Claims and the Philadelphia Fed came in hot enough to rescue the premise every bull case rests on, and the S&P closed within twenty-one points of a gamma flip it had been a hundred and fifty under. Then six trillion of notional lined up to roll off at the open, alongside the quarter's largest index rebalance.
What would change the read: reclaiming the flip in futures terms restores the dampened state, and the overnight December contract sits twenty-eight points beneath that translation — closer than all month, inside a single opening range. Industrial production at 09:15 is the smallest of the week's prints; the expiry outranks it.
| Instrument | Last | Change | Note |
|---|---|---|---|
| S&P 500 cash close Sep 17 | 7,637.76 | +85.95 · +1.14% | Through yesterday's roof; inside today's corridor. |
| Nasdaq Composite cash close | 26,418.30 | +439.87 · +1.69% | Led the rebound; led the loss too. |
| Nasdaq 100 cash close | 29,446.98 | +501.92 · +1.73% | Mega-cap did the lifting. |
| Dow Jones cash close | 51,778.04 | +316.14 · +0.61% | Half of Wednesday's 631 back. |
| Russell 2000 cash close | 2,874.63 | +15.82 · +0.55% | Last again. |
| ES · NQ Dec overnight | 7,700.25 · 29,745.50 | −0.09% · +0.01% | First session with December front. |
| YM · RTY Dec overnight | 52,249 · 2,895.50 | +0.06% · −0.05% | Flat into the expiry. |
| WTI Oct · Brent Nov | 101.08 · 103.85 | −0.81% · −0.93% | Third session softer, still a hundred handle. |
| Nat gas Oct | 2.863 | −1.31% | The one energy contract with no war premium. |
| Gold · Silver Dec | 4,390.20 · 66.41 | −0.22% · +0.48% | Settled +1.90% and +3.77% Thursday. |
| Copper Dec | 6.6445 | −0.26% | Within 3% of its 52-week high. |
| US 10Y | 4.939% | Thu close 4.947% | Back under five — the whole rally. |
| US 2Y · 5Y · 30Y | 4.683 · 4.792 · 5.286 | — | The 2Y sits 81bp over the funds midpoint. Section 11. |
| JGB 10Y · 30Y · Bund · Gilt | 2.981 · 4.086 · 3.491 · 5.239 | — | Gilts, the highest ten-year in the G7. |
| DXY · USD/JPY | 100.247 · 157.09 | flat · yen −0.72% | Yen weaker after a BOJ hike. |
| EUR/USD · GBP/USD | 1.1485 · 1.3370 | +0.10% · +0.11% | — |
| VIX · VVIX · SKEW | 15.44 · 87.72 · 145.70 | −12.82% · −8.06% | Vol-of-vol fell harder than vol. |
| Bitcoin · Ether | 77,321 · 2,475 | +1.06% · +1.10% | Tracking equity risk, not rates. |
| Nikkei · Hang Seng · Shanghai Friday | 65,272 · 24,798 · 3,916 | +1.77% · +0.79% · +1.04% | Asia rallied through the BOJ. |
| Stoxx 600 · DAX · FTSE Thu close | 642.60 · 25,717 · 10,816 | +0.55% · +1.40% · +1.19% | Green before we were. |
| GNRC · VICR Thu close | +18.34% · +17.66% | — | Amazon generators; AI power licensing. |
| FLNC Thu close | −15.36% | — | Guidance cut on plant delays. |
| GFL · NVDA · TSLA · LEN · GOOGL | +5.26 · +2.54 · +2.27 · +1.71 · +1.30 | — | Per cent, Thursday's close. Details in Section 09. |
| Gauge | Reading | What it says |
|---|---|---|
| CNN Fear & Greed | 28.7 Fear | Still deep in fear after a strong session; a month ago, 59.1. |
| AAII week ended Sep 16 | 28.8 / 17.9 / 53.3 | Bulls, neutral, bears. Two technicians independently called this a contrarian buy. |
| CBOE put/call | 0.79 total · 0.52 equity | Index ratio 1.07. Single-stock positioning is not defensive; index hedging is. |
| Breadth · 200-DMA | 50.29% | Back over the line after dipping to 48.11 on Wednesday. Section 06 has what that dip triggered. |
| Breadth · 50-DMA | 30.81% | The shorter series is the one that has broken. |
| Fed path · Oct 28 | ~55% hike | Secondhand read on exchange pricing; the tool would not render. Prediction markets: October 51%, December 66%. |
| Dealer gamma | NEG · session 6 | Set from the prior close against the flip. Levels in Section 05. |
The tape and the gauges swapped places. Yesterday the gauges were bearish and the tape was heavy; today the gauges are just as bearish and the tape has ripped. Fear and Greed at 28.7 after an eighty-six-point session, bears at 53.3% and a bull-bear spread at a two-year low describe a crowd that did not participate in the rebound — the setup a flow desk wants to fade in the other direction. Against that, supply has not improved: Citadel Securities told clients in early September that the buyback window begins closing around September 12, so the corporate bid is thinning rather than returning. The one number that genuinely changed is the flip, which fell 45 points while cash rose 86 — the option book repricing beneath the tape, not the tape catching up to the book, on the last full session before that book expires.
