Closing Briefing — Tuesday, September 1, 2026

Source: Finviz group screener, Performance table view (g=sector&v=140&o=name), read after the close. 1-Day is the Change % column, 1-Week Perf Week, YTD Perf YTD. Finviz classification, not GICS.

YTD reconciliation, and it is the tightest of the reporting window. Compounding each group's 31 August YTD by Tuesday's one-day move reproduces the published YTD to within 0.01 percentage points for all eleven groups, with a maximum deviation of 0.010 pp in industrials. Worked examples: energy 1.3839 × 1.0169 = 1.4073 → +40.73% against a published +40.73%, deviation 0.001 pp; technology 1.2409 × 0.9875 = 1.2254 → +22.54% against +22.53%, deviation 0.009 pp. The real-estate drift this report tracked for three sessions last week remains closed, at 0.001 pp.

The rotation is defensive with an energy engine, and it is the first genuinely defensive tape of the reporting window. Four groups closed green and three of them are defensives — utilities +0.67%, healthcare +0.63%, consumer defensive +0.18% — with energy at +1.69% the only cyclical bid. CNBC's GICS tally agrees on the shape, marking energy up 1.3% and consumer discretionary down 1.9% intraday against Finviz's +1.69% and -1.76%, and counting the same four positive sectors. The utilities number is the mirror of Monday's trap: the same broad market-cap-weighted bucket that hid a 23% collapse in Edison now hides an 8.93% rebound, and the group closed up less than seven-tenths of a percent.

The composition traps to name. Basic materials at -1.95% was the worst group on copper's 2.25% fall, with Freeport-McMoRan -4.32%, Vulcan Materials -3.67% and Sherwin-Williams -2.62% — a growth signal filed under commodities. Communication services at -0.55% looks resilient only because Meta rose 1.08% inside it while Alphabet A fell 1.28%. And technology at -1.25% understates the semiconductor damage badly: SOX fell 2.14%, because Finviz's technology bucket also carries Apple, which rose 2.61%. On the year the extremes are wider than a week ago: energy +40.73% and technology +22.53% against consumer cyclical -5.18%, communication services -2.52% and utilities -1.20% — and consumer cyclical has lost 1.70 percentage points of its year in two sessions.

Higher

Lower

Analyst and corporate actions

Sourcing, disclosed. The Earnings Whispers day pages remain behind a cookie-and-usage-agreement consent banner, which this unattended session did not accept. The rosters below are captured from the Nasdaq earnings calendar API for each date and screened name by name against an S&P 500 constituent list. Nasdaq publishes a before-open / after-close bucket rather than a clock time, so no clock times are asserted this session; confirm every time against company investor relations before trading a date.

Wed 9/2. BMO: Brown-Forman (BF.B). AMC: Broadcom (AVGO), Hewlett Packard Enterprise (HPE), NetApp (NTAP).

Thu 9/3. BMO: Campbell's (CPB), Toro (TTC). AMC: Lululemon Athletica (LULU). Timing bucket not published: Copart (CPRT) — the reviewed calendar carries no before-open or after-close designation for this date; confirm with company investor relations.

Fri 9/4. No S&P 500 reporter on either bucket.

Mon 9/7. U.S. equity markets are closed for Labor Day. The Nasdaq capture returns no S&P 500 reporter for the date.

Tue 9/8. Timing bucket not published: Oracle (ORCL) — the reviewed calendar carries no before-open or after-close designation; confirm with company investor relations.

Wed 9/9. AMC: Cooper Companies (COO).

Thu 9/10. AMC: Adobe (ADBE).

Fri 9/11. BMO: Kroger (KR).

Changes vs. the prior calendar (8/31 report):

Source: U.S. Department of the Treasury Daily Treasury Par Yield Curve Rates, Text View for September 2026, read after publication. Changes are versus the 31 August official row (1-day) and the 25 August official row (1-week).

