Sweep
Desk liquidity sweep indicator for TradingView

1. WatchStart with a confirmed high or low.
2. SweepPrice passes it, then closes back across.
3. ConfirmA later qualifying close prepares a zone.
4. AcceptDefault: enter at the qualifying confirmation close.

This long setup mirrors for shorts above a high. Defaults: Confirmation close and Single gaps. Close inside and limit entries are alternatives.

Swings: highs/lows confirmed five candles later by default. Equal levels: nearby highs/lows merged within an ATR tolerance. Sessions: extremes available after a session ends. Higher-timeframe swings and previous day/week levels add context.

“Liquidity pool” is the tool’s name for these price-derived references; it does not observe waiting orders. “Strong” means equal, session, higher-timeframe or day/week by type. It does not establish a better outcome.

Any wick strictly beyond a pool consumes it, even without enough depth or a close back inside. Touching exactly does not. Entry filters cannot make consumed liquidity fresh again. Age and budget prune swing-type pools; the drawing-distance setting only hides distant lines.

iFVG, the default: after a sweep, a new close through an opposing gap’s far edge can prepare that gap as the zone. In our long example, exactly 90 fails; above 90 can qualify. The gap must also meet age, location and movement checks.

Default HTF prerequisite: a completed 60-minute FVG must receive its first valid touch. It grants six observed chart bars of directional context, ages zero through five, and cannot refresh on repeated touches. An accepted setup consumes it; a pending return retains its frozen context.

Optional SMT compares synchronized prior closed bars and needs a meaningful manually chosen pair. Momentum, a minimum observation count and confirmation-close macro windows are separate checks. Our definitions are explicit; more checks do not prove better results.

MSS, the alternative: price closes beyond the opposing short-term swing captured at the sweep, with enough movement. A qualifying gap in that shift leg becomes the zone. SHIFTED means it is still searching; ARMED means the zone is ready. Nearest picks its newest gap; Deepest picks its oldest.

Either accepts either route, with iFVG first if both qualify together. Confirmation allows twelve candles after the sweep; the next candle is too late. MSS allows eight extra candles after a shift to find the gap.

Default sweep distance is at least 0.05 ATR and the close must return strictly across the level. Confirmation needs at least 1 ATR of movement; gap width at least 0.1 ATR. ATR measures recent price movement including gaps; its length is fixed at fourteen.

Confirmation close, the default: the qualifying inversion or MSS-zone acceptance close is the entry reference. Its earlier wick cannot also exit that plan.

Close inside, the alternative: a later candle closes within either edge, including exactly on an edge. Its actual close becomes entry. The confirmation candle cannot fill too.

FVG edge / CE midpoint: use a paper limit at the zone’s near edge or center. Contact can fill without a close inside. These are alternative entry rules.

The zone is fixed after arming. Close-inside entry recalculates stop and targets using the actual fill close and current ATR. The pending sweep extreme can extend before confirmation.

Recorded reasons include no confirmation in time, no rejection, invalidation, a close through the wrong side of an inverted zone, price reaching the first target unfilled, or expiry.

A prepared setup has thirty candles to fill. Close inside first checks the frozen pending stop: a close beyond it cancels before entry and cannot move the stop to manufacture a fill. Otherwise, a qualifying fill wins over missed-target and expiry checks, including on the final allowed candle. Once expired, a later return cannot revive it.

Pro labels recorded cancellations. A fresh qualifying sweep can also replace an unarmed idea without a cancellation tag. Armed and filled setups are protected from replacement.

1R is initial entry-to-stop distance. Default stop: sweep extreme plus 0.1 ATR buffer. Zone far edge is an alternative. Auto can choose a nearer anchor when the sweep is too far; its 2.5 ATR cap does not constrain the default Sweep extreme setting.

Targets default to 1R, 2R and 3R. TP1 optionally moves stop to entry; TP2 always moves it to TP1. Neither takes partial profit. Protective stop changes apply from the next bar, after the current bar is checked against its existing stop. Breakeven names an entry-price reference; costs can still leave a loss. This tool has no continuous ATR trail.