| Time ET | Event | Consensus | Prior |
|---|---|---|---|
| 09:15 | Industrial production, Aug | +0.2% m/m | +0.2% |
| 09:15 | Capacity utilisation, Aug | ≈76.3% | 76.3% |
| 09:30 | Gov. Bowman — Vice Chair for Supervision, London | stress testing | — |
| 10:00 | Conference Board Leading Economic Index, Aug | +0.2% m/m | — |
| 13:00 | Baker Hughes rig count · 15:30 CFTC positioning | — | — |
| 16:00 | Quad witching + quarterly rebalance — 122 indices, 25 S&P 1500 changes | — | — |
| done | Bank of Japan +25bp to 1.25%, 7–2 | as priced | 1.00% |
Consensus for the 09:15 prints and the LEI are aggregator estimates rather than a polled survey, and are indicative. The Bowman event, venue and billed topic are confirmed from the Board's own calendar.
| Gamma level | SPX | ES Dec · +69.49 | Role in today's tape |
|---|---|---|---|
| Gamma flip | 7,658.27 | 7,728 | The regime boundary, 20.51 above cash — the narrowest gap of the month, from 151.52 in one session. It fell 45 points while cash rose 86; over a third of the convergence came from the book, not the tape. |
| Call wall | 7,650.00 | 7,719 | Ceiling, 12.24 above cash. Up 50 points overnight, the first shift in five sessions — and it sits beneath the flip, so reclaiming the roof does not by itself flip the regime. |
| Put wall | 7,600.00 | 7,669 | Floor, 37.76 under cash, unchanged for a seventh session. Cash sits inside a fifty-point corridor, nearer the roof. |
Levels are from a public dealer-gamma (GEX) model on the Sep 17 close, all-expiry. Regime is read from cash against the flip and nothing else.
Per-strike open interest tells a different story from the aggregate, and that difference is this morning's whole point. On the October index chain — the first book that exists after today — the heaviest call open interest above spot sits at 7,800 and the heaviest put open interest below spot at 7,500, with a lighter put shelf at 7,600; the December E-mini's own near-dated book peaks at the same two strikes. A one-day futures book and a twenty-eight-day index book agreeing on both sides says the wide structure is not an artefact of one expiry. So the aggregate corridor is fifty points wide and the one that survives the open is nearer three hundred. Neither is wrong — the aggregate is dominated by the September quarterly retiring at this morning's bell, and once it is gone the remaining book has far less to defend near cash. One caveat: heavy open interest sits at the top of the observable window, so the upper peak reads as the start of a shelf rather than a confirmed maximum.
The term structure is in contango and steepened through the rebound rather than flattening on it. October VX settled at 17.97, November 18.65, December 19.03 — a 16.4% premium on the front contract against spot, the widest of the week. The market marked down immediate risk and marked up everything past the expiry — what you would expect if the whole vol bid is concentrated in what happens once today's book disappears. On the pivot table, the December E-mini pivot is 7,681.83 with R1 at 7,746.17 and S1 at 7,641.42; the overnight contract sits eighteen points above the pivot and well inside the first band either way — no technical opinion going into the quarter's largest mechanical print.