Off-table bills, extracted and reported here because they carry the financing story. 1.5 Mo 3.88% (+2 bp on the day, +10 bp on the week), 2 Mo 3.89% (+1 bp, +9 bp), 4 Mo 3.97% (+1 bp, +8 bp), 6 Mo 4.00% (+1 bp, +5 bp). The entire bill curve moved 0 to 2 bp on a day the 3-year and 5-year moved 6. The 6-month, which fully spans the 16 September meeting, cheapened a single basis point while the 2-year cheapened five — a second consecutive session in which the bill strip declines to confirm the coupon curve. See Section 9 block b.

The read: a belly-led bear flattener, and the exact mirror image of Monday. Walk the curve outward and the shape is a hump — 1-month 0, 3-month +1, 6-month +1, 1-year +2, 2-year +5, 3-year +6, 5-year +6, 7-year +4, 10-year +4, 20-year +3, 30-year +2. The maximum move sits at three and five years and decays in both directions. That is the signature of policy-path repricing, not term premium: the tenors that price the next eighteen to sixty months of the funds rate moved three times as much as the thirty-year. 2s30s flattened 3 bp to 88 and 2s10s 1 bp to 40, reversing Monday's 3 bp and 2 bp steepening; 3M10Y widened 3 bp to 87 only because the bill is anchored at the front of it.

The diagnostic that matters is what did not move, and it is the long end. On a day when Britain's 10-year rose 16 bp to 5.22%, the highest since June 2008, the 30-year gilt 10 bp to 5.8909%, the highest since March 1998, Germany's 30-year touched the highest since 2011 and Australia's set a record, with a Bloomberg index of global sovereign yields at its highest in almost two decades — the U.S. 30-year rose two basis points. The American long bond outperformed every developed peer in a global long-end rout. Bloomberg's own framing is the counterweight: the 30-year at 5.27% is back to the level "seen moments before Bessent announced the move on Aug. 19," so the buyback bought a fortnight, not a level, and the 10-year is more than 10 bp above where it sat then. Bank of America's Mark Cabana: "The rates market has not been able to hold any type of significant rate decline." The expanded operations do not begin until 9 September, and Treasury said only that it would "at least double" their size.

The week now has one unambiguous shape and it is a bear flattener. On five sessions the 2-year is 22 bp cheaper at 4.39%, the 3-year 21 bp, the 5-year 20 bp and the 1-year 17 bp, against the 10-year's 15 bp, the 20-year's 11 bp and the 30-year's 10 bp. 2s30s has flattened 12 bp on the week and 2s10s 7 bp. Jackson Hole moved the front, oil moved the back for one session, and the front took it back with interest. The vendor gap is again negligible: Bloomberg's live board marks the 10-year at 4.80% with a +5 bp change against the official par 4.79% and +4 bp, a one basis-point difference in both level and change and a routine artefact of a 4:59 p.m. live mark against a 3:30 p.m. bid-side construct. Bloomberg separately marks the 2-year at 4.40%, +6 bp, against the par curve's 4.39% and +5 bp.

Source: Federal Reserve Bank of New York Economic Indicators Calendar for September 2026 (all times Eastern). Consensus figures are carried where independently verified against the Wall Street Journal's U.S. economic calendar or Bloomberg's economist survey; where none is verified, the sensitivity note describes what the market is positioned for instead of asserting an expectation.

Current week (Aug 31 - Sep 4) — still to come

Next week (Sep 7 - Sep 11)

Current target range: 3.50%-3.75%. Two vendors twelve minutes apart, a 1.3 percentage-point gap, and a prior-session live read that this time proved exact.

CME FedWatch headline — 16 September 2026 meeting. Data as of 1 Sep 2026, 04:57:35 p.m. CT (5:57 p.m. ET), read from the FedWatch probability table.

Provenance of every column, stated — and Monday's live read vindicated. CME's NOW column carries a 04:57:35 CT timestamp against a wall clock after 5:00 p.m. ET and a 4:00 p.m. CT ZQ session close, so it resolves as p.m. and sits within the hour after the close. The 1 DAY column carries the legend date 31 August and prints 65.4% — precisely the figure this report published from CME's live column on Monday evening. After two consecutive sessions in which the live read overstated the settled figure by 1.5 pp and then 2.7 pp, the third reproduced to the decimal, which argues the earlier gaps were a settlement-window artefact rather than a systematic bias. Treat the NOW column as indicative to roughly one to three percentage points and confirm it in the next edition. 1 WEEK (25 August) and 1 MONTH (31 July) carry genuine reference dates and are used in calculations. The Investing.com matrix below is timestamped 1 Sep 2026, 05:45 p.m. EDT and is the primary source for parts (a), (b) and (c).

