SMB CapitalIndicator Playbook AM ATR v4.21-CL

Adaptive Mean + ATR Bands

It measures how far price has wandered from its own moving average, counted in how many normal days of movement that distance represents — then trades the two moments that distance becomes useful. Where a setup fires it marks the entry, prints the stop, and tracks the trade to its exit.
Basis 5DMA intraday · 20DMA daily+ Scale Daily ATR(20) Signal Daily close Execution Any intraday chart Status Live · not backtested
0 / 0 ↑ Top
1

What this indicator watches

Start here
One measurement. Everything else on the chart is built from it.
In one sentence
It measures how far price has wandered from its own moving average, counted in how many normal days of movement that distance represents — then trades the two moments that distance becomes useful.

The three things it reads

It looks atWhich means
A moving averageThe 5-day average on any intraday chart, the 20-day on daily and higher. Called the basis or the mean throughout. It is the middle price recently traded around.
The daily ATRHow much this stock typically moves in one day, top to bottom. MSTR moves about $6.83. A $30 stock might move $0.60. It is the stock's own yardstick.
The gap between themPrice minus the average, divided by that daily range. This is the number the whole tool runs on.
A worked number
MSTR trades at 119. Its 5-day average sits at 105. It moves about 6.83 in a normal day.
Price is 14 points above the average. Divide by 6.83 and you get 2.05. So MSTR is sitting two full days of normal movement above where it has been trading. Not "up 13%" — two days of travel away from the middle. That is what the tool means by extended, and it means the same thing on a $30 stock and a $900 one.

The two moments it trades

AVERAGE +2 −2 FADE — too far from the average, turning back. Bet on the return. RECLAIM — came back and held. Bet it keeps going.
Both trades use the same measurement. One enters far from the average, one enters at it.
SetupFires whenYou are betting
FadePrice got more than 2 days of travel from the average, then closed back toward itThe stretch is finished and price returns to the middle
ReclaimPrice spent days on one side of the average, dug in, then closed back through itThe return is real and continues past the average

What it does once one fires

It is not an alert-only tool. When a setup triggers it marks the entry bar, prints the stop price, calculates the targets, and then tracks that trade to its exit — printing a label and firing an alert at every scale-out and at the stop. Each trade gets a number so you can tell one from another on a busy chart.

What it does not do: size the position, check the trend, read the news, or close anything at the bell. Sections 8 and 12 are explicit about where the tool stops and you start.

One naming warning. This indicator's "reclaim" is not the desk's Pullback & Reclaim checklist — same idea, different entry rules, so the two will disagree about which trades exist. Section 13 lays out exactly how, and how to use both together. Read it before you assume a signal here would pass the checklist.

2

The one number

Concept
Everything on the chart resolves to this.
Extension
extension = (close − basis) ÷ daily ATR
How many normal days of travel price sits away from its own mean. A reading of 2.0 means price is two average days' range from the middle. Everything — the bands, the tags, the state, both setups — is quoted in this one unit.
+4 +3 +2 +1 MEAN −1 −2 NORMAL STRETCHED EXTENDED back toward the mean
The shaded band is normal territory — inside ±1 ATR. Each rung out is one more average day of travel from the mean.
ReadingStateWhat it tells you
under 1NORMALDrifting. Nothing stretched about it.
1 – 2STRETCHEDMoving with purpose. Not yet a fade.
2 – 3EXTENDEDMean-reversion odds start to matter. This is the fade zone.
3 +EXTREMERare, and rarely sustained into the close. Chasing here pays for someone else's move.
The ±N ATR tags are not signals

They mark the first time each tier is reached in a session and nothing more. They fire whether or not a trade exists, and they are tied to no entry, stop or target. Read them as a ruler.