Her bull case is not that the Fed is nearly done — it is that the destination does not matter. "The rate hikes are not coming simply from supply shocks. The rate hikes are a reaction to a strong economy and the AI ecosystem trade. The capex is not going to be slowed or deterred because we get 25, 50 or 75 basis points more hikes." The instruction: "Let the bond market be the bond market and the yields be the yields… The signal is strong earnings."
The evidence is a comparison against her own former self: a sceptic in January at 22.5 times forward earnings, she puts it at 19.4 today with less concentration and better revision breadth — which is why she calls this a healthier market than the one she feared. She concedes where the damage lands without hedging: small caps, input-cost-sensitive discretionary, housing. Note which index finished last again yesterday. Identity guard: Mike Wilson holds a separate Morgan Stanley CIO title with a separate target.
He dissented from the hike in the opposite direction from the week's loudest dissent. Asked whether he agreed with Gundlach's call for fifty: "I do not… the market had boxed them into a corner with a 92% probability of a 25 basis point hike… I myself would have preferred that they don't hike."
His analytical point is that the bond market was never pricing what everyone said it was. "For all the talk about credibility tests, about inflation concern — if you really look at the ten year… they haven't moved that much over the last few days." What did move yields: "an imbalance between increasing demand, especially from the hyperscalers for bond financing, and fewer reliable buyers and holders." Hold that sentence; Section 11 takes it where nobody on air took it. His conclusion refuses to pick a scenario, which is itself the call: "There is no dominant scenario, and investors should expect a much wider range of possible outcomes."
He told clients to fade the exact rebound the television bench spent the afternoon celebrating — on mechanics, not on a view. "The supply/demand setup into month-end remains unfavorable, the technical backdrop is still working against equities, and we continue to think equities can trade lower over the next two weeks." The reasoning is inventory: "several important sources of demand are either fading or already deployed, while the potential sources of supply are increasing." His own firm's work names the supply event — this morning's expiry. A specific mechanism on a two-week clock, and the most gradeable thing anyone said this week.
He produced the most-quoted bearish datapoint of the week and the market undid it in one session. Writing Wednesday: "The percentage of SPX names > 200 DMA closed at 49%. This is the first time since April '00 when this closed below 50%, while the SPX itself was at least 4% above its 200 DMA and within 4% of an all-time high." A twenty-six-year precedent, correctly conditioned — for Wednesday's close. By Thursday the series was back over the line, and the signal had lasted a day.
What has not recovered is the shorter series in Section 03, and his framework is better served by it: "you just aren't getting rewarded for buying strength. More and more stocks are breaking support and their 200 DMAs." None of his three conditions for calling the correction over was met yesterday — and the first, less complacent sentiment, is the live dispute.
He refused the rates question on air, twice. Asked about financial conditions after the decision, his entire first answer was: "Did rates go up?" Told they had: "it doesn't matter in long term investing… 25 basis points this way or that way, or 50 or 75, it just doesn't matter."
The contrarian line inverts the standard fear. "The biggest impediment to the data centers being built and AI factories being built is power… we just can't build it fast enough." Then: "If there's a slowdown, that's really good because we will be able to build."
