← Back to All Reports
DAILY REPORT · MONDAY 09/21 CLOSE · THE POST-EXPIRATION RIP PRINTED ON DAY ONE AND OUTRAN THE WEEKLY BAND · A CAP-WEIGHTED MELT-UP ON A FLAT EQUAL-WEIGHT INDEX · THE EQUITY OPTIONS TAPE TURNED NET BUYER WITH HOLDERS HARVESTING · THE FRONT OF THE VOLATILITY CURVE GOT BID · THE PAYROLLS PATH MAP IS VOID

Daily Report — 09/21/26 · “The Rip Nobody Priced”

The first session after the quarterly expiration did what the September map said it would do after the dealer book rolled off — and did it harder than any band on the sheet allowed. The S&P 500 closed outside its band for the day and its band for the whole week on the first day of that week; the Nasdaq-100 did the same on three catalyst names. Underneath, the equal-weight index barely moved, small caps sat at the bottom of their monthly range, banks did not participate and energy broke. This edition separates the mechanism from the fuel, reads the options tape after structure is stripped (it was a net buyer, with holders locking gains at the same time), shows why the dark-pool sell tags on the index wrappers are supply met rather than a signal, measures the shift in the volatility curve that re-opens the calendar trade, grades a report that called the mechanics wrong, and lays out Tuesday and the week with triggers rather than opinions.

The session in one line — the rip printed, and it was narrow

The S&P 500 (SPX) closed 7,764.70, up 1.55%, through both the daily upper band for the session and the weekly upper band for the whole week; the Nasdaq-100 (QQQ) closed 741.47, up 2.77%, through its daily and weekly uppers and its zone high. The move was the shape the September projection registered on 09/04 — a base into the inflation print and the rate decision, then a rip once the September expiration’s dealer book rolled off — and it arrived on the first possible session with a catalyst cluster stacked on top: META up 11.34% on a price-target raise and an app-download story, INTC up 12.14%, AMD up 9.95% to a lifetime high, ARM up 17.16%, QCOM up 9.29% on a customer validation, crude down about 4% on the Iran tape, the ten-year yield down five basis points.

The breadth tells the other half. The equal-weight S&P (RSP) rose 0.18%. Small caps (IWM) rose 0.52% and closed at their monthly lower band for a fourth session. The Dow (DIA) rose 0.76%. Financials (XLF) were flat at 55.90 and never approached the 56.73 repair line the prior edition set. This was a cap-weighted melt-up on the names that had a reason to move, not a broad repair — the same narrowness the last edition described, with the sign flipped from down to up.

TAPE SPX daily band for 09/21 was 7,607.22 / 7,693.78 and the weekly 7,551.84 / 7,749.16; the close cleared both · QQQ daily 715.85 / 727.05, weekly 709.26 / 733.64, zone high 735.26 · RSP +0.18%, IWM +0.52%, DIA +0.76% against QQQ +2.77%

The regime under it — the event card expires, the real yield still caps

Nothing at the top of the hierarchy changed: the Federal Reserve is tightening, the dollar sits over 100, and the real ten-year yield at 2.68% is still above the 2.34% breakeven. The rate decision’s event card, which read UNCLEAR and kept the index-direction machinery flat since Wednesday 09/16, expired with this session — Monday was its last window day — so from Tuesday the buy-the-dip signals that were mechanically suspended are back in force. That is a process fact, not a bullish input.

The day’s rate move helped the tape and did not change the regime. The ten-year fell to 4.96% and the long-bond fund (TLT) closed 81.80, back over the 81.35 line the prior edition made the condition for a durable rotation into rate-sensitive names — but that condition was written as a weekly close, and a Monday close is one of five. The curve flattened again (tens minus twos at 0.20%), which is the mechanism the banks have been losing to since the hike, and they lost to it again. With the real yield above breakeven the multiple cap on high-multiple names stays on; a day like this one is a bet on earnings growth outrunning it, and on Monday that bet was placed in three names.

Credit is calm (HYG 78.68, above its 78.35 line), the dollar block on new metals exposure is active (DXY 100.43), and the sentiment gauge for this session is not yet posted; the last reading, Friday’s 43.2, was neutral with both contrarian arms inactive, and it is cited here as Friday’s, not as Monday’s mood.

The four timeframes — through the weekly upper on day one, which changes what the week can do

On the daily frame the S&P closed 2.6 daily standard deviations above Friday’s close; on the weekly frame it closed above the 7,749.16 upper band on the first of five sessions; on the monthly frame it sits at plus 0.31 of a monthly move inside 7,428.73 / 7,943.55; on the quarterly frame it sits at plus 0.47 inside its quarterly band of 6,929.27 to 8,069.45 (the quarterly upper is 8,069.45, the quarterly lower 6,929.27). The two near timeframes are stretched and the two swing timeframes are not. The consequence is specific: a week that spends its whole expected move on Monday has, on the market’s own pricing, little left to give — the rest of the week is more likely to reconnect and hold than to extend, and an extension through the 7,800 call wall would need the breadth that did not show up today (small caps over 287.62, the equal-weight index participating). The monthly and quarterly bands say the regime is intact and there is room to the upside on a swing horizon; the daily and weekly say do not chase it here.

The Nasdaq-100 is the sharper case. QQQ at 741.47 is above its daily upper, above its weekly upper, above its zone high, and 1% under its monthly upper at 748.97; on the quarterly frame it closed over the quarter’s starting level for the first time this quarter. The swing ceiling is the monthly band; the near-term tell is the daily lower for Tuesday at 735.64, which is also the zone high — a close back under it is the first reversion signal, and holding above it is what keeps the monthly band in play. Small caps are the mirror: IWM at 285.58 is below the middle of all four of its bands and sitting on its monthly lower at 281.09; its reversion trigger is a close over 287.62.

