DAILY REPORT · TUESDAY 11/17 CLOSE · THE S&P 500 -1.03%, QQQ -0.73%, RSP -1.45% · 150 UP, 578 DOWN · SPX UNDER ITS 200-DAY · 10-YEAR 5.109% · VIX 25.46 OVER ITS 3-MONTH · FOM SENTIMENT 14.5 · NVDA BEAT, +1.4% AFTER HOURS

Daily Report — 11/17/26 · “The Index Lost Its 200-Day as the Selling Rotated, NVDA’s Beat Faded Overnight — and the Gap Under the Market Is a Level, Not a Magnet”

The S&P 500 fell to 7,288.34 at 09:50 CT on an outage headline and settled 7,340.53, down 1.03%: 64.40 under Silva’s weekly lower 7,404.93, 4.05 under Tuesday’s daily lower 7,344.58, 4.00 under its 200-day 7,344.53 and under the 11/05 low close 7,353.73 - a weak reversion setup whose twin failed the session before, voided under Wednesday’s daily lower 7,248.42, with November’s monthly lower 7,239.64 on the same shelf. QQQ closed 656.95, a second close under its quarterly lower 669.25. 578 names on our census fell against 150, but the selling rotated: the AI basket closed flat while software, the financials, the insurers and the equal weight led down, and the 10-year slipped to 5.109% - not a rate day. NVDA beat and raised with its margin guide lower and held about +1.4% after hours (unofficial) against about 10.3% priced. The FOM sentiment gauge printed 14.5, its first capitulation reading - one print, not the cluster our rule needs. SPY’s April gap, 659.61 to 671.46, is a level and not a magnet: about 1 in 4 by the end of March on our base rates, with the middle path higher than Tuesday’s close. Our PPI card still suspends index-direction shorts through Wednesday: two branches, no lean.

Tuesday in one line — the index lost its 200-day while the selling rotated out of AI, and NVDA’s beat has not been paid for

A second broad down day, shallower than Monday, and a different one: 578 names on our census fell against 150 that rose, the median stock fell 1.26% (2.09% on Monday), and the selling left the AI names for software, the financials, the insurers and the equal weight. SPX opened 7,411.02, never traded over 7,415.37, fell to 7,288.34 at 09:50 CT on an outage headline and settled 7,340.53 (-1.03%): 4.05 under Tuesday’s daily lower 7,344.58, 64.40 under Silva’s weekly lower 7,404.93, 4.00 under its 200-day average 7,344.53 and under the 11/05 low close 7,353.73. That one close voids Monday’s reversion setup and fails the rebound that began on 11/05. Our close recap labels Tuesday’s close a reversion setup again, the default for an index under its weekly lower, and calls it weak — the same label failed the session before: triggered by a close back over 7,404.93, graded on Wednesday’s close over 7,340.53, voided by a close under Wednesday’s daily lower 7,248.42. QQQ closed 656.95 (-0.73%), a second close under its quarterly lower 669.25 — Monday’s continuation read stands, on the AI and software repricing, wrong on a close back over 669.25. SPY closed 732.44 (-1.01%), IWM 269.05 (-1.43%), the equal-weight RSP 215.76 (-1.45%). This page was built before Wednesday’s open for Wednesday 11/18 and the week; after-hours prints on it are unofficial.

The regime under it — not a rate day, again

**The 10-year settled 5.109%, two basis points lower, and front crude rose 0.3% to 86.01: the inputs that moved Tuesday were a news headline, company news and soft growth data, not rates.** At about 09:45 CT a headline reported an outage on the NSA’s network — the fifth outage at a large institution in five days; each earlier one was given a technical cause, this one has no established cause and there is no evidence of an attack. The session low printed five minutes later and held all day. Software fell on its own headlines (CRM, NOW), a broker cut its NAND price forecast (MU, SNDK), and a story that NVDA’s departed founder had asked to return ran into its report. Industrial production fell 0.4% against a +0.1% forecast and the NAHB builder index printed 25 against 29, the lowest of the 64 readings we hold; import and export prices ran hot (1.1% against 0.7% and 0.5%); retail sales met their forecast at 0.2%. None is a top-rank print: context, never a direction on its own. The Fed is tightening — it raised rates 25 basis points on 9/16 to 3.75% to 4.00%, held on 10/28, and a December increase is priced.

TAPE SPX 7,340.53, 64.40 under the weekly lower 7,404.93 after a 7,288.34 low; the 200-day 7,344.53; the 11/05 low close 7,353.73; QQQ 656.95, 12.30 under its quarterly lower 669.25; the 10-year 5.109%; real yield 2.92% against a 2.20% breakeven; crude 86.01; VIX 25.46 (session 22.36 to 27.73) over its three-month index 24.98; DXY 103.56; high-yield spread 3.34; FOM sentiment 14.5.

The four timeframes — three lines lost on one close, and four lines on one shelf under it

The swing frames lead: QQQ is under its quarter for a second close, SPX is under its week and its 200-day, and the next shelf under SPX is one where the day, the week, the month and the zone sheet all agree. Bands are Mike Silva’s expected-move boards; where he carries no row we name our census implied-volatility band.