The CME-versus-Investing.com gap, quantified. CME puts the September hike at 66.9% at 5:57 p.m. ET; Investing.com at 65.6% at 5:45 p.m. ET — a 1.3 percentage-point difference across twelve minutes, against 2.0 pp on Monday and 3.8 pp on Friday, so the vendors are converging as the meeting approaches. The gap is large in probability and trivial in price. Investing.com publishes the September future at 96.290, down 0.5 bp from Monday's 96.295; because the 16 September meeting sits mid-month, only about 47% of the contract's averaging period is affected, so one basis point of ZQ price is worth roughly ten percentage points of hike probability. The entire 1.3 pp vendor gap is therefore thirteen-hundredths of a basis point of contract price. Bloomberg reports traders "pricing in some 17 basis points of tightening at the Fed's Sept. 15-16 meeting, or odds near 70%," a third construction sitting above both vendors; the three-way spread of about 4.4 pp is under half a basis point of price. That leverage remains the most important number in this section.

One-day, one-week and multi-day momentum. The September hike rose 1.5 pp on CME's own columns (65.4% to 66.9%) and 1.2 pp on Investing.com's (64.4% to 65.6%) — a fifth of Monday's move and a fifteenth of Friday's, on a day when the one high-sensitivity release missed. That is the tell: the front meeting barely moved on a soft ISM, while the entire 2027 strip cheapened. On a one-week view it is 66.9% against 39.6% on CME and 65.6% against 35.0% on Investing.com, so 27.3 pp and 30.6 pp in five sessions. The one-month column remains the discipline: on 31 July CME priced 67.0% at 375-400, so Tuesday's 66.9% is one-tenth of a point below where the market sat a month ago — a completed round trip, not a new regime. Further out, every 2026 horizon extended: October's cumulative-above went to 77.4% from 73.5%, December's to 91.0% from 88.9%, and the peak of the strip is now 97.1% across June and July 2027 against 96.0% on Monday. The probability of a cut at any 2026 meeting remains 0.0%.

(a) Current-year meeting distributions

Investing.com Fed Rate Monitor, updated 1 Sep 2026 05:45 p.m. EDT. Format: current [prior day] [prior week]. Modal range in bold.

Sums are 100.0%, 100.0% and 100.1% on the published figures. Three observations. First, September's hold column fell only 1.2 points on the day but 30.6 points on the week, from 65.0% to 34.4% — the meeting is now priced, and the marginal information is arriving further out. Second, the deep tail is where the day's move went: December's +75 bp bucket rose to 13.5% from 10.3%, a 3.2-point jump, and October's +50 bucket to 22.6% from 17.8%, a 4.8-point jump, while both meetings' +25 buckets fell. Third, the modal outcome at December is two hikes at 41.9% for a third consecutive session, with the gap over one hike widening from 0.8 points to 6.3. The market has stopped adding confidence to the first hike; it is adding hikes.

(b) Next-year meeting path

Modal range, its probability, and the cumulative probability above and below the current 3.50%-3.75% range, with the contract price that draws it.

The 2027 strip moved, and after a week of not moving that is the most important line in this section. Every contract cheapened: 95.960, 95.855, 95.800, 95.740, 95.725, 95.725, 95.725, 95.755 against Monday's 95.985, 95.885, 95.835, 95.780, 95.770, 95.780, 95.775, 95.810 — 2.5 to 5.5 basis points across all eight points, with the largest moves at the September and December 2027 tails. The trough deepened to 95.725 and now spans July, September and October 2027 rather than sitting alone in July. July 2027's modal range flipped to 4.25-4.50% at 30.3%, one-tenth of a point above 4.00-4.25% at 30.2% — a statistical tie, but the first time in the reporting window that any meeting's mode has sat at three hikes. Cumulative-above rose at every point: December 2027 to 94.7% from 92.7%, June to 97.1% from 96.0%. The first non-trivial cut probability is still only 0.5% at 3.25-3.50% in December 2027, and it fell from 0.8%. Read against Monday, when an eight-point jump in the September probability produced nothing beyond eighteen months, this is a market that has stopped treating the oil shock as transitory in level.

