined mechanically, before anything else is discussed:

• Entry = the close of the confirming bar.
• Stop = the opposite side of the opening range (OR low for longs, OR high for shorts).
• 1R = |entry − stop|, in points.
• Target 1 = entry + 1R (long) or entry − 1R (short).
• Target 2 = entry ± 2R.

Worked example, New York, US100. The 13:30–13:45 GMT window prints OR high 20,150 and OR low 20,110. Range = 40 points, mid = 20,130. At 13:52 an M1 bar closes at 20,162 — above the high — on 2.1x the prior 20-bar average volume. Both gates pass. The math:

Notice 1R is 52 points, not the 40-point range width. The confirming close landed 12 points above the high, so the stop distance includes that overshoot. A violent breakout bar buys a worse R. That is a real cost, and it is one reason a close 30 points above the high on a 40-point box is often a worse trade than a close 3 points above it.

Short example, London, XAUUSD. Window 07:00–07:15 GMT prints high 2,412.4, low 2,408.9 — a 3.5-point box. At 07:41 a bar closes at 2,408.1 on 1.9x volume. Entry 2,408.1, stop 2,412.4, 1R = 4.3 points, Target 1 = 2,403.8, Target 2 = 2,399.5. Every alert embed carries exactly these lines: Entry, Stop with 1R in points, Target 1, Target 2, and the volume ratio.

The teaching layer adds three rules on top of the mechanical targets. First, express every target in R before entry — a wall at 2R is a trade, the same wall at 0.5R is a skip. No R math, no trade. Second, cap targets inside the option-implied expected move; the EM boundary is where hedging pressure stalls price, and asking a target to live beyond it is asking for the improbable. Third, when taking profit, prefer mapped institutional strikes — the walls where open interest and hedging flow actually cluster — over round numbers. Round numbers are decoration; strikes are positioning.

Invalidation and the latch

The trade is invalidated at the opposite side of the opening range. For a long, the stop belongs beyond the OR low; for a short, beyond the OR high. The logic is structural: the box was the auction's disagreement zone, and a full traverse back through it means the breakout side has been absorbed and reversed. There is no "give it room" beyond that. The risk lesson allows one refinement — placing the stop beyond the nearest mapped level instead, when a mapped level sits inside the box — but the default is the far side of the range, and it is what every alert prints.

Two invalidation mechanics worth knowing precisely:

• The latch. Once a side fires, it is done for the day. If the long fires and price then reverses and produces a confirmed close below the OR low with 1.5x volume, the short fires — that is the second and last signal. Both sides used, session finished. The engine will not chop you long-short-long-short through a rotational day; it structurally cannot.
• Session status. The published levels carry invalidated as a status alongside pending, active, and complete. When you see it, the session's structure is spent. Stand down on that session.

The teaching layer adds two situational skips that sit on top of the mechanical rule. A confirmed breakout aimed straight into a nearby gamma wall gets skipped — the trigger was valid, but the first target would have to chew through the level where dealer hedging leans hardest against the move. And a setup firing against the day's institutional flow bias gets labeled "weak" — still a valid signal, still logged, but graded down. Skips and grades are decisions about a signal the engine already produced. They never change what the engine produces.

Session personalities

Same rules in all four windows. Different expectations in each. Calibrate or bleed.

Instrument fit follows from this. US100/NQ is a New York instrument first; its Sydney box is usually noise-sized. Gold and Silver trade both London and New York seriously, with Asia as a secondary window. Nothing stops the engine from computing all four boxes on all three instruments — it does, every day — but your attention and your size should follow the liquidity.

Risk sizing

Size comes from the stop, and the stop comes from the box. The order is fixed: signal → 1R in points → position size. Never the reverse.

The formula: size = account risk per trade ÷ (1R in points × value per point). Decide the account risk as a fixed fraction — 0.5% to 1% per trade is the framework taught in the risk lesson — and let the box set the divisor.

Using the US100 example from section 6: account $10,000, risk per trade 0.5% = $50. 1R = 52 points. On a CFD worth $1 per point per lot, size = 50 ÷ (52 × 1) ≈ 0.96 lots... except that instrument's minimum step may force 0.9 or 1.0 — round down. If the same day had printed a 150-point 1R, size drops to 0.33 lots. Same $50 at risk either way. The wide-range day does not get a wider ambition; it gets a smaller position.

