SomerQuant Daily · Friday, 18 September 2026 · EOD options flow & dark-pool research · SomerQuant
18 September 2026 — EOD read. The quarterly expiration did what quarterly expirations do, and then some: our $100M-and-above dark-pool board came to $117.73B across 379 prints in 128 names, more than the previous four sessions combined. Almost all of it cleared in the closing cross, and the single largest print on the tape was $9.36B of SPCX at 152.71, the exact closing price. The index tape itself barely moved. SPY finished at 761.69 against 762.60, down 0.12%, on a day it went ex-dividend $1.888834, which means the fund was fractionally higher on a total-return basis while the quote showed red.
Market. SPY 761.69 on a 4.03-point range; ex-dividend day masks a flat-to-higher session.
Flow. Net options premium sat below zero for 52 of 79 readings, then finished at its high.
Next session. 757.97 is the level to watch; the rebalance prints leave no directional read.
SPY opened at 761.31 and spent the first fifteen minutes inside a 760.59 to 761.76 range, then gave that up and worked lower through the middle of the day. The session low of 757.971 printed between 12:25 and 12:30 PM ET, and the high of 762.00 did not arrive until the final five minutes. The close was 761.69. Ten regular-session dark-pool prints cleared the $100M screen in SPY, grouped into three marks on the figure, and every one of the ten carried a contingent or derivative-priced condition. All ten printed at 762.693, a price the tape never traded, which is the whole reason those marks are drawn hollow. The largest was $3.03B at 11:23 AM ET. That one level carried all ten prints. Of the five heaviest premium shelves, only the 758 strike sat inside the day’s range at −$9.8M net; the 745 shelf at −$19.7M, the 755 shelf at −$13.4M, the 720 shelf at +$12.0M and the 785 shelf at −$10.4M all sat off the price scale.
The tide opened at +$43.5M and held positive for the first hour. It crossed below zero at 10:30 AM, fell away through the late morning and reached its trough of −$67.3M at 12:35 PM ET, within ten minutes of the session low in SPY. From there it ground back, turning positive again in the last half hour and finishing at +$51.3M, which was also the highest reading of the day. Twenty-seven of the 79 five-minute readings sat above zero and 52 below, and the series crossed zero six times. A day that finishes at its own high after spending two-thirds of itself underwater is not a day with a clean directional read.
17 September’s radar — OI confirm
| Contract | Prior vol | OI change | Ratio | Verdict |
|---|---|---|---|---|
| SPX 19 Mar 2027 $5,900 puts | 1,352 | +1,276 | 0.94 | ✓ Confirmed |
| VIX 21 Oct 2026 $24 calls | 50,859 | +45,209 | 0.89 | ✓ Confirmed |
| CRWV 15 Dec 2028 $150 calls | 2,221 | +1,937 | 0.87 | ✓ Confirmed |
| MU 2 Oct 2026 $1,000 calls | 12,347 | +8,335 | 0.68 | ✓ Confirmed |
| SPX 18 Dec 2026 $7,710 calls | 4,232 | +2,265 | 0.54 | ✓ Confirmed |
| SPX 18 Dec 2026 $7,710 puts | 4,179 | +2,132 | 0.51 | ✓ Confirmed |
| SPX 19 Mar 2027 $8,000 calls | 1,323 | +31 | 0.02 | ✕ Not confirmed |
| IWM 18 Sep 2026 $315 puts | 48,145 | +2 | 0.00 | ✕ Not confirmed |
| QQQ 18 Sep 2026 $745 puts | 2,031 | −27 | — | ✕ Not confirmed |
Verdict describes the open-interest update only, never the contract. Six of the nine cleared the 0.40 line and none landed in the partial band, so the tally is cleanly split. The two that failed hardest were both contracts expiring today: the IWM 315 puts turned 48,145 contracts of volume into two units of new open interest, and the QQQ 745 puts saw open interest fall by 27. That is what closing activity into an expiration looks like, and it is the correct outcome rather than a miss in the screen. The SPX March 8,000 calls are the genuine non-confirmation: 1,323 contracts traded and 31 stayed.
What happened
This was the September quarterly expiration, and the tape behaved accordingly. Volume in SPY sat near its ordinary level for most of the day and then arrived in two bursts: 9.15 million shares in the 11:15 AM bar and 9.09 million in the 12:30 PM bar, both of them tied to the derivative-priced blocks rather than to any move in price. The range for the whole session was 4.03 points, from 757.971 to 762.00, and 2.09 of those points came in the final half hour as the close was pinned.