(c) Year-end probability ladders

Year-end 2026 — the 9 December meeting.

Cumulative above the current range: 91.0%. Cumulative below: 0.0%. Sum: 100.1%.

Year-end 2027 — the 8 December meeting.

Cumulative above the current range: 94.7%. Cumulative below: 0.5%. Sum: 99.9%.

Rounding, transparently. Every figure is reproduced at the vendor's own one-decimal precision. Column sums of 99.9% or 100.1% are rounding artefacts of that precision, not missing probability mass; no cell has been rescaled, and cells shown as 0.0% are ranges the vendor's own card either omits entirely or publishes as zero, which under CME methodology means a probability below the rounding floor. The December 2026 row sums to 100.1% and the January, June, July, September, October and December 2027 rows to 99.9% for exactly this reason.

(a) IG and HY credit spreads

ICE BofA option-adjusted spreads via FRED. FRED publishes with a one-business-day lag: the levels below are as of 31 August 2026, not the 1 September close. Same-day direction is cross-checked against the cash-market proxies underneath and against Bloomberg's credit coverage.

CDX — the full six-step ladder was worked again and is reported so the gap stays auditable. (1) Bloomberg in Chrome: /markets and /markets/rates-bonds were rendered; the Fixed Income Indices table and the global government-yield boards populate, a full-text scan returns zero occurrences of the string, and no CDX line appears. (2) WSJ Market Data bonds page: rendered, zero occurrences in a full-text scan. (3) Cbonds carries dedicated CDX.NA.IG 5Y and CDX.NA.HY 5Y index pages but masks the levels behind a subscription; S&P Dow Jones Indices and ICE publish index-family, options and constituent-list documentation, not the daily running spread. (4) FT Markets Data and Reuters credit wraps returned no dated CDX quote for 1 September. (5) TradingView and Barchart symbol searches for CDX resolve to an unrelated ETF, and the CME CDS index product pages carry specifications rather than levels. (6) Cash-market proxies, labelled as proxies: HYG closed $79.10, -0.89%, and LQD $105.22, -0.93%, against 31 August closes of $79.81 and $106.21. HYG's duration is roughly a third of LQD's, so a 4 bp move in the 10-year explains perhaps 0.14% of HYG's loss and the remaining 0.75% is credit — the same direction as the CCC series and a genuine same-day widening signal. No CDX level is published here, because an undated third-party digest number is not a CDX level.

The tail dissented, loudly, and the aggregate barely noticed. CCC widened 16 bp to 1,042 bp on the 31 August update — the largest single-day move in the series in the reporting window, and enough to push the CCC-minus-HY differential to 779 bp from 766, a 13 bp widening that takes it back above the extreme this report flagged last week. Against that, IG widened one basis point to 80 and HY three to 263, so the aggregate is still 18 bp through where 2026 started at the HY level and one basis point wider at the IG level. This is the configuration this section has described for three weeks arriving in the data: an index at a cycle tight concealing a tail that is repricing. Two sessions ago the tail had tightened twice in four days and this report cut the protection trade to a quarter on that evidence; one 16 bp session has undone both narrowings and more, which is an honest mark against that decision (Section 12). The cash market corroborates directly: HYG fell 0.89% while the 10-year moved only 4 bp, and the single-name widening Bloomberg documented on Monday in PG&E and Edison long bonds sat behind a legislative outcome that then reversed on Tuesday.

(b) Money-market and funding plumbing

New York Fed reference rates, published ~8:00 a.m. ET for the prior business day. The rates below carry the 31 August 2026 effective date. Reverse repo take-up is the same-day 1 September operation.

