The setup at 9:29
| Input | Value | Read |
|---|
| Overnight | O 7785 · H 7793 · L 7769 | 100% long inventory, "crowded" |
| Now | 7781 | Above Tue value (7746–7775), under the 7791–7803 ★★★ stack |
| VIX | 16.95 | Under 17 = green for buying dips |
| Card lean | — | Rotation day; buy 7766–7778 on absorption; fade 7785 on a lower high |
A crowded overnight long is a two-sided fact. It "tends to hold" in the backtest. When it doesn't, it liquidates fast. The open decides which.
Act 1 · 9:30–9:50
The overnight longs get tested
9:30 (right): −258 / −287 / −299 / −269, nothing over 300, and it took the whole bar to travel 15 points. Low 7765.50. 9:35 opens with +325 / +426 sitting right on top of the low.
Modest selling, no single big print, and the undercut of the 7766 stack was half a point. Then the answer: +426 against −299, bigger than the selling it was answering. Two of three on the bottom rule inside two bars.
9:35 closed green with +646 at 7770 (the trigger). 9:40 printed +491 at the top and fell 12 points on small reds to an equal low. 9:45 held 7765.50 a third time and ran to 7779.
Equal low ≠ lower low. A retest of the same price on shrinking reds is the bottom rule's second hold; the third hold is as strong as it gets.
The live mistake, and the correction that came out of it: the 9:40 bar looked like the trap pattern (big green at the top, price down 12). But the trap rule requires new lows, and 7765.50 was the same low as 9:30. Reds were −159 vs −299 — shrinking. The bar after that said so with +646 → +325 → nine greens.
Act 2 · 10:00–10:15
The break
10:00: −955 at 7770.50, price sitting on it, bar still open.
10:00 closed as −888 with price 5 under it. 10:05 stacked reds to 7755. 10:10: +472 at the top, then −376 / −517 below it and a new low. The green got sold through.
Three questions on the −955: it was mid-box (not expected traffic), the next bar fell 5 points (push-through), nothing answered. All three said "real." The fourth test of 7765.50, driven by a −888, broke — a resting buy-limit under the triple bottom would have filled into the move, which is why it was pulled.
The 10:10 bar is the trap shape in miniature: +472 printed, −517 answered below it, price made a lower low. Fourth green print of the morning to fail. The B-period single prints 7752–7760 on the day's profile are these fifteen minutes.
Act 3 · 10:20–11:20
Repair on air, then the breakout
10:25–10:30: seven greens in a row, none over +360, price up 7. No red answer at all.
10:35: −344 at 7764.50, price gave 4 and stopped. 10:40: nine green prints in a row. 10:45: +300 / +263 through the −888 shelf.
Price rising on small prints means there are no sellers, not that buyers are weak. Delta is effort; when nobody leans on the offer it takes very little to lift it. The one real seller (−344) was absorbed and answered. That is the entry on a drift: the first close back above the broken level, not a big print — there was never going to be one, because nobody was fighting.
10:51: POC 7766 is the −888 shelf. The market fought there, lost it, took it back, and now calls it fair. D singles below value = the dip was rejected hard.
10:50: lower high under the 7781.50 IB high, greens 299 → 161, first big red −369. Two of three on the top rule. 10:55 closed 7775.75 — a quarter point above the line. Not a trigger.
Then the greens grew: +371 at 11:00, +392 at 11:05. That's the top rule's "fails if." POC migrated 7766 → 7775 — value building at the top of the range, which is what a P-shape does before it extends.
11:15 closed above the IB high on +264 / +277, CVD 6,867 new high, VIX 16.63 falling. 11:20 dipped to 7780.50 — the first pullback to the breakout. Setup A: buy 7780.50, stop 7775.75, targets 7785 → 7791–94.
The trade you take is the one where price, CVD and VIX all agree. The ones you skip are the ones missing a referee.
Act 4 · 11:40–2:45
Two ledges and a squeeze
Into the stack: −753 at 7793, −712 / −408, −524. Greens 611 → 485 → 215 → 86. Highs flat at 7796 for thirty minutes. Both A targets hit; the runner came off at 7790.
12:10–12:35: a ledge. Nine red prints over 300 and four greens over 390 inside bars one to two points tall. Both aggressive sides absorbed; passive players in charge.
A ledge is what acceptance looks like: time and volume at a price. It's different from the 9:50 standoff (small prints, nobody active). Here everyone is active and nobody can move it. Who's losing? Count the effort: reds bigger and more frequent, and price didn't drop. The aggressive sellers are the trapped side; the passive buyer at 7789–90 has staying power.
12:40: 7792 → 7806 in one bar, biggest print +386. 12:45: back to 7797.50 on small reds.
Fourteen points on moderate prints = the −753 / −712 / −528 sellers covering. A short squeeze, not new buying — which is why it retreated instantly; nobody was defending the new highs because nobody had bought them with conviction. The test was whether the old ledge top, 7796, held as support. It did. Acceptance, and a second ledge formed at 7800–7811 for two hours. The F-period single prints 7782–7794 on the profile are the squeeze bar.