The mega-caps against their bands

The bands show the rotation: the chips held their day into NVDA’s report, and the damage moved to software, Alphabet and the bitcoin names. Seven of the fifteen names on our expected-move roster closed inside every band; 104 census names closed under their weekly lower against 15 on Monday. Each name below carries its Wednesday zone or a declared zone gap.

TAPE SPX quarterly band 7,089.48 to 8,213.60, monthly band 7,239.64 to 7,920.50, weekly band 7,404.93 to 7,718.41, Wednesday’s daily band 7,248.42 to 7,432.64; QQQ 669.25 (quarter), 647.08 (month), 656.28 (week), 643.99 to 669.91 (Wednesday); the four-line shelf 7,239.64 to 7,248.42.

NVDA after its report — a beat and a raise, a lower margin guide, and a reaction that has not arrived

**NVDA reported after Tuesday’s close: revenue $115.84B against a street near $109.3B, and a guide of $130.0B (plus or minus 3%) against about $125B — with gross margin guided down to 71% to 72% from 74.9%.** The business first. Adjusted earnings were 2.68 a share against 2.47 to 2.48; data-center revenue was $107.6B, up 110% on the year; free cash flow $43.77B. The guided quarter carries an extra week, so by our arithmetic (not confirmed on the call) the per-week run rate grows about 4% against about 20% in the headline. The guide assumes no China data-center compute. Supply commitments and guarantees grew; the board said it does not intend to reinstate the departed chief executive; the quarterly filing comes in the coming weeks and is due by Friday 12/04, so it is not late.

The single names — 197 names owed a read; the damage moved from the chips to what had held

Across our roster the picture changed in one session: 162 names closed at 20-session lows (125 on Monday) against 4 at highs (40), and the new lows were CRM and NOW, JPM, UNH, MSFT and the funds SPY, IWM and RSP — not the chips. On our trend-model board 171 of 302 names read confirmed lower and 34 names left the confirmed-higher list in one session. The names below each carry their Wednesday zone or a declared zone gap.

Software and cyber — sold on their own headlines, and not bid on the outage

Crypto — MSTR under 100, and a dated seller on 11/30

The financials and UNH — Monday’s shelter, sold

What held — the most beaten, not the best

The exhaustion tags — one setup, described

Our setup scan tags nine names with exhaustion — a vertical run, call buyers chasing, holders selling calls into them: SQQQ, BRK/B, PAR, WRBY, RIOT, UAL, NXPI, GE and DAL. It is the one bearish setup we have validated from flow (35% of tagged names fell 10% or more within ten sessions against a 13% base). DAL 84.30 (-2.66%, DAL ZONE GAP) is on our list in the third session of its window; we add none, for reasons given under Top Trades. NVDA, AMD and AVGO carry no such tag, so nothing bearish is read from their flow.

The sector map — Monday’s order turned over

Ten of eleven sectors fell on market-cap weight, and the ones that held on Monday fell the most: Financials -1.91%, Utilities -1.64%, Health Care -1.34%, Communication Services -1.32%, Consumer Staples -1.17%. Technology fell 0.98% and the chips inside it were close to flat. The one sector that rose, Industrials (+0.52%), rose on a single name — SPCX added more than the whole gain. Across our 48 census groups Tuesday’s order ranks only +0.18 with the move since the 10/06 high, against +0.86 on Monday: the shelter-first order that had run since SPY’s record close stopped describing the tape.

The bottom test — the shape changed, the answer did not

Our bottom test ranks each day’s move against how far each name had fallen since 10/06, measured through the prior close. On Monday the most beaten fell the most (+0.71 across the census, +0.64 in the chips). On Tuesday the rank was -0.10 and -0.11: the first of the nine SPY down days since 10/29 on which the most beaten did not fall the most. 22 of the 64 worst names since October closed green (none on Monday), and the most beaten third of the chips rose. But quality broke — TSM -3.68%, MSFT -2.27%, UNH -5.42%, JPM -2.44%. The most beaten stabilising into a report while the large franchises are sold is the shape of a squeeze into an event. A bottom is the reverse: an up day led by the best names with breadth behind it. Tuesday was not one, and Wednesday is the first session that can be.

Group calls — three stand, two fired Tuesday, nine groups are one input short

A bearish group call is stated when three of a group’s four own inputs line up — price, bands, rising vol as it falls, lagging a flat SPY. None is a recommended expression here: no sector view is registered on any of the three.

Dark pools — volume only, no side

The vendor’s redesigned dark-pool feed carries no buy or sell side, so every figure here is size. The tape was $195.29B, flat on Monday like for like, 47.8% of it the closing auction: the second down day did not add volume. AMD drew the most on the file, $9.99B, about $7.43B of it on two prints between our 12:45 and 14:30 cuts — a size, never a side. SPY’s $8.91B was a third less than Monday’s $13.36B; NVDA took $6.00B on the session of its report; XLF’s volume nearly doubled.

The options tape — puts bought before the headline, put premium sold after the low

The option tape was two-sided, and its one clear feature is timing: single-stock puts were bought about two to one in the ten minutes before the outage headline, and put premium was sold after the low. Our deduplicated options file held $27.91B of gross premium, up 74% on Monday — but the whole increase was SPX spread tickets, financing and not direction. 57% of the file was structure — matched legs, box spreads, deep in-the-money stock substitutes — and the citable, sided layer was 38% of gross: readable, two-sided.