The turn was not benign after all, and it cleared in one day. Monday's edition read a $6.726bn reverse repo take-up and a SOFR level with IORB as cash finding the facility rather than bidding the market, and asked whether take-up would fall back under $1bn on 1 September. It did — $725m, a ninefold drain — but the reference rates published Wednesday morning for the 31 August effective date tell a different story about the turn itself: SOFR printed 3.68%, three basis points through the 3.65% IORB, its first print above the administered rate in the reporting window, on window-high volume of $3,056bn with the 99th percentile at 3.77%, four basis points wider than Friday's. Tri-party and broad general collateral both jumped 4 bp to 3.67%. Repo, not the unsecured market, took the strain: EFFR and OBFR were unchanged at 3.63% with EFFR volume falling to $105bn from $123bn. So the month-end squeeze was real, it was collateralised, and it lasted a day. Reserve balances have no new print — the H.4.1 series still reads $2.9249tn for the week ended 26 August, $68bn below the 5 August peak; the observation covering the turn publishes Thursday.

The bill strip declined to confirm the coupon curve for a second session, and this time the message is cleaner. The entire bill complex moved 0 to 2 bp — 1-month 0, 3-month +1, 6-month +1, 1.5-month +2, 2-month +1, 4-month +1 — while the 2-year cheapened 5 bp and the 3-year and 5-year 6 (Section 6, off-table tenors). Monday's version had the 6-month richening on a day the futures strip added eight points, which looked like month-end cash. Tuesday's version has the bills simply anchored: with the 16 September meeting two-thirds priced, a bill maturing before or barely after it has almost nothing left to reprice, so every incremental hawkish basis point has to go into the two- to five-year sector. That is the mechanical reason the curve flattened from the belly. On the week the bills are up 5 to 10 bp against the 2-year's 22, so the divergence is a week old and it is a duration story, not a financing one. Watch whether SOFR returns below IORB on the 1 September effective date, and the 9 September buyback operation.

(c) Rates volatility and swap spreads

The MOVE series updated again, to 75.32 on the 31 August stamp, with a day range of 70.97-75.32 and an open of 70.97 — internally consistent within its own day and reconciling exactly to the 70.97 this report published for 28 August, which makes the +4.35 point move on Monday publishable. What remains disqualifying is the stated "previous close" of 95.74, twenty points outside the series' own day range for a fifth consecutive session, so the 1 September change is still withheld and the level carries its vintage. That matters more than usual, because Bloomberg's Cameron Crise argued on Tuesday that "rates volatility is generally well-contained; that matters, because most of the significant downdrafts in stock prices over the past few years have coincided with spikes in rate vol" — and the series that would test the claim for the current session cannot be read. What can be stated: equity volatility rose 9.52% on a 0.71% index decline, a thirteen-to-one ratio, on a session in which the 2-year cheapened five basis points and Bloomberg reported traders paying millions in premium for Treasury options struck at a 5.7% thirty-year yield by late November, with 10-year structures targeting 4.85% and 5-year structures 4.6%. Options skew on long-bond futures now favours puts, so hedging a selloff costs more than hedging a rally, and JPMorgan's client survey shows neutral positioning at its lowest since September 2025 with both outright longs and shorts up two points. That is not a well-contained rates market by the positioning measure, whatever the index level says. The swap-spread basis is unchanged in substance: Treasuries have outperformed equivalent-maturity swaps since the buyback announcement, which Bloomberg cites as evidence the intervention did something even as the level round-tripped.

(d) Issuance, leveraged loans and private credit

Levels from the TradingEconomics currency board taken after the U.S. close. The vendor's date column reads Sep/01 for GBP, NZD, CHF, CAD, KRW and TWD and a live clock for DXY, EUR, AUD, JPY and CNY; every live row's %Chg was verified against the prior edition's levels for the same vendor and reconciles to within 0.03 percentage points, so the published changes are genuine 24-hour moves and are reproduced (Data Notes). USD/CNH is from Investing.com, whose session had rolled, so its move is computed against the prior edition. Quote basis: USD per unit for EUR, GBP, AUD and NZD; units per USD for JPY, CHF, CAD, KRW, TWD, CNY and CNH.