| Voice | Stance | Takeaway |
|---|---|---|
| Tom Lee Fundstrat · 8,200 | BULL held | Hardened, not changed: "A max hawkish Fed is a bullish setup." |
| Ed Yardeni Yardeni Research · 7,900 | BULL held | No new number after the cut from 8,400. Reads the war and higher-for-longer oil as "the deciding factor" in the vote. |
| Savita Subramanian BofA · 7,400 | BEAR raised | Raised from 7,100 on Sep 14, 12-month 7,800 — still the Street's lowest, and now below spot. |
| Andrew Sheets Morgan Stanley, Fixed Income | HAWK | House call: two more 25bp hikes, December and March, to 4.25–4.50%, then on hold through 2027. Not Mike Wilson's call. Gapen and Hornbach read the committee as "thinking in terms of more than one move." |
| Ben Snider Goldman Sachs | NEUT | Base rate: a 2% average decline over three months at the start of past hiking cycles, a 9% gain over twelve. Section 08 has his sharper point. |
| JPMorgan trading desk no individual named | CAUT | "Tactically Cautious/Neutral." What flips it bullish: "lower oil/bond yields" — over the same fortnight Rubner named. |
| Mark Newton Fundstrat | BULL held | Equities "remain in good shape despite the threats of higher Crude and long-term interest rates." His breadth measure is the Russell 3000, a different universe from Section 06's. Chris Verrone, now billed Baird Strategas, had nothing fresh. |
| David Keller Sierra Alpha Research | REVERSED | A complete turn in twenty hours — the bull flag "failed" below 7,600 in the morning, "The Bulls Just Took Back Control of 7600" by night. |
| Brian Levitt Invesco | BULL | The cycle ends "when something breaks in the AI trade… But that's not the current environment." |
| Victoria Fernandez Crossmark Global | NEW CAUT | Took the other side of Levitt on the same panel: if capex goes, "who is going to step up?" Section 11 has her second-order point. |
| Phil Palumbo Palumbo Wealth | NEW BULL | The most specific path call from any guest: "one and done. Two and done, maximum." Hedged for the alternative anyway. |
| Mark Mahaney Evercore ISI | NEW BULL | Single-name, not index. Alphabet to $450 — Section 09. |
| The Kobeissi Letter | DARK | Thirteen posts in window, none independently verified. The most-shared numbers on the tape this week have no trusted carrier. |
| Thursday's prints | Actual | Consensus | What it settled |
|---|---|---|---|
| Philadelphia Fed manufacturing | 37.8 | 30.5 | The premise test, passed by seven points. A seventeen-point step-down was written into expectations; the survey declined to deliver it. |
| Initial jobless claims | 196K | 208K | Below every estimate; continuing claims 1.73M against 1.78M. No crack. |
| Housing starts | 1.275M | 1.31M | −2.6% on the month. The rate channel working as designed. |
| Building permits | 1.394M | 1.41M | −2.7%, and forward-looking. |
| Pending home sales | +0.3% | +2.0% | A seventh of the expected bounce, index down 4.7% year on year. |
Two economies printed inside the same hour and the market read one of them. Manufacturing sentiment and the labour market came in strong enough to validate the hike; everything touching a mortgage came in soft enough to show the hike already working. That is not a contradiction — it is transmission on schedule, arriving in housing first because housing is where it always arrives first. The question the tape skipped is whether a committee that sees the strong half keeps tightening until the soft half reaches the aggregate data, which by construction takes two more quarters.
Ben Snider's framing at Goldman explains yesterday better than the data does. Equities react not to the level of yields but to the speed of the move — more than fifty basis points in a month, or thirty in a fortnight, is where stocks historically stop absorbing it. On Thursday the ten-year went the other way for the first time in weeks and the response was immediate and disproportionate: the same mechanism in reverse, and the reason a five-basis-point move in a yield produced an eighty-six-point move in an index.
Abroad, two central banks moved in one week and neither market cared. The Bank of Japan raised to 1.25% overnight, the highest since 1995, on a 7–2 vote — and the yen promptly weakened while the Nikkei rallied. Ueda gave no forward guidance. The Bank of England held with three votes for a hike. The pattern is identical across all three decisions: fully priced, no guidance, reaction in the currency rather than the rate. The strain sits in the long end, where the thirty-year JGB above four is a number that did not exist in this century until recently.
Generac was the tape's clearest statement of what it actually wants to own. The stock added 18.34% on an agreement to supply generators for Amazon data-centre capacity, reported at $2.4 billion initially with a path toward eight. A backup-power manufacturer repricing by nearly a fifth on one hyperscaler contract is the same trade as Vicor's 17.66% on an AI power-delivery licensing deal — the electricity constraint Brookfield's chief executive described in Section 06, arriving as a supply-chain scramble rather than a theme. Fluence took the other end, down 15.36% after cutting guidance on Houston plant delays. Adjacent businesses, opposite outcomes, one week — the dispersion El-Erian said to expect, and which the index-level gauges cannot see.