The dashboard — every vendor net leaned negative, and the vendor’s gross disagrees with ours

The vendor’s market net flow was negative all session, deepest at 14:30 CT and better into the bell. Our rebuild of it read -$69.8M at the close (the vendor’s own figure -$71.8M) after -$12.3M at 09:15, -$34.2M at 10:15, -$66.7M at 12:45 and -$89.7M at 14:30; it improved by $19.5M in the last half hour before NVDA’s report. These are vendor series, quoted and never counted as a signal.

The dealer map — the sign flipped at every cut and the close was split (vendor estimates, unproven)

The vendor’s estimate of SPX dealer gamma changed sign at every one of our cuts — +$0.99B, -$0.32B, +$0.09B, -$3.98B — and closed -$1.35B, while SPY (+$0.51B) and QQQ (+$0.21B) stayed positive all day. These are estimates of a book nobody can observe, under a forward test with nothing graded yet, and never merged with Silva’s levels: his 11/17 put lines 7,350 and 7,300 both traded through at the low, and his gamma flip 7,600 sits 259 points over the close. The split did not show up as a visible divergence in price — SPX, SPY and QQQ moved within a tenth of a percent of each other through the afternoon — so we read it as where hedging size sits, not as a regime.

Unusual trades — the structures worth naming

SPX at 10:04 CT — a November put sold against a December straddle bought

Fourteen minutes after the low, in one second, between 9,800 and 10,000 SPX November 20 7,500 puts were sold against 8,624 to 8,800 each of the December 18 7,320 calls and puts bought — about $472M of premium in one package (our two counts of the tape differ by one repeated clip, so we give the range). By structure it is a deep in-the-money put sold against an at-the-money straddle: long volatility moved out a month with a put kept at 7,320, not a hedge cashed. The December legs are new against lines of 23 and 421 contracts; whether the November sale closed a position is unconfirmed until the next open interest. SPX zone 7,245.64 / 7,495.00 / 7,744.36.

SPY — Monday’s put strip, bought back above the ask

At 10:02 CT the exact sizes of Monday’s SPY November 27 put strip — the 725 and 710 puts, 55,625 a leg — were bought above the ask, 2.24 and 1.37 higher than Monday’s sale, about $20.1M more than its proceeds; 83 seconds later 700 / 675 puts were bought for November 27 and sold for December 4, 44,500 a leg, as one put-spread package. By structure these are packages, not matched call-and-put legs: by shape Monday’s seller of downside bought it back at a loss near the low — unconfirmed until the next open interest. SPY zone 722.51 / 747.23 / 771.95.

SMH at the low — 20,000 long-dated puts sold

At 09:50:47 CT, the minute of the low, 20,000 SMH September 2027 380 puts were sold for $67.0M — the day’s largest single-name ticket, new against a line of 129 contracts, and sold at 45.9 implied vol against an at-the-money 51.8: one-sided, not a box spread at parity, the body of the curve and not the wing. A seller of downside 17% under the market for ten months is a holder’s view, never a floor. SMH closed 460.10, SMH ZONE GAP, on Silva’s Wednesday band 439.12 to 481.08.

NVDA into its report — two-sided to the last minute

NVDA’s two report expiries were matched within 10% on each side (November 18 calls $18.0M bought against $19.9M sold, puts $21.6M against $19.6M), with no late push; by structure no straddle or two-strike package of size passed our test on them. The lean was far-dated: January 2029 230 and 240 calls bought at the lowest vol on its curve, and seven minutes before the low a June 2027 240 put was sold against a 155 put bought, 3,250 a leg — a deep put moved down to the money, one package, unconfirmed until the next open interest. NVDA zone 139.62 / 165.21 / 190.79.

MSTR — long-dated puts bought into the index deletion

MSTR’s December 2027 95 puts were bought 7,012 against a line of 2,460 contracts, about $29.1M — one-sided, not a box spread at parity — with deep in-the-money November puts beside them: hedging into MSCI’s deletion after the 11/30 close, and never read as a bearish signal on its own. MSTR closed 96.50, MSTR ZONE GAP.

XLF — tail puts by contract count, not by premium

XLF’s January 50 puts were bought 131,405 above the ask late in the day, at least 54,000 of them new against the line, at 28.4 implied vol against an at-the-money 19.7 — 1.44 times, an elevated wing, one-sided and not a box spread at parity. It is a large count and a small premium: a tail hedge on the financials the day they led down, with KRE and BAC tail puts beside it. XLF zone 56.33 / 58.01 / 59.68.

Volatility and mood — three washout markers on one close, and what they are worth

On Monday the market paid for this week’s events and not for the quarter. On Tuesday the demand for protection reached past Friday for the first time in this decline, the VIX curve inverted, and the sentiment gauge crossed the capitulation line — all on the close that lost the 200-day. Each is a marker that tends to show up near washouts. None is a signal on its own, and price, our group calls and our earnings regime still read lower.

TAPE SPX at-the-money vol November 20 25.8, December 18 19.9, January 19.3; VIX 25.46 over its three-month index 24.98 (ratio 0.98); FOM sentiment 14.5 (27.4 Monday); the three-month implied correlation index 18.74; SPY front 1.18 times and QQQ front 1.23 times their own realized movement.