3

Setup A — the Reclaim

All required
Price left the mean, went far enough to matter, and came back. Entry is AT the mean.
MEAN STOP — the pullback extreme depth ≥ 0.5 ATR ENTRY — first close that holds 2+ daily closes below the mean
The bowl is everything between leaving the mean and coming back. Its deepest point becomes the stop.
Two or more daily closes on the wrong side Automatic Price must have closed below the mean at least twice before a reclaim can arm. One-day flips are rejected outright — they print a grey tag, not a trade. Setting: Min daily closes, default 2.
The pullback reached at least 0.5 ATR past the mean Automatic A touch is not a pullback. The excursion has to have depth or the reclaim means nothing. That deepest point is the stop. Setting: Reclaim depth, default 0.5 ATR. Never set it to 0.
The mean is flat or turning your way Automatic No buying the stern of a boat still sinking. A reclaim into a mean that is still falling is rejected. Setting: Slope gate, on by default.
A daily close crosses back over the mean Automatic The trigger itself, evaluated once per day at the rollover. This is a daily decision on any chart — a 2-minute chart does not look for 2-minute reclaims.
The next session's first confirmed bar holds the right side Automatic The signal arms silently and waits. A session that gaps and fails at the open never prints an entry — it prints a fail tag instead. This is where your fill actually happens, and the reason a 30-minute chart fills 28 minutes earlier than a daily one.

Your judgement still sits on top of this. The indicator does not check trend, coil, MA cluster, sector, or news. Run the desk checklist for that. These five lines are only what the code enforces.

Both sides, spelled out

MEAN LONG — pull back BELOW, close back ABOVE stop = the LOW of the pullback entry MEAN SHORT — rally ABOVE, close back BELOW stop = the HIGH of the rally entry
Identical machinery, mirrored. Everything the long does below the mean, the short does above it.
LongShort
Persistence2+ daily closes below the mean2+ daily closes above the mean
DepthDipped ≥ 0.5 ATR belowRallied ≥ 0.5 ATR above
Slope gateMean flat or risingMean flat or falling
TriggerDaily close crosses up through the meanDaily close crosses down through the mean
ConfirmFirst bar of next session closes above the meanFirst bar closes below the mean
StopThe lowest low of the pullbackThe highest high of the rally
T1Next whole band above entryNext whole band below entry
T2Close below a falling trailClose above a rising trail
RunnerDaily close back below the meanDaily close back above the mean
Void (day 1)Close below mean − toleranceClose above mean + tolerance
Label#7 ▲ LONG#7 ▼ SHORT
4

Setup B — the Break-confirm

Second chance
The reclaim was rejected by a filter. Structure gets one chance to prove it anyway.
MEAN ARMED — prior session high reclaim filtered — no trade close through the level = ENTRY
A dotted grey line appears on the chart at the armed level. It either breaks within the window or expires silently.
A reclaim printed but a filter killed it Automatic Too shallow, wrong slope, a one-day flip, or it failed at the open. Any of the four arms this path.
The prior session's high (longs) or low (shorts) arms as a tripwire Automatic Drawn as a dotted grey line. If you see one on the chart, a rejected setup is waiting on structure.
A confirmed close takes that level within 3 sessions Automatic The mean must still be non-opposing at the trigger bar. Otherwise the level expires and nothing prints. Setting: Break-confirm window, default 3.

Turning this off cuts your trade count materially. It is the second of the two entry paths, and it is the one that catches turns the strict reclaim filter refuses.

Both sides, spelled out

LongShort
Armed levelThe prior session's HIGHThe prior session's LOW
TriggerConfirmed close above that highConfirmed close below that low
Mean check at triggerFlat or curling up over the last quarter-sessionFlat or rolling over
Cancelled ifA daily close returns below the mean, the 3-session window expires, or the short side armsA daily close returns above the mean, the window expires, or the long side arms
StopThe pullback low, capped at 2 ATRThe rally high, capped at 2 ATR
Label#7 ▲ LONG (brk …)#7 ▼ SHORT (brk …)

The exits are the same ladder as Setup A — this path changes only how you got in, never how you get out.