Three rules that keep the framework honest:

• The stop is set by structure, never by size. If the 1R the box demands makes your minimum lot size exceed your risk budget, the trade is a skip. You do not tuck the stop inside the range to make the size work — a stop inside the box sits inside the noise the box exists to measure.
• One risk number, decided before the session. Not resized after a winner, not doubled after a loser. Expectancy — (win% × avg win) − (loss% × avg loss) — only means anything across trades of constant risk.
• Journal in R. Session, setup grade, planned R, realized R, every trade. R makes a Gold trade and a US100 trade comparable in one column. Dollars hide the process; R exposes it.

Flow context: confluence, not trigger

The engine's trigger uses four inputs and only four: the fixed session clock, the 15-minute range, the M1 closing price, and tick volume against a 20-bar average. No moving averages, no oscillators, no pattern recognition. The institutional option flow layer — branded the ORB Flow Engine — sits entirely on top. It answers one question the trigger cannot: is this breakout worth taking, and where does it likely stall? It never fires a trade.

What the layer contributes, and where each piece plugs in:

• Expected move — the option-implied daily range. Price closes inside it roughly 2 days in 3. Use: cap targets inside it; treat its boundaries as stall levels. A 2R target beyond the EM boundary gets trimmed to the boundary.
• Gamma regime (GEX) — terrain, not prediction. In positive gamma, dealer hedging dampens moves: expect breakouts to grind and stall at walls. In negative gamma, hedging amplifies: follow-through improves. The flip point marks the regime border — above it, take profits early at the walls; below it, let runners run.
• Call wall / put wall — the maximum and minimum GEX strikes. The call wall is a lid and a magnet: a take-profit for longs. The put wall is support: a maximum target for shorts. A confirmed breakout aimed into a nearby wall is a skip.
• Smart-money strikes — up to three strikes showing repeated flow over 7 days. Preferred take-profit levels over round numbers.
• Flow bias — Bullish/Bearish/Neutral from the calls-vs-puts ratio (1.5x threshold with net flow sign), with deep-ITM structural flow (collars, rolls, synthetics) filtered out first; when the weighted score is too small, bias is vetoed to "Neutral — structural flow filtered" rather than guessed. A breakout against bias is labeled weak, not blocked.
• Whale prints, sweeps vs blocks, ask vs bid — one $1M+ print is a data point, not a signal. Repetition, aggression (sweeps over blocks), and side (ask-bought vs bid-sold) turn prints into context.

The hierarchy is strict. The core engine produces every signal and is complete without any of this. Flow grades signals, places take-profits at real positioning instead of round numbers, and occasionally vetoes a target or a trade. If you strip the entire flow layer away, the method still stands: fixed box, confirmed close, 1.5x volume, 1R/2R, stop at the far side.

Common mistakes

Every one of these is a live way to lose money with a correct box on the chart.

• Entering on the wick. Price tags 20,158 against an OR high of 20,150 and you buy the touch. The bar closes 20,147. No confirmation existed; you bought a stop hunt. The rule is a closed M1 bar beyond the level — the wick is noise by definition.
• Trading a close that failed the volume filter. A bar closes 2 points above the high on 1.1x volume and you take it because "it closed above." Half the rule is not the rule. No volume confirmation, no trade — a 1.1x break is exactly the rumor that snaps back into the box.
• Building the box on the wrong clock. Your platform runs broker server time, you chart "the open" at the wrong hour, and your box covers 13:00–13:15 GMT instead of 13:30–13:45. Every level is wrong all day and looks fine. The definition is GMT — Sydney 22:00, Asia 00:00, London 07:00, New York 13:30 — convert your chart, or use the published levels, which are computed DST-safe.
• Re-entering a latched side. The long fired, stopped out, and price closes above the high again an hour later — so you buy again. The engine fires each side once per session for a reason: after a failed break, the level has proven it cannot hold, and repeat entries at the same level are how one planned 1R loss becomes three.
• Moving the stop inside the range. 1R feels too big, so you park the stop at the mid "to improve the R:R." The box is the measured noise band of the opening auction; a stop inside it gets hit by rotation that says nothing about the trade. If structural 1R is too big for your risk budget, the answer is smaller size or a skip — never a closer stop.
• Taking every session with the same expectations. A Sydney gold box is thin, small, and fakeout-prone; a New York US100 box is the day's main event. Same rules, different quality distribution. Treating a 22:15 GMT signal like a 13:52 GMT signal ignores what the sessions are.
• Skipping the R math on targets. "It broke out, resistance is at 20,190, I'll target that." That is 28 points on a 52-point risk — 0.54R. A wall at 2R is a trade; the same wall at 0.5R is a skip. If you did not express the target in R before entry, you did not evaluate the trade.
• Targeting through a wall. The break is confirmed, but Target 2 sits 20 points past the call wall. Hedging flow leans against price hardest exactly there. Cap the target at the wall or skip — hoping the level fails is not a plan.
• Sizing to the instrument instead of the stop. "I always trade 1 lot" works until a 150-point 1R day triples your normal risk. Size = risk budget ÷ (1R × point value), recomputed every trade, rounded down.