Underneath, the composition of the day was unusual in a way the headline index numbers cannot show. The $100M-and-above dark-pool universe printed $117.73B across 379 prints in 128 names. The four sessions before it came to $114.74B between them. The ETF share of that notional collapsed to 25.3% from 65.5% yesterday, which is the signature of an index rebalance: the money moves through single names, not through the wrapper. SPCX alone accounted for $18.81B across 17 prints, all of them within a tenth of a percent of 152.71, and Bloom Energy put up $5.34B across 13 prints at 265.63.
The options side was quieter and more two-sided. Market-wide net premium finished at +$51.3M after a trough of −$67.3M, and the flagged-events screen was dominated by SPX structures rather than single names. The largest lines were paired: 1,000 March 2027 7,500 puts at $21.7M against 1,000 March 2027 8,000 calls at $21.1M, and a matched 740-lot in December 7,720 calls and puts at $15.4M and $15.3M, each against open interest under 750. Paired prints of that shape are typically a spread or a roll being written in one go, and they say less about direction than their size suggests.
Why it matters
A $117.73B board is not a sentiment reading. It is a plumbing event. When the quarterly rebalance clears, index funds have to hold the new weights by the close, and the cleanest way to do that is a single cross at the closing price. That is why 17 SPCX prints all landed at 152.71, why Bloom Energy’s 13 landed at 265.63, and why the biggest print of the session sits at exactly 0.00% from its own close. None of it is a view. Reading it as accumulation would be a mistake, and we are flagging that explicitly because the notional is large enough to tempt exactly that reading.
The part that does carry information is what the rebalance crowded out. On an ordinary session the ETF share of this board runs high because index products are how large books express a market view. Today it fell by more than forty points. Set against an options tape that spent 52 of 79 readings below zero and still closed at its high, the honest summary is that very little directional money changed hands today. The positioning gauge says the same thing by saying nothing: no leveraged-long, inverse or volatility ETF cleared either screen, so all three read $0.
The ex-dividend detail is worth holding onto into next week. SPY’s $1.888834 distribution is 0.25% of the price, so the fund’s 0.12% decline is a total-return gain of roughly 0.13%. Anyone comparing SPY’s chart against QQQ at +0.63% over the next few sessions will be reading a gap that is partly an artefact of today’s payment date.
What to watch into Monday’s session
- 757.97 was today’s low and the level to watch. A close below it would mark the first break of the range that has held since Tuesday.
- 762.00 was the session high and it printed in the last five minutes. Holding above it on Monday would say the pin was a floor rather than a ceiling.
- The 745 strike carries the heaviest net put premium in SPY at −$19.7M across all expiries, some 2.2% below the close. It is the nearest shelf of consequence on the downside.
- 720 is the only call-led strike among the ten heaviest at +$12.0M, and it sits 5.5% below spot. Call premium stacked that far under the money usually reflects existing positions rather than new expression.
- Micron reports on 30 September and the 25 September 1,010 puts traded 41 times their open interest today. That contract expires before the print, so whatever it is, it is not an earnings position.
- Monday’s dark-pool board will show whether 128 names was a rebalance artefact or the start of something wider. Anything above 60 names would be notable; the trailing four sessions ran 44 to 85.
Names on our radar
Every row below is frozen at publication and scored mechanically in the Accountability Ledger at +5 and +21 sessions — misses stay on the page. Options prints are checked against the next morning’s open-interest update in the following session’s post.
| Ticker | What printed | Read |
|---|---|---|
| SPX | 1,000 March 2027 7,500 puts, $21.7M, against open interest of 12,739 | The largest single line on the flagged-events screen, and it is paired with a same-size call line. |
| SPX | 1,000 March 2027 8,000 calls, $21.1M, against open interest of 59,272 | The other half of the pair. Matched size on both sides reads as structure, not direction. |
| SPX | 740 December 2026 7,720 calls, $15.4M, volume 1.39× open interest of 697 | Volume above open interest on both legs of the same strike. Unconfirmed until Monday’s update. |
| SPX | 740 December 2026 7,720 puts, $15.3M, volume 1.38× open interest of 702 | The put leg. Same size, same strike, same second. |
| XLI | 22,500 November 2026 170 puts, $11.7M, on the low-historic-volume floor rule | A two-trade floor print in industrials, against open interest of 55,346. Size without a matching call leg. |
| SPXW | 1,486 September 25 7,650 calls, $6.2M, volume 2.24× open interest of 1,270 | A week-out call line written on the bid on floor. The nearest-dated line of size on the board. |
| MU | 900 September 25 1,010 puts, $2.1M, volume 41.4× open interest of 59 | Expires five days before the 30 September earnings print, so it is not positioning for the number. |
| DOCN | 3,190 October 2026 120 calls, $5.1M, all opening, against open interest of 132 | A large-cap floor trade flagged as fully opening, at 24× the prior open interest. |
| SPCX | 3,000 December 2026 150 puts, $4.2M, against open interest of 8,401 | Written on the ask in the name that carried the single largest dark-pool print of the session. |
Positioning into Monday
We are carrying no directional lean out of this session, and the reason is the composition of the tape rather than any balance of risks. A board whose largest component is a mechanical index cross tells you where shares had to move, not where anyone wanted them. The positioning gauge reads $0 across leveraged-long, inverse and volatility ETFs, which is genuinely unusual and reads as an absence of hedging rather than confidence in either direction.