5

Setup C — the Fade

Opposite trade
The mirror image of the reclaim. Entry is FAR from the mean, betting on the return.
Why this is not the same trade
The reclaim enters at the mean, betting the return continues. The fade enters two ATR away from it, betting the stretch is finished.
Same measurement, opposite ends of it, opposite direction of travel. They are separate state machines and both can be open at once — which is why fades carry F-numbers and reclaims carry #-numbers.
MEAN +1 +2 STOP — the extreme of the stretch ENTRY — closed back inside +2 T1 — half off at +1 T2 — remainder at the mean
The fade says the move is exhausted. Price making a NEW extreme past that point says it is not — which is exactly where the stop sits.
Extension pushed beyond ±2 ATR Automatic The stretch has to be real before there is anything to fade. The furthest point reached becomes the stop.
It closed back INSIDE ±2 Automatic The re-entry is the trigger, not the stretch. A confirmed close back through the band — that bar is your entry.
The MA state agrees Automatic The filter MA must be rounding the right way, or a timeframe break must have occurred. Rejected fades print a grey tag if you enable them. Setting: Fade filter.
No open reclaim in the opposite direction Automatic The regime gate refuses to fade against your own live position. Setting: Regime gate, on by default.
You have decided the exit is yours if it stalls Manual Nothing closes this at the bell. It ends on its stop or at the mean, however long that takes. If you intend to be flat by the close, that flatten is yours to execute.

Both sides, spelled out

MEAN +1 +2 FADE SHORT — stretched ABOVE +2, closes back inside stop = stretch HIGH entry T1 +1 T2 mean MEAN −1 −2 FADE LONG — stretched BELOW −2, closes back inside stop = stretch LOW entry T1 −1 T2 mean
Both fades travel toward the mean. The stop always sits at the furthest point the stretch reached.
Fade shortFade long
StretchExtension pushed above +2 ATRExtension pushed below −2 ATR
TriggerConfirmed close back below +2Confirmed close back above −2
FilterMA rounding down, or a break of the prior lowMA rounding up, or a break of the prior high
Regime gateBlocked while a long reclaim is openBlocked while a short reclaim is open
StopThe highest high of the stretchThe lowest low of the stretch
T1 — half offThe +1 bandThe −1 band
T2 — remainderThe meanThe mean
LabelF3 ▽ FADE SHORTF3 △ FADE LONG
6

Exits

Automatic
Every one of these prints a label and fires an alert. You do not have to watch for them.

Every exit below is symmetric. Read "beyond" as above for a long and below for a short; a falling trail for a long is a rising trail for a short. The code runs one set of rules and flips the comparison on playDir.

Reclaim & Break-confirm — the ladder

30%Tranche 1
At the next whole ATR band beyond entry Promotes upward so the first target is never behind the fill — a trade born at 2.04 ATR targets the 3 band, not the 2.
50%Tranche 2
Confirmed close through the trail, once the trail has turned Not a price — a condition. Requiring the trail to have turned stops a single poke from paying.
20%Runner
Opposite daily close through the mean The market closing on the far side of the mean, on the timeframe that generated the signal, ends the idea.
STOPAny day
Confirmed close beyond the stop closes whatever is left Live from v4.19. Before that the stop was printed but never acted on.
VOIDDay one only
Close back through the mean before T1 pays Premise failure, not price failure — the reclaim did not hold through its own first session. Fires at the mean, usually well before the stop.

Fade — two targets and a stop

50%T1
At the ±1 ATR bandWhere a stretched move typically first pauses — close enough to reach inside the session, far enough to pay.
50%T2
At the meanReversion means reaching the mean. Arriving there is the trade completing, not a trailing decision.
STOPThesis dead
Close back through the extreme of the stretchA new extreme says the move was not exhausted after all.
Neither setup is time-bounded

There is no session flatten and no maximum age anywhere in this indicator. A fade stalled between its stop and the mean is still open next week, and the labels and alerts keep firing for a position you may already have closed by hand.

7

Where every stop and target comes from

Reference
No level here is an arbitrary multiple.
LevelWhere it sitsWhy there
Reclaim stopThe pullback extremeThe trade says that low held and buyers reclaimed the mean. Price back below it says they did not.
Reclaim T1Next whole ATR band beyond entryOne ATR is a normal day's travel. Books a real gain on the leg you were paid for, without calling the top.
Reclaim T2Close through a turned trailAsks whether the trend carrying you is still carrying you.
RunnerOpposite daily close through the meanThe mean is the thesis. Closing beyond it ends the idea.
VoidThe mean, day one, pre-T1Premise failure — knowable before the stop is reached.
Fade stopExtreme of the stretchThe fade claims exhaustion. A new extreme refutes it. Mirror of the pullback extreme.
Fade T1 / T2±1 band, then the meanFirst pause, then the destination the whole trade is named after.
Both stopsCapped at 2 ATRStructural stops vary wildly in width. The cap bounds risk so sizing stays comparable across setups.