Liquidity selects the nearest qualifying fresh opposing pool, skipping closer pools below the minimum distance: 2R by default, checked after rounding toward entry. Clustered quality requires at least two confirmed merged swing observations. Choose Liquidity or TP3 as final; Liquidity falls back to TP3 when none qualifies. The selected final target freezes at acceptance; later pool changes cannot move it.

Close entries have no entry-bar outcomes. On later bars, a stop-gap open is the exit reference; a favorable gap beyond the target uses the fixed target. Ordinary both-touch bars resolve stop-first and disclose ambiguity. Unknown intrabar limit fills never invent a same-bar target success. These paper assumptions are separate from broker fills. Default timeout is two hundred candles.

State shows the stage and timers; Pool / Confirm / Zone identify the setup. Entry / Stop / R ladder / Liquidity describe its plan. Record / By pool / By model summarize closed paper outcomes without costs.

1. SweptSelected qualifying sweep ideas.
2. ConfirmedIdeas reaching confirmation.
3. ArmedPrepared entry zones.
4. FilledRecorded paper entries.

These counters exclude sweeps blocked by enabled filters and new sweeps while armed or filled. Sweep quality is descriptive; it gates nothing.

Co-Pilot explains NOW, WHY and PLAN from the same engine. Pro adds pool, sweep, structure, zone and cancellation labels; Full adds session shading. It does not draw every stored gap.

Four display defaults differ, labelled Indicator and Lab. The Lab has its own broker dashboard; other defaults are shared.

Strategy Lab only

Lab sizing

Cash risk

Cash risk in account currency

100

Fixed units alternative

1

Maximum notional versus current equity

90 percent

Quantity increment

Symbol minimum

Manual quantity increment alternative

1

When required, volume must be positive with a positive twenty-bar mean. Zero or unavailable volume does not pass. The separate observation count is descriptive, not a private grade or probability.

The indicator has twelve named conditions: long/short armed, long/short filled, TP1, TP2, final target, stopped, cancelled, liquidity swept, FVG inverted and higher-timeframe pool swept. Final target covers TP3 or liquidity.

Any alert() function call sends only ARMED, FILLED, TP1, TP2, EXIT and CANCEL. It does not send separate sweep or inversion payloads. Use Once Per Bar Close.

Both scripts send JSON or readable text. The sweep-desk.v2 schema combines same-bar paper events with entry, initial/current stop, target, times and observed exit. These are not broker fills. The twin submits its bracket with entry and owns its position and deadline; a paper EXIT does not close it. Default Confirmation close uses on-close processing with declared slippage. Recreate alerts after source or setting changes.

1. Keep defaults and follow the state through a complete session.
2. Include waiting, cancelled and losing examples.
3. Load the twin on a second chart. Set realistic size, costs and slippage.
4. Change one input at a time, then check a separate period.

Lab defaults: USD 100,000 capital, cash risk 100, maximum notional 90 percent of equity, commission 0.02 percent per order and five ticks of slippage. Fixed units is an alternative. The broker emulator can differ from the indicator’s cost-free reference ledger. Review Properties and quantity increment for the instrument.

Use the current Lab and its complete trade list. The older Strategy Tester picture belongs to the historical v1 article and is not a result for this release.

The current v2 MNQ Strategy Lab sample lost money after costs. That small observation does not establish an edge.

The linked gold and bitcoin measurements belong to the historical v1 engine. Its close-inside baseline is not evidence for current defaults; v2 revises entry timing and broker ownership.

Keep historical findings and current validation separate. Software checks verify behavior, not economic edge.

Signals commit on closed candles. A live candle can change before its close. Different history, feeds, timeframes or inputs can recalculate results. Pivots appear after confirmation; session pools after their session ends. Higher-timeframe context uses its last closed candle.

The engine sees prices, not news, resting orders or execution. Its paper record excludes costs. It does not predict prices. Trading risk is yours; this is research tooling, not financial advice.