The regime read stays Neutral, unchanged from the 17 September close, and the MPI sits at 62. Our observational markers are the two ends of today’s range. 757.97 on the downside and 762.00 on the upside bracket a session that resolved nothing, and the first clean close outside them is the first piece of new information Monday can offer.
The boards — data appendix
| Measure | 14 Sep | 15 Sep | 16 Sep | 17 Sep | 18 Sep | Change vs 17 Sep |
|---|---|---|---|---|---|---|
| $100M+ notional | $18.52B | $39.81B | $28.82B | $27.59B | $117.73B | +326.7% ($117.73B from $27.59B) |
| Prints | 74 | 127 | 86 | 101 | 379 | +275.2% (379 from 101) |
| Distinct names | 44 | 85 | 48 | 49 | 128 | +161.2% (128 from 49) |
| ETF share of $100M+ notional | 46.8% | 62.9% | 84.3% | 65.5% | 25.3% | −40.2 pp (25.3% from 65.5%) |
Every measure on the strip set a high for the five sessions we hold, and not by a small margin. Notional at $117.73B is 4.27 times yesterday’s $27.59B and 2.96 times the previous widest session on 15 September. Print count at 379 is roughly triple yesterday’s 101, and 128 distinct names is half again the 85 that 15 September carried. The one measure that fell is ETF share, down 40.2 points to 25.3% from 65.5%. Those four movements together are one fact stated four ways: an enormous amount of single-stock notional crossed at one moment, and index wrappers were not the vehicle.
| Strike | From spot | Call premium | Put premium | Net |
|---|---|---|---|---|
| 785 | +3.1% | $3.2M | $13.6M | −$10.4M |
| 770 | +1.1% | $11.3M | $18.1M | −$6.8M |
| 764 | +0.3% | $9.6M | $16.5M | −$6.9M |
| 761 | −0.1% | $51.3M | $60.0M | −$8.7M |
| 759 | −0.4% | $63.8M | $70.7M | −$6.9M |
| 758 | −0.5% | $25.7M | $35.5M | −$9.8M |
| 755 | −0.9% | $16.8M | $30.2M | −$13.4M |
| 745 | −2.2% | $6.1M | $25.8M | −$19.7M |
| 740 | −2.8% | $8.3M | $14.8M | −$6.5M |
| 720 | −5.5% | $17.7M | $5.7M | +$12.0M |
Nine of the ten heaviest strikes carry more put premium than call premium. The heaviest of all is 745 at −$19.7M net, 2.2% below the close, built from $6.1M of calls against $25.8M of puts. The nearest shelf on the other side is 720 at +$12.0M net, 5.5% below spot, and it is the only call-led strike in the set. Note that the two lean the same way in practice: a put shelf below the market and a call shelf further below it both describe premium already sitting under spot rather than anyone reaching upward. The at-the-money strikes carry the largest gross premium, with 759 at $63.8M of calls against $70.7M of puts, but their net figures are small because both sides are crowded.
The universe below is every off-exchange print of $100M or more on 18 September 2026, retrieved complete by pagination and de-duplicated on trade identifier, premium and size, with canceled prints dropped. It totals $117.73B across 379 prints in 128 names.
| Board | Reading |
|---|---|
| Participation | 128 distinct names across 379 prints totalling $117.73B. ETFs were 94 prints and $29.73B; single names and ADRs were 285 prints and $88.01B. No trailing baseline is computed in the cloud edition, so the five-session strip above is the comparison. |
| Concentration | Technology led the single-name notional at $38.99B, 44.3% of the single-name total, with Industrials second at $26.28B and 29.9%. SPCX was the largest single name at $18.81B, which is 16.0% of the whole board. |
| Positioning gauge | Leveraged long $0, inverse $0, volatility $0. No ETF in any of the three baskets cleared either the flagged-events screen or the $100M dark-pool screen, which reads as an absence of expressed hedging rather than a directional stance. |
| Threshold count | 379 prints at or above $100M, complete. |
Relative size is each print’s share count divided by that name’s 30-day average daily volume.