Three names closed beyond their monthly two-standard-deviation bands and each carries a consequence rather than an opinion. META at 741.24 is 3.7 monthly moves above its August close and above its weekly two-sigma upper at 726.09, inside a zone of 626.61 / 689.35 / 752.09 with 1.46% of room to the high; the reversion setup triggers on a close back under 726.09, which opens the weekly one-sigma upper at 695.92. AMD at 615.52 is above its monthly two-sigma and its weekly upper at 591.03 and through its zone high (zone 481.51 / 547.53 / 613.55); that is a continuation read because the input that moved — hardware capital spending — moved twice in three sessions, and it is graded on a hold of 591.03. INTC at 121.78 is above its weekly upper at 116.26 and inside its monthly two-sigma, one percent under its zone high (zone 92.6 / 107.84 / 123.08); a continuation read carried from two weeks ago with the zone high at 123.08 as its target. Energy went the other way: XLE at 62.46 closed under its zone low of 62.78 on a range sheet whose trend column has turned negative, and that is a reversion setup with a trigger (a Tuesday close back over 62.78) and a cause (the physical driver reversed).

TAPE SPX daily band for Tuesday 7,726.47 / 7,802.93; the Asher implied-volatility bracket runs wider at about 1.26x the Silva band on the daily frame and 1.17x on the weekly · META monthly 2-sigma upper 663.82, weekly 2-sigma upper 726.09 · AMD monthly 2-sigma upper 577.38, weekly upper 591.03 · XLE zone 62.78 / 64.35 / 65.91, range column minus 31.8

Who led, who sat out — read from price

The leaders were the names with a catalyst, and every one of them carries a sponsorship tag: price up, call buyers present, holders writing calls against them and staying. META, INTC, AMD, ARM and QCOM did the index’s work; equipment ran with them (AMAT up 4.42%, LRCX up 4.92%, KLAC up 3.94%, the semiconductor fund SMH up 4.02%); SHOP rose 7.33% and rolled its long calls out to December on the way; the crypto sleeve ran a third straight session as bitcoin cleared 87,000 (MSTR up 9.47%, IBIT up 6.50%); software followed at a distance (IGV up 2.66%, CRWD up 4.92%, NOW up 1.63%) while services lagged (CRM down 0.63%). The memory pair split: MU held (1,043.96, up 2.77%) and SNDK slipped (1,766.64, down 1.41%) — the two-session reversal off the 09/16 low did not extend for SNDK, and MU reports next Wednesday.

TAPE META +11.34% to 741.24; after structure is stripped the options residue is a net buyer at +$117.1M with $212.8M of calls sold, 49% of them at or below spot (holders locking gains) · INTC +12.14% to 121.78; Nov-20 70 calls 15,000 lots sold at the bid at 2% extrinsic — a holder cashing a deep in-the-money line, never a short · ARM +17.16% to 322.90, residue +$16.3M readable

The mega-cap frame, name by name, because the index is these names. NVDA closed 227.38, up 2.30%, inside its weekly band of 214.48 / 230.06 with the zone high at 231.92 two percent away; its Oct-16 220 calls were the largest readable buy on the tape, though 37% of its raw premium carried no side tag, so the net is quoted rather than counted. AAPL closed 338.98, up 0.85%, inside weekly 327.33 / 344.93 — the all-time high sits at that weekly upper, as Silva noted — and its holders sold puts and wrote calls 71% at or below spot: an overwritten name. MSFT closed 501.61, up 1.59%, inside its zone of 486.13 / 497.12 / 508.11 and its weekly 480.64 / 506.92, coiled in the triangle Silva named (a hold over 505 to 508 confirms it, a close under 490 voids it); 72% of its sold calls were at or below spot, holders harvesting. GOOGL closed 354.97, up 1.55%, 0.7% under its zone high of 357.47 with the weekly upper at 361.31 — the flag Silva says opens the weekly upper; its options tape is structure-dominated and carries no readable direction. GOOG closed 350.87, up 1.88%; no Silva board carries a GOOG row, so the census implied-volatility bracket is its band — weekly 328.93 / 359.89, monthly 299.06 / 371.76 at 32.5% implied — and it sits inside both; its options tape is likewise structure-dominated. AMZN closed 258.45, up 1.87%, inside weekly 245.59 / 261.83 and one percent under its zone high of 261.07, with a flat options residue. TSLA closed 375.30, up 3.03%, inside weekly 348.27 / 380.27 and a point under its zone high of 378.99, residue balanced. AVGO closed 362.66, up 1.41%, inside weekly 344.03 / 371.19; more than half of its options premium is unsided and the rest is a structure-dominated artifact, so nothing directional is read from it. SPCX closed 151.85, down 0.56%, inside weekly 144.24 / 161.18 and monthly 126.50 / 160.88; the day’s print in it was a June 2027 risk reversal (120 puts sold against 230 calls bought), a bullish stock-like shape dated a long way out and quoted at low confidence.

TAPE NVDA $5.5B of dark volume on 32 prints (a magnitude marker, not a direction) · AAPL puts sold $37.4M after the strip, calls sold $40.1M, 71% at or below spot · SPCX Jun-17-2027 120P 15,800 at the bid at 10.50 with 230C 9,000 at the ask at 11.32, 11:01:49 · GOOGL $2.74B of dark volume on 23 prints; options residue structure-dominated

The names that sat out are the ones the last edition said had not repaired, and they still have not. The money-center banks were flat to down on a day the index rose 1.55% (JPM 352.04, GS 959.39, BAC 57.96), which is six sessions under the financials fund’s zone middle and a confirmed downtrend read on price, not a flow claim. Small caps are the weakest index on every timeframe. Defensives and metals were sold as risk came on (XLU down 1.07%, XLP down 1.06%, GLD down 0.70%, GDX down 1.10%). Energy was the clean loser: XLE down 2.88%, XOM down 3.20%, COP down 3.26%, the refiner MPC down 5.30%, the crude fund USO down 3.68%. NVO fell 7.96% on its 2030 obesity-sales reset to 39.80, under its weekly and monthly implied-volatility lowers — the second name the entry checklist marks ready, on the vol axis, because its long-dated implied volatility sits in the third percentile of its own history.