The take: the dollar converted, but only because everything else broke. Monday's edition wrote that the dollar had failed the easiest test it would get all year — a hawkish repricing plus an oil shock, and it fell anyway. Tuesday delivered a smaller hawkish repricing (1.5 pp on CME against 8.4) and a larger oil shock, and DXY rose 0.22% to 99.646, recovering the whole of Monday's loss. The mechanism is not American strength; it is that the alternatives all had worse days. The euro fell 0.19% on the day euro-area inflation printed 3.3% with a 14.3% energy component and an ECB hike moved to almost fully priced — a currency that cannot rally on its own central bank turning hawkish is being sold for the energy import bill, not for the rate. Sterling fell 0.23% while the gilt cheapened 16 bp at ten years and 10 bp at thirty: a bond market demanding more compensation and a currency refusing to reward it is the textbook fiscal-risk signature, and it is the single most informative cross on this board.

The Canadian dollar is the tell that oil no longer buys anything. USD/CAD rose 0.31% to 1.38959 on a session when WTI gained 6.04%. Twenty-four hours earlier the same pair fell 0.35% on a 3.42% crude move and this report called it the orthodox response. Two sessions, two opposite reactions to the same-signed shock, and the difference is the U.S. two-year: on Monday it did not move, on Tuesday it cheapened five basis points. When the front end of the U.S. curve is moving, the terms-of-trade channel loses to the rate channel in every G10 cross — which is also why the Swiss franc weakened 0.38% into a second consecutive Hormuz escalation and why gold and silver fell 2.37% and 3.51% (Section 11).

The Asian crosses inverted, and Korea is the important one. USD/KRW rose 0.58% to 1375.20, ending four consecutive sessions of won strength, on the day Korea reported total exports +68.7% year on year to $98.25bn and semiconductor exports +209%, an all-time record. A currency that weakens on a record export month is a currency whose bid was never about the trade balance — which corroborates rather than contradicts the BNP Paribas explanation this report carried on Monday, that the recent strength was corporate repatriation to fund capacity rather than portfolio inflow. Repatriation is lumpy, and it stopped. Meanwhile USD/TWD fell 0.14% to 31.6620 as the Taiex rose 1.78%, so the two Asian chip currencies moved in opposite directions on the same semiconductor news, and USD/CNH at 6.7220 barely moved on a private manufacturing PMI beat at 51.5.

Settlement basis, stated. All eight rows are quoted from the Investing.com per-contract historical series as read after the close, which is this report's settle series of record, so every day change below is like-for-like within one series. WTI, RBOB, heating oil and natural gas are on the October contract; gold, silver and copper on December; Brent on the front. The prior edition quoted WTI at $86.25, Brent at $90.68 and gold at $4,499.40 from vendor real-time boards rather than this series, which carries 31 August at 85.76, 88.37 and 4,481.50 — so three levels are not directly comparable to Monday's published numbers, and the reconciliation is in Data Notes. Weekly and year-to-date columns are TradingEconomics spot returns on the vendor's Sep/01 date stamp.

\*YTD figures marked with an asterisk are TradingEconomics spot year-to-date returns, not futures returns on the contracts quoted above. Weekly columns are TradingEconomics spot weekly changes on the same caveat. The daily settles, changes and percentage moves are on the futures basis named in each row. Mixing the two would be a basis error; they are presented in separate columns for exactly that reason.

The metals broke as a complex, and copper is what makes it diagnostic. Monday's version of this story was that gold and silver were trading the expected real policy rate rather than the geopolitical premium — a monetary explanation. Tuesday added the industrial leg: copper -2.25%, platinum -2.51% (TradingEconomics), gold -2.37% and silver -3.51%, with gold closing at $4,375.40 within five dollars of its low and the gold-silver ratio out to 67.69 from 66.90. A pure discount-rate story would leave copper alone. Copper falling with the precious complex on a day the barrel rose 6% says the market is pricing a higher policy path and weaker end demand at the same time — which is what an ISM at 54.6 with prices at 71.1 and employment at 51.2 describes, and what a JOLTS report with hiring down 278,000 confirms. Bitcoin agrees, -2% at $77,247.48 with ether -2.4% at $2,412.93 (Bloomberg). Gold is now up only 0.21% on the year on the TradingEconomics spot basis, having been up more than 3% at Monday's close — one week has taken the entire 2026 gain out of the debasement trade.