Alphabet is the one large-cap carrying a fresh, fully specified call. Evercore ISI's Mark Mahaney raised his target to $450 from $420 on his tenth proprietary search survey, at 25 times a 2028 estimate of $18.11 — roughly 30% above the current price. The argument is narrow and testable: "Search is still 50% plus of revenue and probably 70% plus of the profits at Google." The survey's finding is a reversal — three years of share loss to ChatGPT through early 2025, share regained since, "because Google's Search product got just dramatically better — thank you to that competition." He named his own risk unprompted: "Nobody's guaranteed any leadership position."
The week's real damage sits in a sector nobody is discussing. Goldman Sachs fell 7.6% on the week and Bank of America 7% — the worst week for either since April 2025 — after both chief executives guided investment-banking and trading revenue lower at an industry conference. The banks wore the hike and then did not participate in the recovery. Elsewhere: Lennar's miss is the housing data in single-name form and the tape ignored it; Nvidia added 2.54% into reports that Huawei unveiled two AI chips for 2027, a competitive story the market has now declined to trade three times; Tesla recovered 2.27% against a federal probe of its robotaxi programme.
The blackout lifted Wednesday evening and the Board has posted nothing since — no speeches, no research, three days after a unanimous 12–0 hike to 3.75%–4.00%. Governor Bowman is the first official voice at 09:30, in London, billed on stress testing rather than policy. The projections are talking instead: sixteen of eighteen participants see at least one more move this year and the 2027 median matches 2026. October prices near a coin flip on the exchange's measure and 51% in prediction markets, December at 66%, with essentially nothing priced for a cut at either — a hike-versus-hold market, not a hike-versus-cut one. Watch whether Bowman takes a policy question on a foreign stage, whether the ten-year holds under five without help, and whether the morning's prints move the October number at all.
Three people described one mechanism yesterday and none joined the pieces. El-Erian said the yield surge is about hyperscaler demand for bond financing meeting fewer reliable buyers, not credibility or inflation. Levitt said the cycle ends when a hyperscaler pulls back on investment. And Fernandez, taking the other side of Levitt on the same panel, was the only one to follow the chain to its end: if capex slows, "what does that mean in terms of the debt issuance that we're seeing from them?… there's a component of the higher yield story that it's coming from that competition for capital. If that pulls back… perhaps you actually see yields come down a little bit."
That breaks the correlation every allocator is relying on. If AI capex is a meaningful share of the marginal supply of corporate paper, the event that damages AI equity is the same one that rescues duration — the index down and the ten-year with it, which looks like an ordinary risk-off day except that it arrives with the Fed still hiking, so the front end would not follow. Every "bonds will work when equities break" argument made this week quietly assumes those two bids are independent. Fernandez's point is that they are two sides of one flow, and nobody asked her the follow-up.
Read the settles rather than the headline. Equities rose — and so did the long bond, gold, silver and copper: the E-mini +1.10%, the Nasdaq contract +1.63%, the thirty-year +1.11%, gold +1.90%, silver +3.77%, copper +2.67%, all in one session two days after a hawkish hike into an energy shock. Stocks and bonds and precious and industrial metals do not rally together on a growth story or a policy story, because those stories need one of them to be the funding leg. They rally together when the price of money itself is being marked, or when something was being forced out and has stopped.
The only complexes that fell were energy and the softs — the opposite of what a growth re-acceleration produces. Thursday has been narrated as relief over a regional survey, which is a great deal of work for a five-basis-point move in a yield. A universal bid across four asset classes on the session before the quarter's largest expiry is more consistent with position-covering into that expiry than with anyone's revised view of the economy — and if so, the rally is an artefact of the structure that disappears at the bell, not evidence against it.