TAPE XLE 62.46 under its zone low 62.78; USO 148.16, minus 3.68% · XLF 55.90, +0.07%, sixth session under the zone middle · NVO 39.80, implied-volatility weekly lower 40.57, monthly lower 41.95; long-dated implied 39.2, third percentile of its history

The one bearish setup this flow can see — exhaustion, a vertical run with call buyers chasing in the top decile and holders selling calls into them in the top quartile, which fell 10% or more within ten sessions in 35% of cases against a 13% base — is tagged tonight on RMBS (up 10.68%), HUT (up 4.74%, and someone bought 4,650 Jan-2027 105 calls above the ask into it), RBRK, TEM and KEYS. It is not tagged on META, INTC or AMD, whose call buying is heavy but whose holders are still writing against them rather than being chased out. Abandonment — no call bid for weeks — is tagged on the industrials fund XLI, the S&P fund VOO and ISRG; that is a description, not a call, while price holds.

TAPE RMBS 96.96, +10.68%, exhaustion tag (call buying and call selling both in the top ranks on a 20-tape momentum in the top quartile) · HUT 103.42, Jan-15-2027 105C 4,650 lots above the ask at 21.49, implied 95

The equity options tape — a net buyer, with holders locking gains at the same time

After matched legs, stock-substitute legs and same-second packages are stripped, the single-stock options tape was a net buyer for the first time since the hike: about $554M of net premium toward calls bought and puts sold, on $1.97B of calls bought against $1.58B sold and $671M of puts bought against $832M sold. Where that premium was dated matters more than the sum. October monthly calls were bought (the largest single readable print on the tape is NVDA’s Oct-16 220 calls, 60,000 lots at the ask). December calls were bought as long call lines in BKNG and SHOP were rolled out from October. Puts dated to March and June 2027 were sold. And November deep in-the-money calls were sold in size — AMD’s Nov-20 430 calls and INTC’s Nov-20 70 calls, both at 2% to 4% extrinsic, both at the bid — which is what a holder does when a position doubles: cash the intrinsic and keep the rest. That is why the raw net-premium panel shows AMD and INTC negative in places where the stripped tape shows them as buyers: the raw headline is the delta-one sale, the residue underneath is fresh buying.

The monetization test says the same thing in one number: 91.3% of all sold-call premium on the file came from holders monetizing positions, against a premium-weighted trailing move of plus 4.7%, which is a HARVEST verdict. Combined with the dark-pool campaign test (which finds no single-stock campaign it can trust on a closing-print-dominated tape, and reads the index wrappers as supply met by cash), the two channels together say institutions were taking profits through the options market while staying long — and a harvest tape predicts capped-not-broken upside into the next expiration, crushed volatility once the chase ends, and call supply that dies at the October expiration. Nothing in it is a short.

TAPE NVDA Oct-16 220C 60,000 at the ask at 12.35, open interest 110,315, 15:37 ET, no same-second mate · AMD Nov-20 430C 5,500 at the bid at 188.75, 4% extrinsic · INTC Nov-20 70C 15,000 at the bid at 52.25, 2% extrinsic · file-wide monetization 91.3%, far-out-of-the-money opening call-writing 2.2%

The index book — puts dated to the quarter and the midterms, not to the week

Strip the boxes and the index tape is a put buyer for October monthly, November, December and January — and left the front alone. Over $11B of the S&P’s raw premium was institutional financing: 7000 / 8000 boxes in Dec-2031 (about $1.9B in one second), Sep-2027, Dec-2027 and Jul-2027, plus a 5000 / 10000 box in Mar-2027. None of it is direction and none of it is a wall; the 280,170 Dec-18 8000 calls and 303,600 Dec-18 7000 puts that top the large-open-interest panel are the same box family. What is left after the strip is about $517M of index puts bought against $204M sold, with the buying dated Oct-16 (a put-spread collar and a 7400 put), Dec-18 (7300 puts), Jan-2027 (7500 puts), and November through the VIX December 18 / 20 collar. The Oct-02 payrolls line, by contrast, saw puts net sold on this basis.

Two synthetic positions are worth naming because they are stock, not opinion, and they point opposite ways. At 13:30 someone bought 6,000 Dec-18 7750 calls and sold 6,000 Dec-18 7750 puts in the same second — a synthetic long S&P at 7,750 to December, a stock substitute. At 12:14 someone did the mirror at 7795: sold the calls, bought the puts, 2,520 lots, opening. Same-strike call-and-put pairs carry no readable direction as options, which is why the structure test sets them aside; read as stock they say one desk wants December delta at 7750 and another wants to be short it at 7795, forty-five points apart.

TAPE SPX Dec-2031 7000C 3,000 at the ask / 8000C 1,800 at the bid / 7000P 1,800 at the bid / 8000P 3,000 at the ask, 14:57:32, open interest 7,715 to 11,402 · Dec-18 7750C 6,000 at the ask at 253 with 7750P 6,000 below the bid at 177.70, 13:30:26 · Jan-15-2027 7500P 3,850 above the ask, Dec-18 7300P 3,850 at the ask, Oct-16 7400P above the ask, 14:39:27

The wrappers in the dark — supply met, not a signal

All three index wrappers printed at the bid in the dark pool on a day they rose: SPY $2.72B net at the bid on 45 prints with volume down 49% from Friday, QQQ $4.09B net at the bid on 48 prints with volume up 63%, IWM $250M at the bid on 25 prints. The vendor’s card calls QQQ the day’s largest dark-pool inflow; the number behind the card is volume, and the tag on it is sell. A single day of a wrapper being offered in the dark while the cash market closes at the high of the week is a description of supply that was absorbed — index rebalancing, closing-cross prints (SPY’s closing-auction share was 48%), short-the-pop positioning — not a direction. It is noted because last edition’s measurement showed that the dark-pool tag echoes the day’s price about 60% of the time and today the wrappers were in the other 40%.