Earnings — misses sold four for four, one beat in three paid, and NVDA’s is still to be graded

Our earnings reaction regime stays bearish: on Tuesday every miss was sold and only one beat in three was paid. The label describes how the market is treating reports; it is never a forecast for the next one.

Tuesday’s day one

After Tuesday’s close — NVDA and SQM (unofficial marks; day one is Wednesday)

The week ahead

The commentary — the bears were right on the close and wrong about what was sold

Tuesday’s tape sided with the bear and sell-the-rally camp for a second session — 10 of 10 of its entries on our shelf confirmed or partly confirmed, against 0 of 3 bottom claims that named Tuesday — but not on the thing that camp had been selling. The selling left the AI names, the 09:50 CT low held all day, and the VIX made no higher high (27.73 against 28.36 on 11/05). Every desk below is a named outside view, weighed against our numbers and never a trigger; across our record desk calls on index direction have been right 50% of the time against a 50% base, so a camp count for one session is a description, not evidence of skill. No Mike Silva transcript or slide deck landed for 11/17, so his own verdict on the day is not on our shelf; his morning brief, his levels and his evening post are.

Silva’s levels and his follow-through day

Both put lines on his morning brief, 7,350 and 7,300, traded through at the low and the close sat between them; his gamma flip 7,600 is 259 points over the close. His evening post counted three distribution days in a row on SPY, which our bars confirm. His follow-through day on QQQ took a third distribution day and failed intraday on its fourth session — QQQ traded 646.64 against the 647.05 low the attempt rested on — while closing 9.90 over it. That is his rule read on our bars, not a call of ours.

The mechanics desks — they mapped the day better than either camp

The bear camp — right for the session, and not paid on semis

Tim Knight (twice), Ilya Spivak, Darius Dale, Todd Horowitz, Rob’s Child and Krown all read lower and the close agreed: 578 names down, SPX under its 200-day, QQQ under its quarter. What the count hides: the bears who were short semis were not paid on Tuesday — Knight says his own new shorts are elsewhere, and Spivak’s title, that NVDA is not the problem, is the day in four words. Knight’s ES level near 7,241 and Volsignals’ 7,250 are both our four-line shelf at 7,239.64 to 7,248.42. Knight’s claim that the odds of a higher market in three months are zero is a horizon to Wednesday 2027-02-17, and our depth record below is not zero-sided.

The bottom-callers — two claims we rebut

The Maverick, decomposed against the tape (a standing perma-bear: his framework is used, his mood is not)

The gaps under the market — levels, not magnets, and a path that takes months

**Is a move down to the open April gap in play? As a level, yes: SPY’s gap runs 659.61 to 671.46, with 666 inside it. As a pull, no: old gaps have shown no measurable attraction. What the gap stands for is depth, and depth has a base rate — about 1 in 4 that SPY reaches it by the end of March and about 3 in 10 by the end of April — on a path whose middle outcome is higher than Tuesday’s close and whose clock runs in months, not days.** Everything measured in this section is in-sample, a single run, not pre-registered: we quote it as a base rate and count none of it as a signal. A gap is a level — never an expected-move band, never a target.

Where the gaps are

Do old gaps pull price toward them? We tested it

We took every session on which SPY had an open gap at least 60 sessions old sitting 7% to 12% under the close while the index was 4% to 9% under its high, and asked whether price reached the gap more often than it fell the same distance on comparable days with no gap. On the first day of each episode SPY reached the gap within 60 sessions 21% of the time (19 episodes) against 15% for the control; QQQ 7% of 15 episodes against 17%. No measurable pull. A gap marks where a fall of that size would land; it does not make the fall more likely. Most gaps do fill — of 115 SPY up-gaps of half a percent or more since 1993, 70% filled within 60 sessions — which makes a gap seven months old a survivor, not a debt.

Depth and time — the base rates that answer the question

On Tuesday the S&P 500 closed 6% under its record for the first time in this decline (6.12%), under its 200-day, with short rates higher than six months earlier. Of the 58 completed pullbacks since 1950 that reached 6% under a record close:

GroupCountWent on to 10%Went on to 14.5% (the gap’s depth)Went on to 20%Median sessions to 14.5%.
All5845%28%19%67.
Under the 200-day on that day1850%39%22%103.
Short rates rising over the prior six months1867%44%39%98.
Both667%50%33%115.

What is lined up for the lower path, and what slows it

Two checks the next editions will grade

1. The early check: does a low inside about 20 sessions (to about Wednesday 12/16) hold the 706 to 722 area in SPY — about 7,075 to 7,240 in SPX? Holding it is the shape of the 42 pullbacks that did not go on, and our four-line shelf is the top of that area. A close through it is not that shape. 2. The relief-rally check: where does the next rally stop against 749.60 to 763.04? A failure under it is the shape of the 18 that went on; a reclaim of 749.60 to 750.36 on rising volume is the shakeout reading.

For NVDA the first level is the 147.96 to 148.80 gap over its weekly lower 146.38, graded on Wednesday’s and Thursday’s closes: no basis as a setup on a single name, a level. Nothing in this section is a registered view; it is an analysis with its clock — months — set beside it.

Timing — the window opened with a lower close, and the rebound leg is invalid

The timing charts are clocks, read for timing, direction and structure only — never for a price — and our own research found their dated turns land at about the rate random dates do. With that caveat stated first: the expiration window of 11/17 to 11/19 opened with a lower close, and the rebound leg that began on 11/05 is invalidated on a close.