The two-year yields 4.683% against a funds midpoint of 3.875% — eighty-one basis points of tightening embedded in the front of the curve, which Morgan Stanley's macro desk reads as roughly three more moves. Prediction markets put December at 66% and "another hike in 2026" at 84%: call that one move with change. Those are not the same forecast. One market is priced for a cycle and the other for a gesture, and the gap between them is worth something like fifty basis points on the two-year.
Every desk note this week quoted whichever measure suited its argument — the hawks cite the curve, the one-and-done camp cites the odds — and none observed that they cannot both be right. The resolution matters more than October does: it is the difference between a front end already carrying the full cycle and one with two repricings left. The technician quoted in Section 07 came closest and stopped short of it. The two-year is telling a different story from the one everybody is quoting.
This publication is provided by Cannon Trading Company for informational and educational purposes only. Content may include market commentary, technical observations, analyst opinions, and aggregated material derived from publicly available sources. While such information is believed to be reliable, Cannon Trading Company does not author, independently verify, endorse, or guarantee the accuracy, completeness, or timeliness of any thirdparty information referenced or summarized herein.
The information, opinions, market data, and commentary contained in this publication are subject to change at any time without notice and do not constitute investment advice, a solicitation, or a recommendation to buy or sell any security, futures contract, option on futures, foreign currency transaction, or any other financial instrument.
Past performance is not indicative of future results.
Trading Futures, Options on Futures, retail offexchange foreign currency transactions, and other derivatives involves substantial risk of loss and is not suitable for all investors. You may lose all or more than your initial investment. Carefully consider whether trading is appropriate for you in light of your experience, objectives, financial resources, and other relevant circumstances.
Cannon Trading Company does not guarantee any profits and makes no representation that the strategies, ideas, analyses, or information presented will result in profitable trades or avoid losses. Any market views, analyst calls, forecasts, or thirdparty commentary referenced reflect the opinions of their respective authors and may or may not align with the views of Cannon Trading Company.
HYPOTHETICAL PERFORMANCE RESULTS HAVE MANY INHERENT LIMITATIONS, SOME OF WHICH ARE DESCRIBED BELOW. NO REPRESENTATION IS BEING MADE THAT ANY ACCOUNT WILL OR IS LIKELY TO ACHIEVE PROFITS OR LOSSES SIMILAR TO THOSE SHOWN. IN FACT, THERE ARE FREQUENTLY SHARP DIFFERENCES BETWEEN HYPOTHETICAL PERFORMANCE RESULTS AND THE ACTUAL RESULTS SUBSEQUENTLY ACHIEVED BY ANY PARTICULAR TRADING PROGRAM.
ONE OF THE LIMITATIONS OF HYPOTHETICAL PERFORMANCE RESULTS IS THAT THEY ARE GENERALLY PREPARED WITH THE BENEFIT OF HINDSIGHT. IN ADDITION, HYPOTHETICAL TRADING DOES NO INVOLVE FINACIAL RISK, AND NO HYPOTHETICAL TRADING RECORD CAN COMPLETLEY ACCOUNT FOR THE IMPACT OF FINANCIAL RISK IN ACTUAL TRADING. FOR EXAMPLE, THE ABILITY TO WITHSTAND LOSSES OR TO ADHERE TO A PARTICULAR TRADING PROGRAM IN SPITE OF TRADING LOSSES ARE MATERIAL POINTS WHICH CAN ALSO ADVERSELY AFFECT ACTUAL TRADING RESULTS. THERE ARE NUMEROUS OTHER FACTORS RELATED TO THE MARKETS IN GENERAL OR TO THE IMPLEMENTATION OF ANY SPECIFIC TRADING PROGRAM WHICH CANNOT BE FULLY ACCOUNTED FOR IN THE PREPARATION OF HYPOTHETICAL PERFORMANCE RESULTS AND ALL OF WHICH CAN ADVERSELY AFFECT ACTUAL TRADING RESULTS.
Cannon Trading Company is registered solely as a commodities broker. Nothing contained herein constitutes the provision of investment advisory services.
© 2026 Cannon Trading Company, Inc. All rights reserved.
cannontrading.com • 18004549572 • (310) 8599572