The single-stock table is read the way the closing-print finding requires: as the closing auction with a side tag. NVDA’s $5.5B of dark volume on 32 prints, META’s $3.31B with 87% of its window at the bell, MU’s $3.12B on three prints, INTC’s $2.91B with 88% of the day at the close (the dark-pool test itself withheld a verdict on it) — these are magnitude markers, and they mark the names where the day’s action was. AMD is the one leader whose dark volume rose with price (up 67%). The one thick non-wrapper is WBD: 131 prints and 464 blocks, $2.54B of dark volume on an 857% jump, and an options tape of far-out puts against near calls — a deal-arbitrage book being built while the Warner case sits in court. It has no census file tonight and no price claim is made on it.

TAPE QQQ $9.76B of dark volume, +63% day over day, net minus $4.09B, thick tape · SPY $9.27B, minus 49%, net minus $2.72B, closing-auction share 47.87% · WBD Sep-25 27P 57,330 at 0.04 and Nov-20 22P 40,000 at 0.06 against Oct-16 29C 80,227 at 1.85 and Nov-20 28C 30,118 at 2.95

Flow map, flow timeline, dealer book — what the three say together

Together the three panels say the mechanical support that produced Monday is gone, the hedge that matters was rebuilt for October 2, and the equity book bought the October and December expirations. The dealer book’s only large row was the Monday expiration itself — about 15.8B of call delta that expired at the bell — and every row behind it is under 2.5B, with the largest put weight in November. So Tuesday opens with a small dealer book and no pin: the zero-day gamma map that closed with a positive node at 7,740 under a negative cluster from 7,750 to 7,780 (which is why the last fifteen minutes faded toward it) expired with the session and carries nothing forward except the lesson that Monday’s book was built below the close.

The flow timeline is the panel that moved. The October 2 index line fell to a new low near minus 70 from about minus 30 on Wednesday — the payrolls hedge was rebuilt into the rip, in size, on a basis that includes the packages the stripped tape sets aside. The September 25 line came off its low (near minus 57 from minus 75), which is protection expiring or being monetized into Friday’s crypto-expiry cluster. The Monday line spiked to plus 37 and is gone; Wednesday’s reads plus 14. The equity flow map puts the week’s call buying at Friday (plus 23 calls with plus 4 puts stacked) and the largest bar at the October monthly expiration (the panel labels it 10/23; the file puts the premium at Oct-16, where NVDA’s 220 calls sit), with the January 2027 row carrying plus 15 calls against plus 8 puts. Read forward: the next two dated risks are Friday (crypto expiry cluster plus the timing model’s turn window into Monday 09/28) and October 2, and the market is hedged for the second more than the first.

TAPE Dealers Diary 09/21 row about 15.8B call delta and 1.0B put delta, expired; 09/24 1.7B calls against 1.5B puts; 11/18 about 2.3B puts, the largest put row · Flow Timeline 10/02 line near minus 70 (new low), 09/25 near minus 57, 09/23 plus 14 · Market DEX plus 3.0B, the largest bar since 08/04

Volatility — the front got bid, and the calendar trade re-opens

Friday’s edition said the front of the curve was the low-priced contract almost everywhere: 4 of 95 names priced the front week at or above 1.10x the back month. Tonight it is 33 of 84, with six at or above 1.20x. The front got bid with the move. QQQ’s front three expirations rose 16% to 19% on the day with spot up (September 25 at 16.5 implied, October 2 at 17.4) while SPY’s front fell — the index volatility bid was a Nasdaq bid, the fast-squeeze signature of upside vol repricing rather than fear. The single names did the same: META’s front week jumped to 52.1 implied (up 31%) against 41.9 in October, a term ratio of 1.25x; INTC to 77.3; AMD to 62.2; MSTR to 90.0; BABA to 62.5 against 46.7 in October and 45.6 in December, the richest liquid front on the board for a second session at 1.37x.

That shift is the actionable one in this edition, because the structure discipline written last week — sell the front the tape inflated, own the back body, never a naked wing — needs a rich front to work, and last week there was none. Now there is, in the names where the checklist’s price point allows it: BABA, KWEB (front 37.4 against 28.9, 1.23x, on a confirmed reversal), MCD (27.4 against 22.2, 1.24x, a calm name where overwrites measured as the one cell that pays). Where the front is rich but the name just ran vertically — META, SHOP at 1.31x, IBIT at 1.16x — the short leg of a calendar is the wrong side of a fast tape and nothing is proposed. The index put wing is the other half of the surface: the October 2 five-percent put wing prices 1.80x its at-the-money and the September 30 wing 1.78x, the paying-up contracts, while the November body sits at 13.5 with a 1.28x wing — fair. So the hedge shape is the same as last week’s with new strikes: own the November body, sell the October 2 wing. Metals volatility was crushed as risk came off them (GLD October implied down 14% against its baseline, SLV down 18%); credit volatility is gone (HYG’s October 2 line at 7.0).

TAPE SPX at-the-money by expiry: 09/25 10.6 (+14% on the day), 10/02 10.5, 10/16 11.6, 11/20 13.3, 12/18 13.7; five-percent put wing 1.80x at 10/02, 1.42x at 10/16, 1.19x at 11/20 · QQQ 09/25 16.5 (+17%), 10/02 17.4 (+19%), 11/20 19.8 · term ratios: SHOP 1.31x, META 1.25x, QCOM 1.25x, MCD 1.24x, KWEB 1.23x, MSTR 1.18x, IBIT 1.16x, MSFT 1.13x

Sentiment — the gauge is a day behind the tape

The FOM sentiment gauge for Monday is not posted at the time of writing; Friday’s reading was 43.2, neutral, up 0.7 on the day and 2.7 on the week. A neutral reading before a plus 1.5% session is neither the capitulation threshold nor the euphoria arm; it is a crowd that had not committed, which is consistent with the equity book having to buy Monday rather than already owning it. The operator-facing question — am I early or late — is answered by the gated flow, not the gauge: holders were harvesting into the run at 91% of sold-call premium, which means the upside is capped by the people who own it, not sold by people who do not. That is a hold-and-do-not-chase tape. It is not a bearish one.