A breakout on air can't be chased, but it can be judged by the first pullback. Hold the level = buy the retest. Lose it = the same level becomes the short.
Act 5 · 2:20–3:05
The top rule, with twenty-five minutes' warning
Top to bottom: the last big buyers, the first big red, the fade, the close below it, the push-through, and the absorption at the morning's battleground.
| Bar | Biggest green | Biggest red | High | Sign |
|---|
| 2:20 | +694 / +491 | — | 7811.50 | The last big buyers. A high made on big effort can still fail, but the next four bars show the effort leaving |
| 2:25 | +147 | −644 | 7810 | First big red at the top — the warning |
| 2:30 | +174 | −329 | 7810 | Greens shrinking |
| 2:35 | +112 | −136 | 7810 | Lower high |
| 2:40 | — | −196 | 7809 | Reds only |
| 2:45 | — | −130 | 7809 | Close below 7809, the −644 price. Three of three — the short |
| 2:50 | +72 | −532 / −652 / −337 | — | Push-through, −10 |
| 2:55 | +325 | −541 | — | One green, sold through |
| 3:00 | +350 / +280 / +216 | −344 / −199 | — | 28-point bar. Absorption at 7779–81, low 7771.75 |
Short the 2:45 close at 7809, stop 7813.50, target the F singles at 7794 then the breakout shelf at 7780. Thirty points in fifteen minutes. It repaired the F singles on the way down — single prints get filled, same as always — and stopped at the day's POC.
Why it stopped at 7770
Because that's where the day's biggest decisions were made: +646 at 9:35, −888 at 10:00, the repair base at 10:40, the POC at 10:50. A liquidation runs until it finds the last place a real buyer showed himself, and asks whether he's still there. The answer was +350 / +280 / +216 at 7779–81 and +125 / +133 at the low. He was. Value held; the day closed Neutral, not failed.
3:06: two distributions (7752–80 and 7795–7811) joined by the repaired F singles. Value 7759–7799, POC 7777, poor high at 7811.50 (M/L/J, no tail) — unfinished business above.
The hinge: why 7766–7773 keeps mattering
| Day | What 7766–73 was |
|---|
| Thu Oct 15 | VPOC 7769 |
| Fri Oct 16 | D-period single prints — the fast move down, the original ★★★ supply |
| Mon Oct 19 | ON high 7768.50 rejected; 11:40 high 7772 with the −958 seller; Mon high 7773.25 |
| Tue Oct 20 | Broke on the third touch (+448), flipped to support; Tue VAH 7774.75 and Rally High 7768.75 landed in it; ON low 7769 |
| Wed Oct 21 | +646 absorber, −888 breaker, repair base, 3:00 liquidation stop |
Supply, then resistance, then support, then the battleground. Everyone active this week has a decision stored at that price, which is why it produces big prints every time and why moves stop there. It's the line between the week's two distributions: Friday/Monday value below (7695–7760), Tuesday/Wednesday value above (7775–7811).
Tuesday's profile explains the exact stopping price. 7769–7775 is a two-letter column (B, D): price raced through it both ways and nobody traded there in between. Directly under it, 7765–7768 is seven letters thick. Today's drop fell fast through the thin column and stopped at the top edge of the thick block.
Thin letters are highways; thick letters are brakes. The edge of the thick block is the level, not its middle.
What the day added to the rules
Act 1
Equal low is not a lower low
The trap rule needs new lows. A retest of the same low on shrinking reds is the bottom rule's second hold. Three holds at 7765.50 was a floor, and the live read that called it a trap was wrong.
Act 2
A resting limit under a triple bottom is a bet that the fourth test springs
When the fourth test arrives behind a −888 that worked, pull the limit. Re-place it at the next ★★★ and only honor it with a green answer.
Act 3
Drift up on small prints = no supply
Not weak buying. The entry is the first close back above the broken level; there will be no big print because nobody is fighting. Keep the stop close — easy movement gives no warning when it ends.
Act 3
Top rule "fails if": greens growing
Two of three with the greens getting bigger on each push is not a fade. Combined with POC migrating toward the level, the lean flips to the breakout.
Act 4
A ledge resolves toward the absorbing side
Count the effort on both sides. The side printing bigger and more often without moving price is the trapped side; their covering is the fuel for the break. A break on air is judged by whether the ledge top holds on the first pullback.
Act 5
The top rule gives warning
Warning (first big red) → lower high → greens shrinking → close below the warning print. Twenty-five minutes from −644 to the trigger; thirty points in fifteen minutes after it. The target is the single prints below, then the breakout shelf.
Act 5
Liquidations stop where the last absorber lived
Look left for the biggest green print at a level that held. That's where the flush pauses, and where the day's POC usually is.
Hinge
Two-letter columns are weak singles
On the prior-day profile, mark 2-letter runs of 4+ rows as thin ladders and the edge of the thick block beside them as the reaction price. ZoneFinder should flag these alongside true single prints.