Scorecard — grading Monday’s edition on Tuesday’s close (4 of 8 forward lines, 50%, a failing grade on our scale)

The 1116 page read the kind of day correctly and leaned the wrong way on the one index line it leaned on: its weak reversion setup on SPX, SPY and IWM was voided on the very next close. We grade three things and report them apart: the page’s forward lines (the grade rests on these), its option reads (graded on Tuesday’s open interest), and its list of trades to follow (in dollars).

Forward lines: 4 of 8 decided, 50% — under the 60% floor of our scale, a failing grade. That means the page’s regime read, its AI read and its map were sound and its one affirmative index call was wrong.

Why it missed. The bounce cell the page borrowed was measured on weekly-lower touches in a bull regime; the page said so in the same sentence and wrote the label anyway, under a tightening Fed, with an NVDA report two sessions out. It is the second edition running whose index reversion label broke the session after it was written. A label that weak should lead with its void line — which is how this edition writes it. The page also described the selling as Monday’s shape — the most beaten falling most, shelter holding — and Tuesday turned that over; and its vol read, that the demand was confined to Friday’s expiration, lasted one session.

Option reads, reported apart. On Tuesday’s open interest, of the page’s 41 gradable reads 16 were confirmed, 2 over-confirmed, 14 partial, 4 not confirmed and 5 failed; seven are withdrawn or corrected (listed under the options tape). On price alone, of the nine tickets that carried a side, the next session paid six and went against three (the SPY put-strip seller, the seller of QQQ’s December 660 puts, and the buyer of MU’s February call package) — each a matched two-sided tape read by package, never a signal.

The list, in dollars. The 1116 list made about $625 on Tuesday across its six marked entries — the SPY put +$485, the SMH put +$90, DAL +$58, PDD +$31, RUN -$9, WMT -$30 — and stands at about +$464 on twenty-four entries since 9/28, from about -$161. No held entry reached a retire line. The PDD calendar entered at Tuesday’s open.

Corrections to the 1116 page. Its bottom-test ranks (+0.78 and +0.70) included Monday’s own move; measured through the prior close they are +0.71 and +0.64 — the same conclusion, a smaller number. Its 200-day for SPX (7,342.71) was Monday’s; Tuesday’s is 7,344.53. It gave the volatility pop of 11/05 as 28.18, which is that day’s VIX close; the intraday high was 28.36.

Wednesday and the week — two branches and no session lean, under a card in its last session

Wednesday carries NVDA’s reaction, the Fed minutes at 13:00 CT and the VIX expiration, with our PPI card still suspending index-direction shorts through the close. We give two branches with their levels and no lean for the session. The trend read behind them is lower — three bearish inputs against one on both SPX and QQQ, with the Fed aligned — and it is not pressed, because the one input against it is a capitulation print.

Wednesday 11/18 — the lines

The rest of the week

Bottom line

**A second broad down day on which the selling rotated out of the AI names and into software, the financials, the insurers and the equal weight — and it cost the index three lines on one close: SPX settled 7,340.53, under its weekly lower 7,404.93, its 200-day 7,344.53 and the 11/05 low close 7,353.73.** That close is a weak reversion setup by our recap’s label (trigger back over 7,404.93, void under 7,248.42) whose twin failed the session before; QQQ’s second close under its quarterly lower 669.25 is a continuation read, wrong back over that line.

Top Trades to Follow

The institutional-flow list the next edition grades — defined risk, the body and not the wing, owned for a window and retired at a line. Every entry names where the name’s long-dated vol sits in its own history, a level with its number, and an exit or retire line with its number. Our entry check tests the quality of an entry, not the winner: its pre-registered test showed no edge as a picker. Marks are the broker’s closing quotes for Tuesday 11/17. The list made about $625 on Tuesday and stands at about +$464 on twenty-four entries since 9/28.

How Tuesday graded the 1116 list. The SPY December put rose to 24.64 (+$485) and the SMH March put to 67.03 (+$90); DAL marked 3.59 (+$58) in the third session of its window; RUN marked 0.66 (-$9); the WMT calendar marked 2.05 (-$30) with WMT’s report Thursday before the open. The PDD calendar entered at Tuesday’s open: PDD opened 78.14, inside 71.45 to 86.49; the debit on the first five-minute prints was 2.00 (it printed 2.24 five minutes later, and one reading of the opening bars puts it at 2.39, over the 2.30 cap, for the first ten minutes — we carry 2.00 and give the result on 2.24 as well); it marked 2.31 at the close, +$31 or +$7. No held line printed. The QQQ break entry’s condition printed a second time; the SMH diagonal’s test still fails.

What the sentiment print changes, and why no new calendar or exhaustion entry is added. Nothing is closed by a sentiment print — every carried entry sits on its own line — but no new short structure is added at a band-break low on a reading under 15: the armed QQQ entry gains a guard, and the one new entry is a watch that can only enter on a rally. On calendars: 49 of 55 names price their November 20 contract over December by a tenth or more, so the tape is front-heavy and the list already carries two entered calendars and a diagonal on watch. The fronts priced furthest over their own movement are this week’s reporters (TGT 114.6 implied vol against 40.6 realized, PANW 154.0 against 60.3, TJX 76.9 against 32.9), and a calendar entered Wednesday on TGT, TJX or LOW sells a report that has already printed; PANW and BABA read “don’t enter” on our trend board; NVDA is excluded the session after its report; nothing is sold on MSTR. Of the nine names our scan tags with exhaustion, DAL is already on the list and we add none: UAL is a second airline and its long-dated vol reading is stale, GE’s and RIOT’s have too few observations to rank, NXPI, PAR and WRBY have no long-dated vol history, BRK/B has no census file, and SQQQ is an inverse index fund.