The commentary — who called it, who owned the miss

Twenty transcripts and one slide deck landed for this session; the useful ones agree on the mechanism and disagree on what comes next. Mike Silva’s deck and transcript put the conditions back at bull-strong by his own mechanical definition (the S&P over his 7,630 gamma-flip line with the Nasdaq Composite’s ten-day average inclining), name the move as the market having “priced it wrong,” put the S&P range at 7,500 to 7,800 with 7,800 the call wall, give 770 to 777 as Tuesday’s expected move on SPY (the ledger’s band is 769.66 / 777.34), and say most of the week’s move is probably in the books — he is trimming into the upper weekly band, which is the same profit-take the harvest verdict measures. His two setups, MSFT’s triangle (a break and hold over 505 to 508, target near 520) and MasterCard’s downtrend-line coil, are registered here as watches, not owned trades.

Volsignals owned its miss in the title (“I Messed Up”) and supplied the base rate: the Monday after a quarterly expiration has closed up seven of the last eight times with a median of plus 0.7%, with the caveat that the week after can turn hard. SpotGamma’s morning call — volatility releasing to the upside, Nasdaq upside vol higher on up days than down days, INTC’s implied rising into its rally — graded a hit on the day. Cem Karsan’s frame is the longer one: compression into the midterms, an unpinning after the quarter, a beta-one move lower in December or January, buy January-to-March volatility; the index tape’s own dated protection (puts bought for October monthly, November, December and January) agrees with him on where volatility is being bought and says nothing yet about his December call.

The perma-bears did what they do. The Maverick sold QQQ 740 / 745 and 735 / 740 call credit spreads into Friday’s expiration with QQQ closing 741.47, called the move a retail chase with no explanation, and set a five-item checklist (small caps leading, a crude reversal, dollar down, rates down, VIX turning from elevated) of which one item was met; the stripped tape contradicts the chase reading — institutions were net buyers with holders writing against them — and his QQQ 703 line, by his own concession, never broke. Tim Knight is 120% short with EFA his only large line, called it a rally with no cause, and his breadth shorts were right in relative terms (the equal-weight and Dow lagged) and wrong in absolute terms. The bitcoin channel — InvestAnswers, Krown, Tyler S, Benjamin Cowen (who conceded the May high) — has 87,000 printed and 90,000 in October as the open call; the framework’s own crypto-fork abort stays at 73,975. Rob’s Child measured the tens-minus-twos at 0.20%, and Arete’s semiconductor breadth trigger from Friday graded a hit.

TAPE Silva deck: GFLIP 7,630, call wall 7,800 (about 109M), put wall 7,500; ET Tradytics whale targets 775 / 770 / 760 — the day high 774.89 tagged the first within 11 cents · Volsignals: 7 of 8, median +0.7% · Maverick: QQQ 740/745 and 735/740 Sep-25 call credit spreads, ~2.40 and ~3.20 credit

Timing — the flat leg diverged; the next drawn turn is Friday into Monday

No new projection chart landed; the September update pair stands, and the orientation fork it opened on Friday stays open, so no direction is cited from the model. Both twins drew 09/17 to 09/21 flat; Monday realized a 2.6-standard-deviation session, so the structure diverged on that leg — the base-then-turn-then-spike shape can still match if the spike is what printed early, and it is graded again on the minor turn drawn for Wednesday 09/23 into Thursday 09/24. The major turn stays at the Friday 09/25 into Monday 09/28 boundary, with the month’s most violent single session drawn on the Monday. Timing charts give timing and shape, never price targets: if that turn is a high, the levels in play are the weekly two-sigma upper at 7,847.82 and the 7,800 call wall under the monthly upper at 7,943.55; if it is a low, the weekly upper at 7,749.16 (now support), the zone middle at 7,648.54 and the weekly lower at 7,551.84. The framework’s own September projection graded its rip leg a hit on the first session; its peak window (09/22 to 09/25) and the fade into quarter-end are the next grades. No bond timing chart is on the shelf.

Unusual trades — the structures worth naming

1. The five-year S&P box: financing, not a wall

At 14:57:32 four legs printed in one second in the December 2031 S&P chain: 7000 calls bought, 8000 calls sold, 7000 puts sold, 8000 puts bought, 3,000 and 1,800 lots a side, about $1.9B of premium. That is a box spread — an institution borrowing or lending against the index five years out at an implied rate — and the same family printed in September 2027, December 2027 and July 2027 during the day. It is the reason the S&P’s raw premium total is over $11B and its readable residue is a small fraction of it, and it is why the large-open-interest panel’s 8000 calls and 7000 puts for December are not a wall of anything.

TAPE SPX Dec-2031 7000C 3,000 + 500 at the ask at 2,649.42; 8000C 1,800 + 400 at the bid at 2,112.08; 7000P 1,800 at the bid at 523.53; 8000P 3,000 + 500 at the ask at 747.59; open interest 7,715 to 11,402 · Sep-17-2027 7000/8000 at 5,000 lots a leg, 10:22:13

2. Two synthetic S&P positions, forty-five points apart

Same-strike call-and-put pairs read as stock. At 13:30:26 the Dec-18 7750 calls were bought (6,000 at the ask) and the 7750 puts sold (6,000 below the bid) — a synthetic long at 7,750 to December. At 12:14:31 the mirror printed at 7795: 2,520 calls sold below the bid and 2,520 puts bought at the ask, both legs opening against almost no open interest — a synthetic short at 7,795. Neither is an option opinion; the structure test sets both aside as matched pairs. As stock they say one desk wants December exposure from here and another wants to be short it slightly higher, which is a fair description of a market that just spent its weekly move in a day.