The crack spreads moved again, in the same direction, and harder. On a consistent October basis against October WTI:

The physical logic has not changed and is now being paid for in size: diesel, jet and marine gasoil transit the strait, reroute when it is mined and are burned by the rerouted tonnage, while gasoline is a domestic-demand product entering the post-Labor-Day roll-off with a winter-grade spec already in the October contract. Heating oil rose 7.06% against RBOB's 2.67%, and TradingEconomics marks heating oil +122.58% year to date against gasoline's +84.83% on its spot basis. The position this book opened on Monday morning at the differential's low is now its largest winner and the risk has inverted: three-quarters of a two-session move has happened, and EIA petroleum on Wednesday 10:30 is the first hard test.

Equity is marking the barrel unevenly for a second day, and the pattern reversed. Monday was a producer-and-services rally with the refiners left alone; Tuesday was the opposite. Marathon Petroleum +2.59% to $383.00, Phillips 66 +2.21% to $252.02 and Valero +0.86% all made 52-week highs, and XLE and VDE set all-time intraday highs with XOP at a multi-year high (CNBC) — while SLB fell 4.91% and Halliburton 0.14% after Monday's +4.83% and +1.85%. The market has worked out that a Hormuz disruption is a refining-margin event and a capital-expenditure risk at the same time. It took two sessions.

Desk-style ideas for institutional investors. Each carries an explicit expression, catalyst and invalidation. These are not personalized investment advice; verify independently and size to your own mandate before acting.

1. The rates trade — the calendar spread took the day back; hold the quarter and raise the stop

Mark first. Long ZQU6 (September 2026) against short ZQZ6 (December 2026), DV01-matched one-for-one at $41.67 per basis point per contract, entered ten sessions ago at 96.325 / 96.160 for a spread of 16.5 bp, trimmed to a quarter on Friday at 27.0 bp with the stop at 23.0. Tuesday's mark: ZQU6 96.290, ZQZ6 96.015 — a spread of 27.5 bp. That is +1.5 bp on the day, worth +$62.50 per contract pair on the quarter, and it leaves the trade +11.0 bp from entry, a new high.

Why it worked, and the mechanism is now also the risk. The spread needs December to absorb more of every repricing than September. Tuesday delivered exactly that: ZQU6 fell 0.5 bp while ZQZ6 fell 2.0 bp, because a soft ISM barely touched a meeting already two-thirds priced while the deep tail extended — December's +75 bp bucket rose to 13.5% from 10.3% and its cumulative-above to 91.0% from 88.9%. Note what the position now owns: it is short the far tail of the 2026 distribution at exactly the moment the 2027 strip cheapened 2.5 to 5.5 bp and one 2027 meeting's mode flipped to three hikes. The trade has made money on a policy repricing, lost money on a supply repricing, and made money again on a terminal-rate repricing — but the third is a headwind if it continues.

The modal path, the base case and the tails. Modal path: a 25 bp hike on 16 September is modal at 65.6% on Investing.com and 66.9% on CME, with Bloomberg reporting 17 bp of tightening priced and odds near 70%, and ease at 0.0%; a hike is modal at 28 October too, at 54.8% against 77.4% cumulative above; and two hikes by 9 December is the single most likely year-end state at 41.9% against 35.6% for one and 9.1% for none. The 2027 strip is modal 4.00-4.25% at seven of eight meetings and 4.25-4.50% at July 2027. Base case: two hikes into year-end and a terminal rate that has just begun drifting higher rather than staying anchored. Tail one, and it is Friday: payrolls at +55,000. A negative print — Interactive Brokers' Jose Torres notes "a second month in a row of declining payrolls would motivate heavy bond buying" — unwinds a two-thirds-priced hike, and at ten percentage points of probability per basis point of ZQ price this spread can move 3 bp in an hour. Tail two: ADP Wednesday at +47,000, the leading indicator of that. Practical implication: the position is a quarter with 11.0 bp of profit into two dated labour events, one of which is asymmetric against it. Hold the quarter, do not add, and raise the stop to 24.5 bp — banking the day's gain as protection rather than reaching for the last two basis points into a payroll.