FAQ

What exactly is the opening range? The high and low of the first 15 minutes of 1-minute bars after the fixed session open, measured by the GMT clock. It is set once and never redrawn.

What confirms a breakout? Two things on the same closed M1 bar: a close (not a wick) beyond the OR high or low, and tick volume at least 1.5x the average of the 20 bars immediately before it. Both, or nothing fired.

Can both a long and a short fire in one session? Yes, once each. Each side latches after it fires. After both sides have fired, the session is done — at most two signals per session per day.

Which session should I focus on? New York (13:30 GMT open) for US100/NQ — highest volume, best follow-through, full flow context. London (07:00 GMT) for Gold and Silver. Sydney and Asia produce valid boxes but thinner, fakeout-prone conditions; the volume filter rejects more of them, correctly.

What if a bar closes beyond the range but volume is only 1.3x? No signal. The side has not latched, so a later bar that closes beyond the level with 1.5x+ volume can still confirm. But the 1.3x bar itself is not a trade — no volume confirmation, no trade.

Where does the stop go? Beyond the opposite side of the opening range: OR low for longs, OR high for shorts. That distance is 1R, and targets are 1R and 2R from the confirming close.

Is the flow data (expected move, GEX, whale prints) part of the entry signal? No. The trigger is clock, range, closing price, and volume — nothing else. Flow context grades signals, places targets at institutional strikes, and drives skip decisions. It never generates an entry.

Does the bot trade for me? No. The engine is signals-only and places no trades. It publishes levels, breakout alerts with entry/stop/1R/2R and the volume ratio, and educational context. Execution, sizing, and the decision to trade are yours. This is education, not financial advice.

Are the losing signals published? Yes. Every signal, including losers, is logged openly in the results channel. No hype, no guarantees, just data.

How ORB Trading Lab runs this

The method above is automated end to end, and the outputs land in Discord on a fixed schedule:

• Session-open pings at Sydney 22:00, Asia 00:00, London 07:00, and New York 13:30 GMT — the clock the whole method runs on.
• Daily ORB Levels — the moment each 15-minute window completes, an embed per instrument and session with OR high, OR low, and range in points. Confirmed breakout alerts (entry, stop with 1R in points, Target 1, Target 2, volume ratio) go to the Premium channel.
• Morning report — daily at 15:00 Oslo time, before the New York open. One embed per instrument (US100 via QQQ flow, Gold via GLD, Silver via SLV, all levels pre-converted to CFD prices): flow bias, gamma regime, expected move, a color-coded level map with call wall, put wall, and smart-money strikes, and long/short setup lines phrased exactly as the method reads them — break above the opening range high on volume, targets in order.
• Whale alerts — every option print over $1,000,000 premium, checked every 120 seconds, with side, strike, premium, and sentiment.
• One lesson daily at 16:00 Oslo, rotating through the 14-lesson curriculum — from "What is an Opening Range Breakout?" through volume confirmation, GEX, walls, the flip point, sweeps vs blocks, the 1R framework, and expectancy and journaling.
• /flow and /movers — on-demand institutional option flow reads for any US ticker (Premium) and the day's most active symbols.
• Open results — every signal, winners and losers, logged in the results channel.

The standard the community holds itself to is the same one this guide teaches: fixed rules, published levels, R-denominated journaling, and full transparency on outcomes. No hype, no guarantees, just data. Education, not financial advice.