TAPE SPX Dec-18 7750C 6,000 at 253 (ask) with 7750P 6,000 at 177.70 (below bid), 13:30:26, open interest 10,190 / 9,435 · Dec-18 7795C 2,520 at 216 (below bid) with 7795P 2,520 at 200 (ask), 12:14:31, open interest 212 / 12, opening

3. Holders cashing deep in-the-money calls into the run: AMD and INTC

The two largest single-name prints on the tape were sales, and neither was bearish. AMD’s Nov-20 430 calls, 5,500 lots at 188.75 with the stock at 611, printed at the bid at 4% extrinsic; INTC’s Nov-20 70 calls, 15,000 lots at 52.25 with the stock at 121, printed at the bid at 2% extrinsic. Those are stock substitutes being cashed by holders after doubles — the harvest side of the tape, and the reason the raw net-premium panel shows the November expiration as the day’s most negative row for equities while the stripped tape shows both names as net buyers underneath.

TAPE AMD Nov-20 430C 5,500 at the bid at 188.75, 11:28:22, open interest 11,694 · INTC Nov-20 70C 15,000 at the bid at 52.25, 15:25:15, open interest 73,191 · equity Nov-20 row: calls net sold about $179M, the two prints most of it

4. Two long-call lines rolled out to December: BKNG down, SHOP up

Both are same-second two-leg packages — structure before side — and neither leg reads alone (BKNG ZONE GAP: the zone sheet carries no BKNG row). At 15:03:19 BKNG’s Oct-16 162.4 calls were sold (31,000 lots at the bid, against 31,350 of open interest — closing) and Dec-18 150 calls bought (28,500 at the ask, opening): a roll out two months and down in strike, staying long and deeper in the money. At 10:05:36 SHOP did the opposite in strike: Oct-16 135 calls sold (45,000 at the bid) and Dec-18 145 calls bought (43,000 at the ask, opening) before the stock ran 7%. Both are holders extending their horizon into the December expiration, which is where the equity flow map’s December bar comes from.

TAPE BKNG Dec-18 150C 28,500 at 26.75 (ask), open interest 437; Oct-16 162.4C 31,000 at 10.40 (bid), open interest 31,350; 15:03:19 · SHOP Dec-18 145C 43,000 at 8.85 (ask), open interest 340; Oct-16 135C 45,000 at 3.70 (bid); 10:05:36

5. Gold: a holder re-strikes higher on a down day

A same-second matched pair, read as one structure rather than as a sale: at 10:45:44 GLD’s Jan-2027 450 calls were sold (52,339 lots at the bid, against 107,715 of open interest) and the Jan-2027 485 calls bought (52,339 at the ask) — the same size, one second, a roll up in strike that takes about 3.55 a contract out and keeps the long. It printed on a day gold fell 0.70% and the metals’ volatility was being crushed. Long gold persists at the large-holder level; with the dollar over 100 the framework’s block on new metals exposure stays on, so this is noted, not followed.

TAPE GLD Jan-15-2027 450C 52,339 at 6.62 (bid), open interest 107,715; Jan-15-2027 485C 52,339 at 3.07 (ask); 10:45:44 · the Dec-18 450C line on the chain panel 57,723 contracts at 4.75

6. The Warner deal book

WBD was the day’s thickest non-index name in the dark pool (131 prints, 464 blocks, $2.54B of volume on an 857% jump) and topped both the call- and put-volume-change panels, with a tape that only a deal explains: 57,330 Sep-25 27 puts at 0.04 and 40,000 Nov-20 22 puts at 0.06 (deal-break wings, 35% out of the money) against 80,227 Oct-16 29 calls at 1.85 and 30,118 Nov-20 28 calls at 2.95 (deal-close bodies), with the stock near 30.8. That is an arbitrage book being sized while the case sits in court. The name has no census file tonight, so no price claim is made; the structure is the finding.

TAPE WBD Sep-25 27P 57,330 at 0.04; Nov-20 22P 40,000 at 0.06; Oct-16 29C 80,227 at 1.85; Nov-20 28C 30,118 at 2.95; dark pool 131 prints / 464 blocks, net minus $120M at the bid

Scorecard — grading the prior report

Friday’s edition (“The Record Was the Closing Bell”) grades a C: right about breadth and right about discipline, wrong about the mechanics and wrong about META. Its central mechanism — that the two-day lift off the 09/16 low was expiration support that died at the bell — was contradicted on the first post-expiration session, when the market rose 1.55% with no expiration book behind it; its own invalidation (two sessions over 7,693.78 with the banks over 56.73 and small caps over 287.29) is one session of two met with the breadth half absent, so by its own rule it is not yet wrong, and in effect it already is. Its breadth line held: the repair conditions did not print, and the melt-up was as narrow as it said the market was. Its META reversion setup — a Monday close back under the monthly two-sigma upper at 663.82 — missed by 77 points; the reversion base rate it leaned on (89% revert next session) is a base rate, and Monday was the tail. The rotation condition on the long bond (a weekly close over 81.35) is half met on a Monday and graded Friday. The 09/25-into-09/28 window as the next dated turn is pending.

The Top Trades graded better than the thesis. The BABA September 25 / October 16 115 call calendar is alive with the stock at its strike (115.75) and the China fund KWEB above its 23.54 retire line; the front’s implied volatility rose from 50.2 to 62.5 against it, which marks against the position and pays it through Friday. The QQQ October 16 / October 2 put diagonal was armed with a rule — do not enter on a close over 725 — and QQQ closed 741.47: the rule kept a roughly 750-dollar loss off the book while the payrolls path map it belonged to was voided on its first session. That is the discipline working and the map failing; the lesson carried is that a dated call cap (the October 2 725 calls) is a level where supply sits, not a ceiling, and in a bull-strong transition it is run over. The SPY January 720 put tail met its retire test on Monday (S&P over 7,641 with the long bond over 81.35) and is retired at Friday’s close if both hold. The TLT October 16 81 call is in the money. The MSTR harvest-only line (open nothing new, sell no call against it) was right on a 9.47% day.