Every stop is the extreme the setup said would hold. Every target is a band the move is likely to reach, or the mean the idea is built on. Nothing is a round number and nothing is optimised.

8

Risk and position sizing

Yours, not the tool's
The indicator reports risk. It does not size anything.
The discipline that matters most here

The stop is a structural level, not a fixed ATR multiple. The pullback extreme can sit 0.6 ATR below entry on one setup and 1.8 on the next. Sizing a fixed number of shares across both is how this tool hurts people.

1 · Read the stop off the entry label Risk per share = entry − stop for a long. It is printed on the label and in the webhook payload.
2 · Decide the dollar loss you accept on this trade A fixed percent of account is the usual discipline. The indicator has no opinion.
3 · Shares = risk budget ÷ risk per share A wider pullback means a smaller position. Always.
4 · Check the R on the label If T1 pays less than about 1R, the setup is not worth the risk it is asking for.
5 · Check DTR / ATR in the table Past 1.00 the day has already spent its range. Size down or skip. The table turns red but nothing blocks the entry.

The ladder blends your R. Taking 30% at T1, 50% at T2 and 20% at the runner means realised R is a weighted average — not the number on any single label. A trade showing +2.7R at its final exit did not return 2.7R on the whole position.

9

Everything on the chart

Reference
Anything not listed here is not drawn by this script.
ElementWhat it gives you
Basis (gold)The adaptive mean. A true 5-day span intraday, a 20-period MA daily and above — the reference every number is measured from.
Band pairs A & BTwo selectable ATR multiples, 1.0 and 2.0 by default. Only two pairs are drawn — the tier engine still knows all four.
Session tintFaint green or red over a session with a live reclaim. Answers "am I in something?" after the entry label has scrolled away.
Dotted grey lineAn armed break-confirm level. A rejected setup waiting on structure.
Trail MAState-coloured, sized in minutes so it is the same real line on every intraday frame. Drives tranche 2 and gates fades.
MA30 ×3Lengths stated in 30-minute bars and rescaled to your chart, so the same three lines appear on a 5m as on a 30m.
Daily MAs5/10/20/50/200 day, computed as rolling chart bars — continuous, and never repainting on the open.
Session VWAPIntraday only, at or below 65 chart minutes. Context and alerts — never an entry condition.
Eight anchored VWAPs2-Day, WTD, MTD, YTD, YTD High, YTD Low, last and prior earnings. Read them to judge whether a signal is worth taking; they trigger nothing.
Right-edge tagsEach line's price in that line's colour, staggered so they never stack. The number printed is always the true value.
#7 ▲ LONGA confirmed reclaim entry, carrying its stop, its T1 band and the R that target pays. The number is the trade id.
F3 △ FADEA fade entry with stop, T1 and T2. F-numbers keep fades separate from reclaims.
◔ ◑ ● ✗Tranche 1, tranche 2, runner exit, and stop or void — each stamped with its trade's number and realised R.
Grey ✗ tagsSetups the filters rejected. Off by default. Turn them on to judge what a filter is costing you.
±N ATR tagsThe ruler, once per tier per session. Not signals.
Readout tableBasis, daily ATR, live extension, state, VWAP extension, DTR/ATR, day max, and live trade state with open R.

Every label has a tooltip. Hover any of them for the levels, the anchor date, and the reasoning behind that specific mark.