PUT SMH Mar-19-2027 500 put (carried) — at 67.03, +$90 on Tuesday and +$159 on the structure since entry. Long-dated SMH vol sits in the high third of its own history (94th percentile, one-year 42.6): nothing is added at that level, and with SMH under its November lower 468.67 our entry check reads a new short as chasing; the put’s own implied vol is 45.3 against SMH’s realized 67.2. SMH closed 460.10, 39.90 in the money, SMH ZONE GAP, inside Silva’s week 454.56 to 518.94 and Wednesday’s band 439.12 to 481.08, 61.12 under its 200-day 521.22. Owned into March 2027 as the list’s sector hedge for the bearish AI read; retired on an SMH close over 521.22 — the line that invalidates the read it hedges — or at the Friday 2027-03-05 close. Maximum loss as held: about $6,703, its mark.
HEDGE SPY Dec-18 750 put (carried, the index hedge) — at 24.64, +$485 on Tuesday; the long put is +$1,312 from its 11.52 entry, and with the November leg’s -$1,910 the hedge is about -$598 since entry. Long-dated SPY vol sits in the high third of its own history (79th percentile, one-year 19.3), and the put’s own implied vol (18.3) is under SPY’s realized vol (22.7): nothing is added and nothing is sold against it. SPY closed 732.44, 17.56 in the money, on Wednesday’s zone 722.51 / 747.23 / 771.95, inside Silva’s Wednesday band 722.63 to 742.25 and under his weekly lower 739.59. Owned into the December expiry as the index hedge, a carried position and not a new index structure; retired on a SPY close over 772.27, or at the Friday 12/11 close. Maximum loss as held: about $2,464, its mark.
SPREAD RUN May-21-2027 9 / 12 call spread (carried) — at 0.66 (1.41 - 0.75; implied vols 82.4 and 82.0), -$9 on Tuesday and -$22 from its 0.88 entry. Long-dated RUN vol sits in the high third of its own history (81st percentile, one-year 80.7): a spread, never a single call. RUN closed 7.72 (-7.54%), 1.07 under the 8.79 line where adds may start, so nothing is added; RUN ZONE GAP; its implied-vol band for Wednesday is 7.41 to 8.03 and its week bottoms at 7.74, with no measured bounce rate for a single name there. Owned into May 2027; retired on a monthly close under 6.46, or at the Friday 2027-03-19 close. Maximum loss: the $88 debit.
PUT SPREAD DAL Jan-15-2027 85 / 75 put spread (carried — the exhaustion setup, session 3 of its window) — at 3.59 (the 85 put 5.60 at 41.0 implied vol, the 75 put 2.01 at 44.2), +$58 on Tuesday and +$50 from the listed 3.09 entry. Long-dated DAL vol sits in the middle third of its own history (33rd percentile, one-year 41.8): a spread, never a single put. Our setup scan still tags DAL with exhaustion — call buyers chasing (95th percentile) and holders selling calls into them (98th): 35% of tagged names fell 10% or more within ten sessions against a 13% base. DAL closed 84.30 (-2.66%), DAL ZONE GAP, inside its implied-vol bands 82.08 to 86.52 for Wednesday and 81.33 to 91.17 for the week; the 85 strike is now 0.70 in the money. Owned into the ten-session window; retired at the Friday 11/27 close (Thursday 11/26 is the Thanksgiving holiday) or on a DAL close over 92.77, whichever comes first. Maximum loss: the $309 debit.
CALENDAR WMT Jan-15-2027 / Nov-20 125 call calendar (carried — the short leg sells Thursday’s report) — entered at about 2.42 on Monday, marked 2.05 (the January 125 call 7.85 at 30.0 implied vol, the November 20 125 call 5.80 at 103.8), -$30 on Tuesday and -$37 since entry. Long-dated WMT vol sits in the high third of its own history (77th percentile, one-year 28.9): a calendar, never a single call. The short leg passes the test that binds: 103.8 implied vol against WMT’s own 20-session realized 24.9, 4.2 times. It expires inside WMT’s report week (Thursday 11/19 before the open), and that is the point: it sells the report’s premium and keeps January, its loss capped at the debit if WMT gaps far from 125. WMT closed 127.52, 2.52 over the strike, on Wednesday’s zone 122.40 / 127.60 / 132.79, inside its implied-vol bands 121.32 to 133.72 for Wednesday and 118.31 to 137.87 for the week. Owned through the report; retired whole at the Friday 11/20 close (the short leg’s expiry) or on a WMT close under 115.60 or over 135.44, whichever comes first. Maximum loss: the debit, about $242.