The September 11 Top Trades close their book: AMAT’s June 2027 400 calls (464.24, hit) and the MSTR 136 / 142 September call spread (expired at full value, hit) worked; the TLT January 2028 80 put died on the long bond reclaiming 81.35, the SQQQ October 39 call died on QQQ over 720, the IWM 290 / 300 call spread never triggered and its 286.53 kill printed; the HYG October 80 put and the XLY October downside are alive on their lines and not paying. Two hits, three dead index or reversal lines, two hedges idling.

Open theses and planned trades — where each stands after Monday

The September projection’s rip leg is a hit; the payrolls path map is void; the planned diagonals from Friday all watched their underlyings run before their red-day entry could fire. In order of what changed:

Tuesday and the week ahead — the path, with triggers

Base case, about half the weight: Tuesday reconnects and holds. A tag of the S&P daily lower at 7,726.47 that holds the weekly upper at 7,749.16 as support, with the zone middle at 7,648.54 as the floor of the week; QQQ tests its daily lower and zone high at 735.64 and holds it. This is what a market that spent its weekly move in a session does most often, it is what the harvest tape says (holders capping their own names into October), and it is what Silva’s reconnect read and the Maverick’s reversal-candle read both point at from different directions. The press-on case, about 30: through the 7,800 call wall toward the JPM collar and the monthly upper at 7,943.55 — and it requires the breadth that did not show Monday: IWM over 287.62, the equal-weight index participating, the banks off their zone lows. The give-back case, about 20: a close under 7,726.47 with the technology fund XLK back under its weekly upper at 192.86, which opens the daily two-sigma lower at 7,688.24. These are qualitative weights, not a model, and no index direction is registered over the open neutral stance.

The week has three dated events and one drawn turn. Thursday brings the Xi summit (Nadella and the technology leadership attending, which is the tape’s own explanation for part of Monday) and Costco’s report after the close, the consumer read Hans flagged. Friday is the crypto expiry cluster — the CME bitcoin and ether settlement and the Deribit quarterly expiry — on the same day the September 25 index put line comes off and the timing model draws its major turn into Monday 09/28. Quarter-end is Wednesday 09/30 with MU reporting after the close (consensus 31.33; its October 2 implied is 1.21x its November), and payrolls follow on Friday 10/02, where the index hedge book is now dated. The instruments to watch are the ones that did not participate: a small-cap close over 287.62 and a financials close over 56.73 would make the press-on case the base case; an energy close back over 62.78 would say Monday’s crude break was a one-day headline rather than the reversal of a physical driver.

TAPE SPX Tuesday daily 7,726.47 / 7,802.93 (2-sigma 7,688.24 / 7,841.16); weekly 7,551.84 / 7,749.16; monthly 7,428.73 / 7,943.55; quarterly 6,929.27 / 8,069.45 · QQQ Tuesday daily 735.64 / 747.30; monthly upper 748.97 · IWM 287.62 daily upper / 287.29 repair line · MU Oct-02 implied 72.5 vs Nov-20 60.1

Bottom line

The rip the map called printed on day one, harder than the bands allowed, on three names and a crude break, with the equal-weight index standing still. The options tape after structure is stripped was a net buyer and a net harvester at once: institutions bought October and December calls, sold 2027 puts, and cashed their November deep-in-the-money lines — long, extending their horizon, and capping their own names into the next expiration. The index book bought puts for the quarter and the midterms and left the week alone. The wrappers were offered in the dark and absorbed. The front of the volatility curve got bid with the move, which is the fast-squeeze signature and, in the names that did not just run vertically, the condition the sell-the-front calendar needs. The regime did not change: tightening, a real yield above breakeven, a dollar over 100, banks and small caps that have not repaired. So: no chase of the daily and weekly extension, structures that sell the inflated front and own a body sixty days or longer out, a re-struck hedge armed for October 2 and the midterms and entered only on a rejection, and the reversion setups (XLE at its zone low, NFLX and NVO at their bands) taken with triggers rather than conviction.

Top Trades to Follow

Institutional-flow structures the framework would follow, each priced off the surface it trades on, each with the window it is owned into and the line it is retired on. Every entry was run through the entry checklist (vol, level, exit): READY means three of three, WATCH means one point short and it is named. The checklist is an entry-quality gate, not a forecast — its own record as a winner-picker measured no edge — and every back leg here is sixty days or longer. Maximum losses are modelled from tonight’s implied volatilities and hold only if the live debit is under the per-structure cap.