The relative-size tape is a $10M screen and the boards universe above is a $100M screen, so a name can appear in one and not the other. NHIC and CERT clear the first and not the second; the $117.73B headline figure comes from the $100M board alone.
| Ticker | × 30-day avg volume | Shares (of 30-day average) | Notional | Price | Time ET |
|---|---|---|---|---|---|
| NHIC | 5.10× | 3,000,000 of 587,755 | $31.95M | 10.65 | 11:59 AM |
| CERT | 0.98× | 2,021,731 of 2,068,819 | $16.96M | 8.39 | 11:49 AM |
| OTIS | 0.44× | 1,699,400 of 3,825,072 | $116.41M | 68.50 | 11:00 AM |
| VOYA | 0.36× | 346,000 of 953,914 | $34.77M | 100.50 | 2:57 PM |
| BWIN | 0.27× | 1,002,335 of 3,675,632 | $31.99M | 31.92 | 1:34 PM |
Cluster watch groups prints in the same name within a tenth of a percent of one price. The three largest today are all closing-cross groups, and their members are not narrated separately above.
| Ticker | Price | Prints | Aggregate | Window ET |
|---|---|---|---|---|
| SPCX | 152.71 | 17 | $18.81B | 4:00–4:56 PM |
| SPY | 762.67 | 23 | $12.36B | 11:02 AM–4:43 PM |
| BE | 265.63 | 13 | $5.34B | 4:00–4:15 PM |
Notable data points
- The $100M-and-above dark-pool board totalled $117.73B, more than the $114.74B of the previous four sessions combined.
- SPCX printed $18.81B across 17 prints, every one of them within a tenth of a percent of 152.71.
- ETF share of the board fell 40.2 points to 25.3%, from 65.5% on 17 September.
- All ten SPY regular-session dark-pool prints above $100M carried a contingent or derivative-priced condition, and all ten printed at 762.693, above the 762.00 session high.
- Market-wide net options premium finished at +$51.3M, its highest reading of the day, after a trough of −$67.3M at 12:35 PM ET.
- SPY call premium ran $935.6M against $799.1M of puts; QQQ ran $761.8M of calls against $637.2M of puts.
- Nine of the ten heaviest SPY strikes by net premium were put-led, the exception being 720 at +$12.0M.
- Six of nine contracts flagged on 17 September confirmed as new open interest, with no partials.
Method note
Flow and dark-pool data sourced from Unusual Whales. MPI score and regime classifier are our internal composite; daily synthesis is AI-assisted from those inputs. Volume-over-OI activity is unconfirmed as new positioning until the next morning’s open-interest update; confirmations are published in the following session’s post. Relative size is a print’s share count divided by that name’s 30-day average volume as reported with the print, computed by SomerQuant; it is not any vendor’s relative-size or relative-strength score. Option premium by strike is aggregated across all expiries. The session map is drawn from 5-minute bars; the opening range is the first fifteen minutes; dark-pool prints marked contingent carry the vendor’s contingent or derivative-priced sale condition. Board baselines are trailing-session averages computed from our own archived pulls; derived indicators are computed by SomerQuant from licensed market data.
Two vendor data issues are disclosed rather than smoothed. The VIX close returned as the bare integer 14 against a session low of 14.80 and a high of 15.63, and yesterday’s close returned as 15 against a low of 15.38, so both closes are integer-truncated and unusable. We print the session range instead of any computed percentage change. Separately, the SPY 5-minute bar beginning 1:25 PM ET returned a negative volume from the vendor, which is a cancellation correction against the canceled 4.5 million-share contingent print at 1:22 PM; that bar is drawn as zero in the volume panel. To keep the published payload inside its size budget, the session map, the five-session strip and the strike ladder each ship one responsive variant rather than separate wide and narrow drawings; the session tide and the confirm scorecard ship both. The flagged-events screen shows the 50 most recent alerts above $1M in premium, and all statements about it are bounded by that depth.
This is research, not advice. Nothing here is a recommendation to buy, sell, or hold any security.
AZTMM HLDGS LLC is not a registered broker-dealer, investment adviser, or FINRA member. All content is retrospective research published for general circulation — not personalized advice, not trade signals. Options involve substantial risk, including losses that may exceed the initial investment. Full disclaimer.
New here? Start Here · Pulse Lab · Accountability Ledger · Trading Academy · Subscribe for daily updates
Discover more from AZTMM HLDGS
Subscribe to get the latest posts sent to your email.