CALENDAR PDD Jan-15-2027 / Nov-20 80 call calendar (carried — entered at Tuesday’s open; the short leg sells Thursday’s report) — entered at 2.00 (the first five-minute prints, an approximation; 2.24 five minutes later), marked 2.31 (the January 80 call 4.10 at 40.4 implied vol, the November 20 80 call 1.79 at 109.5), +$31. Long-dated PDD vol sits in the low third of its own history (7th percentile, one-year 38.8), so the leg we keep is priced low on its own history. The short leg passes the test that binds: 109.5 implied vol against PDD’s own 20-session realized 30.1, 3.6 times. It expires inside PDD’s report week (Thursday 11/19 before the open): it sells the report’s premium and keeps January. Its exposure, stated: PDD is 11.8% under its 60-session high, in a cohort where 18.8% of 176 reports fell 8% or more against 11.9% that rose as much — a left-tail skew, and a call calendar struck 2.61 over the close loses toward its debit on a gap down; the loss is capped at the debit on either tail. PDD closed 77.39 (-0.92%), PDD ZONE GAP, inside its implied-vol bands 73.44 to 81.34 for Wednesday and 71.45 to 86.49 for the week. Owned through the report; retired whole at the Friday 11/20 close (the short leg’s expiry) or on a PDD close under 71.45 or over 86.49, whichever comes first. Maximum loss: the debit, about $200 (no more than $230 by the entry rule).
DIAGONAL SMH Mar-19-2027 500 / Dec-18 440 put diagonal (WATCH) — WATCH, not entered: a December 18 440 put sold against the carried March 500 put, for a credit of about 19.98 at Tuesday’s mark (the pair 47.05). The missing point is the short leg: it is sold only when its implied vol prints over SMH’s own 20-session realized vol, and it reads 58.6 against 67.2 — 0.87 times, still under (0.83 on Monday). Long-dated SMH vol sits in the high third of its own history (94th percentile, one-year 42.6), which is the reason to sell a leg against the body instead of adding to it. SMH closed 460.10, SMH ZONE GAP, 20.10 over the 440 strike, which sits on Wednesday’s daily lower 439.12; nothing is sold against a fund under its November lower on a sentiment print under 15. Tested on each session’s own quotes. Owned, if entered, into the December 18 expiry; the short leg retired on an SMH close under 460 or at the Friday 12/11 close. Maximum loss: the carried put’s mark less the credit — no more than about $6,703.
ARMED QQQ Jan-15-2027 670 / 640 put spread (armed — the break entry, Thursday 11/19’s open at the earliest, with a guard added) — at 12.07 (the 670 put 33.46 at 27.4 implied vol less the 640 put 21.39 at 29.9; $1,207 a spread): the body bought and the wing sold, both legs under QQQ’s realized 35.6. Long-dated QQQ vol sits in the high third of its own history (71st percentile, one-year 25.0). Its condition — a QQQ close under 669.25 — printed a second time (656.95); our PPI card holds index-direction shorts through Wednesday, so it fills no earlier than Thursday 11/19’s open, and only if QQQ’s Wednesday close is still under 669.25. The guard added tonight: it is not filled while QQQ sits under its weekly lower 656.28 — no new short structure at a band-break low on a sentiment print under 15 — so after a Wednesday close under 656.28 the fill waits for the first close back over it that is still under 669.25. Each session it stays armed it buys less of the 30-point width for more debit (about 1.5 to 1 now, 2.5 to 1 on Friday 11/13). QQQ zone 628.28 / 669.08 / 709.88, Wednesday’s band 643.99 to 669.91. It is the structure for our registered bearish view on QQQ. Owned, if entered, into January; retired on a QQQ close over 723.16, or at the Friday 2027-01-08 close. Maximum loss: the debit at entry, about $1,207 at Tuesday’s mark.
WATCH QQQ Mar-19-2027 670 / 600 put spread (WATCH, new — the lower-path entry, only on a relief rally) — WATCH, not entered: about 23.41 at Tuesday’s closing quotes (the March 670 put 42.42 at 26.6 implied vol less the March 600 put 19.01 at 30.9; $2,341 a spread against a maximum value of 70). Long-dated QQQ vol sits in the high third of its own history (71st percentile, one-year 25.0): a spread, never a single put — the body bought near the money and the strike in QQQ’s April gap (588.98 to 602.12) sold. The sold put prints 30.9 implied vol against the body’s 26.6, 1.16 times — under our 1.3 line, so it is sold to cap the debit at the level the read names, not because it is dear. Why a watch: of the 18 pullbacks that reached the gap’s depth, 15 rallied above their starting close first, on a clock of months — so the entry is the rally, never Tuesday’s low. It is written on QQQ because our registered index view is on QQQ. QQQ closed 656.95, zone 628.28 / 669.08 / 709.88; the entry zone is 669.25 to 695.57 (its quarterly lower up to the 11/11 rally high), and the trigger is a failure inside it: after a close back over 669.25, the first lower close while still inside the zone. Never entered while the PPI card holds (through Wednesday), and only while an index view in its direction is registered (the current one is graded Friday 11/20, so a new one is owed for any later entry). Owned, if entered, into March 2027; retired on a QQQ close over 695.57 — the reclaim that reads as a shakeout — or at the Friday 2027-03-05 close; unentered, the watch lapses on that same close over 695.57. Maximum loss: the debit at entry — no more than about $2,341 at Tuesday’s quotes, less on a rally entry.