CALENDAR BABA Sep-25 / Dec-18 115 call calendar (carried, re-dated to a December back) — WATCH: price sits 9.5% above its weekly implied-volatility lower; the richest liquid front on the board for a second session, front-week implied 62.5 against 45.6 in December, front-over-back 1.37x; own the December body at the 115 strike with the stock at 115.75, financed by the Friday contract the tape inflated; modelled debit about 7.35, maximum loss about 735 dollars; owned into the December 18 window with the front rolled each weekly expiry, retired on a close under 105.71 or on the China fund KWEB closing under 23.54.
SPREAD MSFT Dec-18 500 / 520 call debit spread — WATCH: price sits 3.2% above the zone low; the front week prices 27.4 against 24.3 in October (1.13x) and 28.8 in December (0.95x), so a September / December diagonal was modelled first and failed the per-structure cap at about 2,680 dollars a lot; the expression is the December body with the 520 wing sold at about 1.05x its at-the-money, a fair wing; modelled debit about 9.80, maximum loss about 980 dollars; entered on the triangle’s break and hold over 505 to 508, owned into the December 18 window, retired on a close under 490 to 493 and unconditionally on a close under the monthly floor at 478.14.
CALENDAR KWEB Sep-25 / Dec-18 25 call calendar — WATCH: price sits 3.9% above its monthly implied-volatility lower; front-week implied 37.4 against 28.9 in October, 1.23x, on a confirmed reversal with long-dated volatility in the fifth percentile of its history; own the December body at 25, sell the inflated front; modelled debit about 1.04 a lot, five lots about 520 dollars; owned into the December 18 window with the front rolled each expiry, retired on a close under 23.54, rolled up on a close over 27.
CALENDAR MCD Sep-25 / Dec-18 250 call calendar — WATCH: price sits 1.9% above its monthly implied-volatility lower (MCD ZONE GAP: no zone row on the sheet, the implied-volatility bracket is the band); a calm name (long-dated implied 24.7) with a rich front, 27.4 against 22.2 in October, 1.24x — the one cell where overwriting measured as paying; own the December body at 250 with the stock at 247.88, sell the front; modelled debit about 8.80, maximum loss about 880 dollars; owned into the December 18 window with the front rolled each expiry, retired on a close under 243.28 or rolled up on a close over 260.
SPREAD NFLX Dec-18 70 / 80 call debit spread — READY: long-dated implied in its middle third (39.6, so a spread and never a naked single), price at 73.36 under the monthly lower at 75.30, exit registered; December implied 37.9 against 33.8 in the front week (0.89x — the front is the low contract here, so no calendar; the body near the money is the buy); modelled debit about 4.70 a lot, two lots about 940 dollars; owned into the December 18 window, harvested on a reclaim of 74.33 then 75.30, retired on a Friday close under 69.25.
CALL NVO Jan-15-2027 40 call, single — READY: long-dated implied 39.2, the third percentile of its own history, which is the one case the vol axis allows a single call; price 39.80 under its weekly (40.57) and monthly (41.95) implied-volatility lowers after a 7.96% reset; the front week prices 1.02x the back, so nothing is sold against it; modelled about 3.50 a lot, three lots about 1,050 dollars; owned into a long-dated volatility pop or a reclaim of 45, never held into its final sixty days, retired on a monthly close under 34.15.
SPREAD XLE Dec-18 62 / 67 call debit spread — READY: long-dated implied in its middle third (25.4, a spread), price 62.46 at or under the zone low 62.78, exit line registered; December implied about 26.5 against 28.3 in the front week (1.07x, no calendar); the single-name band-break reversion cell measured 100% revert next session on nine moves in one regime — a base rate, and Monday’s META was its tail; modelled debit about 2.05 a lot, three lots about 615 dollars; entered only on a Tuesday close back over 62.78, owned into the December 18 window, harvested at the zone middle 64.35 then the weekly upper 66.57, retired on a weekly close under the monthly lower at 60.36; never re-entered the week it is stopped.
DIAGONAL SPY Nov-20 765 put / short Oct-02 750 put diagonal — the payrolls-and-midterms hedge, re-struck — WATCH: a hedge bought at the band high does not pass the checklist’s price point, and that is the point named; the October 2 five-percent put wing prices 1.80x its at-the-money (the paying-up contract, sold) and the November body 13.5 with a 1.28x wing (fair, owned), with both payrolls and the November 3 midterm inside the body’s window where the market’s own protection is dated; modelled debit about 9.70, maximum loss about 970 dollars; armed, not entered — entry on a rejection at the 7,800 call wall (a session over 7,802.93 that closes back under 7,749.16) or a close under the daily lower 7,726.47; owned into the October 2 window on the front, then rolled to the October 16 wing, and into the November 20 window on the body; retired on an S&P close over the weekly two-sigma upper at 7,847.82 or at the November expiry.

Sources

Expected moves & levels: Silva daily board, the ZONE DOCUMENT and the Zone Visual for 09/22 (zone document ingested this session, 164 symbols, 82 of 82 census closes reconciled), weekly board 09/21 to 09/25, monthly September and quarterly Q3 boards with the JPM collar levels; Asher implied-volatility bracket for 09/22 (built this session) and the census implied-volatility blocks; the FOM sentiment gauge for 09/18 (the 09/21 gauge not yet posted, declared). Tradytics: Live Options Flow and Darkpool Market Summary end-of-day files for 09/21; the 25-page dashboard capture for 09/21, every page read (the sector-flow line chart on page 17 read for slope only; the sentiment and intraday-target cards on page 1 are the operator’s standing exclusion; no QQQ zero-day panels in the capture). Recon pipeline: the 2026-09-21 census, 131 per-ticker reports read for this edition (SOXX, PCG, WBD, ASST and BNO have no census file and are declared). Ledgers: flow, dark-pool, monetization, setup-scan, expected-move, implied-volatility, long-dated-volatility and volatility-tracker ledgers for 09/21; the entry checklist and loss-budget state. Commentary (09/21): the FOM Silva transcript “They Priced It Wrong” and the 38-page 9/21 slide deck, both read directly; nineteen further polished transcripts (Mike Jones, ET Tradytics 09/21 and 09/22, the Maverick, Tim Knight morning and close, Krown, Cem Karsan, Rob’s Child, Arete Trading, InvestAnswers, tastylive, Options with Hans, Benjamin Cowen, Tyler S, Nanalyze, SpotGamma, Volsignals, Darius Dale) read in full by one reader pass with quotes carried verbatim; the Maverick’s 09/21 morning brief; the five intraday briefs of 09/21 and the analyst-feed digest. Timing: the September projection pair (091 update, no newer file). Market data: the Robinhood earnings calendar for 09/21 to 09/30; FRED series for the real ten-year, the breakeven, the ten-year and the Fed balance sheet (csv pull, the script’s own pull having timed out). Framework: comprehensive_analysis_0921.md, daily_close_recap_0921.md, the 0918 edition and its comp file, THESIS_LEDGER, TIMING_LEDGER, EM_ANALYSIS_LEDGER. Not integrated: no bond timing chart is on the shelf; no FOM morning brief for 09/21 exists; no web search was used.