10

Settings that matter

Five of them
Most inputs are cosmetic. These change which trades exist. Move one at a time.
SettingDefaultWhat moving it does
Reclaim depth0.5 ATRHow deep the pullback must be. Raise for fewer, better setups. Never set it to 0 — the depth test then passes with no pullback at all.
Min days on side2Persistence before a reclaim can arm. Raise to kill flip-flops in chop; drop to 1 to catch faster turns and accept the noise.
Void tolerance0.10 ATRHow far a day-one close must go back through the mean to kill the trade. Raise to survive a messy open.
Break-confirmon, 3 sessionsThe second entry path. Off = clean reclaims only, materially fewer trades.
Max stop distance2.0 ATRCeiling on risk per trade. Tighter structural stops are left alone; wider ones are pulled in.

Leave these alone unless you know why. Confirmed ATR stops the bands repainting. The slope gate stops you buying into a falling mean. The regime gate stops you fading against your own position.

11

When this is wrong

Know these
Every tool has weather it does not work in.
A trending market kills the fade Price rides ±2 to ±4 ATR for days and every fade gets run over. The tell is the tags — +2, +3, +4 printing in sequence with no return. The regime gate only protects you when a reclaim is already open.
A tight range kills the reclaim Price crosses the mean every other day. The 2-day persistence filter catches the worst of it, but a two-day chop still qualifies and whipsaws you on a shallow bowl.
Earnings and gaps beat both There is no earnings blackout. A gap can arm, confirm and void inside one session — and a gap through your stop on day two or later is not protected by the void rule at all.
Thin liquidity makes the R a fiction Every level is a close. If the spread is wide, the R printed on the label is not the R you get. The tool assumes you trade the print.

The edge is in what it refuses. Shallow reclaims, one-day flips, reclaims into a mean moving against you, fades in the wrong regime — all filtered out. Most of this indicator is a list of trades it declines to take. Override the filters and you are trading something else.

12

Known gaps

Before money
Stated plainly so nobody discovers them with a position on.
GapWhat it means for you
No trend filterIt will signal a reclaim in a downtrend. The desk checklist would refuse it. Yours to enforce.
No coil / base testNo volatility-contraction requirement. A ragged base qualifies the same as a tight one.
Nothing is time-boundedNo session flatten, no maximum age. Calling either setup a day trade is intent, not behaviour.
Not fractalThe detector only ever asks the daily question. A 2-minute chart gives you daily reclaims filled with 2-minute precision.
No VWAP setupExtension is measured from the basis only. The session VWAP has alerts but triggers no trade.
Range gauge warns, does not blockPast 1.00 DTR/ATR the table turns red and the entry still fires.
No backtestEvery rule here is reasoned from the code. None of it is measured. The forward test is the test.
Read this before trading it

This is a measurement and signalling tool, not a strategy with verified expectancy. It marks conditions; it does not size, manage risk, or account for slippage, commissions or gaps. Every claim on this page describes how the code behaves — not how it performs.

13

How this differs from the desk checklist

Important
Same idea, different entry rules. They will disagree about which trades exist.

The desk Pullback & Reclaim checklist and this indicator both describe price pulling back to a mean and reclaiming it. They are not the same specification, so the same chart produces different trades. Neither replaces the other.

Desk checklistThis indicator
Trend filterRequired — must be in an uptrendNone
Pulls back toA named major MA you pick: 5/10/20/50/200 dayOne adaptive mean — 5DMA intraday, 20DMA daily+. Not selectable per trade
Base conditionRequired — must have gone quiet, a coilNone
MA clusterTightens the setup and the stopNot considered
PersistenceNot specifiedRequired — 2+ daily closes on the wrong side
DepthNot specifiedRequired — excursion ≥ 0.5 ATR past the mean
SlopeNot specifiedRequired — mean flat or turning your way
Stop~1.7% of priceThe pullback extreme, capped at 2 ATR

A stock coiled on the 50-day in a clean uptrend is textbook on the checklist — and this indicator may never fire, because the pullback never closed two days below its 5DMA. It will equally signal a reclaim in a downtrend with no coil, which the checklist forbids outright.

Use them together like this
The checklist decides whether the setup is worth taking. The indicator decides when the level actually broke, and manages the trade after.
Run the checklist for context and quality. Take the indicator's entry, stop and exits for execution and record-keeping. Where they disagree, the checklist is the judgement and the indicator is the measurement — the disagreement itself is information.