Retired from the list: nothing tonight. Entered: the PDD calendar, at Tuesday’s open.

Not listed: anything on NVDA (a name under its November band on the session after its report: nothing is sold against it and no direction is taken into day one); anything sold into Tuesday’s post-report gaps (JD, RCAT, ESLT, SQM); MSTR (no short call by our standing rule; MSCI’s deletion is a dated event, not an entry, and its low is not callable on anything we measure); a new long in batteries, copper and base metals or drones, our three standing group calls, or in the nine groups one input short of a call — and no put-side structure on a called group either, because no sector view is registered on any of them; new index structure before Thursday’s open (the PPI card) and any new short structure at a band-break low (the sentiment print); gold and the miners (our dollar gate on commodities is active); anything bearish on AMD or AVGO (no exhaustion tag on either); the reporters’ calendars and the other exhaustion names (above); the pre-report tails on WWD, TJX, INTU and ROST, which we describe and do not trade; and any structure struck for SPY’s 666 gap itself — a gap is a level with no pull, and the QQQ watch above is the only expression of the lower path.

Sources

Expected moves. Mike Silva’s (Figuring Out Money) expected-move boards: Tuesday’s daily board for 11/17 (graded; indices and futures only) and the daily board for Wednesday 11/18 (forward; indices, futures and sixteen single names and funds); the zone sheets for Tuesday 11/17 (graded) and Wednesday 11/18 (forward) — the ZONE DOCUMENT, the Zone Visual, the zones and the range-and-trend sheets; the weekly board for 11/16 to 11/20; the November monthly board and the fourth-quarter board with the year-end collar levels; Silva’s 11/17 morning brief and gamma levels; our implied-volatility bands computed from Tuesday’s end-of-day options file and carried in our census; the FOM sentiment index for 11/17 and our sentiment tracker.

Tradytics files. The end-of-day options file for Tuesday 11/17 (72,161 prints) and Monday’s for comparison and the open-interest checks; the end-of-day dark-pool summary for Tuesday, without a buy or sell side on the vendor’s redesigned feed, with its fill-location columns; the options dashboard for 11/17, read from the raw data the redesigned page is drawn from at the close and at four intraday cuts (market net flow, flow extremes, the money map and sectors, the contract tables, the same-day flow and gamma for SPY, SPX and QQQ, the dealer history and the vendor’s chart levels), with the equities page (assets, sectors, themes, the digest, the heatmap, movers, earnings, insiders and congress filings) and the flow board; the vendor’s dealer map at the close, re-read after midnight, and at four cuts; our rebuilt net-flow record; an earlier dashboard capture read for layout only.

Census and our records. Our per-ticker census for 11/17 (734 names) and the prior session, its sector chunks and summary; the name-behavior record (our price history and the week’s movers), the census screen, the AI and speculative basket monitors, the group-convergence record, the trend-model board (current at the 11/17 close), the long-dated vol record, the setup scan, the monetization record, the vol tracker and the implied-correlation record, the cross-asset reading, the event regime with the PPI card, the stance licence and its sector views, the thesis record and the thesis watch, the rebound plan, the audit of our editions of the week of 11/02, and the path-and-edge research program.

Intraday and the close. Our five intraday briefs for Tuesday (pre-open, 9:15, 10:15, 12:45 and 14:30 CT), the cut level frames and the close frame, the cut flow ledgers, sector sweeps and ticker recon, the news ledgers, the scheduled-prints record, the pre-print drawdown screen, the earnings print cards and our print notes on NVDA, SQM, DIS, ESLT, HD and JD, the intraday chart read, the Maverick’s morning brief and our 11/17 close recap.

The gap study. Our research record on the open gaps (the magnet test, the depth and time base rates since 1950, the analog paths and the resampling, the range map), built on daily bars for SPY, QQQ, NVDA and the S&P 500; in-sample, single runs, not pre-registered. The gap edges were verified on daily bars.

Timing. Savino’s November projection (both readings), his November-to-December analog, his Treasury-bond forecast (the 9/30 update), the Maverick’s economic-stages chart, Silva’s 2018 ten-year analog tracker, our rebound plan’s clock, and our timing record. No price on this page is taken from a timing chart.

Commentary. Every polished file on the Market Commentary shelf dated 11/17 (24 files): the Maverick’s 11/17 video; Tim Knight (two); Darius Dale (42 Macro, two); Arete Trading (two); Volsignals (two); Ilya Spivak; Geeks of Finance; FX Evolution; Danielle Shay; Dividend Data; ET Tradytics; Krown; James of InvestAnswers and his posts; On-Chain Mind; Rick Rule; Rob’s Child; Todd Horowitz; Tyler S; Andrew Rader; Mike Jones; SpotGamma’s Brent Kochuba (his 11/16 file, graded); Mike Silva’s 11/17 brief and evening post (no transcript or slide deck landed for 11/17); the external-setup record.

Rates, the dollar and market data. FRED’s published rows for the 10-year real yield, the breakeven, the Fed’s balance sheet and the high-yield spread, fetched directly; our rates block; our dollar-gate reading; TradingView bars for the 10-year, 2-year and 30-year, the dollar index, crude, KOSPI and the after-hours session; the broker’s market-data quotes for the list’s option legs and for QQQ’s daily bars.

Prior edition. Our 1116 daily report and its comprehensive analysis, the 11/16 close recap, and our correction record.