Expression: long ZQU6 / short ZQZ6, DV01-matched one-for-one, quarter size. Catalyst: ADP 9/2 08:15; EIA petroleum 9/2 10:30; claims and ISM services 9/3; payrolls 9/4 08:30; the buyback operation 9 September; PPI 9/10; CPI 9/11; the 16 September FOMC. Invalidation: the spread through 24.5 bp; or the September cumulative hike printing below 45% on either vendor; or any 2026 meeting showing a non-zero cut probability. Sizing: a quarter, at $41.67 per basis point per pair.

2. Long the October distillate crack against short the October gasoline crack — take half the profit, run the rest

Mark: entered Monday at a differential of $56.02 ($100.59 distillate crack against $44.47 gasoline crack, prior-edition basis). On the settle series of record the two legs now stand at $107.39 and $41.75 for a differential of $65.64, and the like-for-like session move was +$9.63 on top of Monday's +$5.77. The honest reading: this is the best two-day entry in the book's history and it happened for exactly the stated reason — heating oil +7.06% against RBOB +2.67% on two supertankers struck exiting Hormuz. It is also now a chase in reverse: the whole thesis has been paid in two sessions and the next catalyst is a data print rather than a headline. Action: take half off, run a quarter. Catalyst: EIA weekly petroleum inventories Wednesday 10:30, where a distillate draw extends the move and a build is the first real test; EIA natural gas storage Thursday as a heating-fuel substitution cross-check; any Hormuz de-escalation or claim of responsibility. Invalidation: the differential back through $56, which is the entry level and a full round trip; or a credible de-escalation headline, since this is a geopolitical supply trade that will be invalidated by diplomacy rather than by data. Sizing: a quarter after the trim.

3. Long the 20-year against the 30-year, on the November refunding — hold the half

Expression: long the 20-year bond against short the 30-year, DV01-neutral, half size. Mark: 20s30s at 0 bp, the 30-year at 5.27% and the 20-year at 5.27% — 1 bp flatter on the day and 1 bp flatter on the week, with the 20-year cheapening 3 bp against the 30-year's 2. A fifth session of essentially no spread. What changed: the environment, not the spread. Bloomberg reports the 30-year back at exactly the level seen moments before the 19 August buyback announcement, with the yield having settled above 5% on 55 days this year, the most since 2006, and averaging 4.96% year to date, the highest annual level since 2004. Options desks are paying for 5.7% by late November. Against that sits a $215bn September corporate calendar and a buyback programme beginning 9 September at an unspecified size. The dislocation this trade waits for needs the calendar to arrive; it has not yet. Catalyst: the 9 September buyback operation and its maturity buckets; Treasury's 4 November quarterly refunding. Invalidation, unchanged: 20s30s through -3 bp, or an explicit Treasury statement ruling out changes to long-end auction sizes. Sizing: a half, unchanged.

4. Protection on the CCC cohort funded in IG — the cut was wrong; go back to a half

Expression: long CCC-exposed credit protection (or short a levered-loan / CCC-heavy vehicle) against long IG cash. Mark: CCC 1,042 bp, +16 bp; HY 263 bp, +3 bp; IG 80 bp, +1 bp on the 31 August FRED update, taking the CCC-minus-HY differential to 779 bp from 766 — a 13 bp gain in a single session. The honest reading, and it is a mark against this desk: on Monday this book cut the position from a half to a quarter on the grounds that the tail had tightened twice in four sessions and the trade was "paying to wait." One session later the tail widened 16 bp, the largest single-day move in the series in the reporting window, and the differential is back through its prior extreme. The cut cost roughly half of a 13 bp gain. The reason to add back rather than merely hold is that the mechanism reasserted itself with corroboration this time: HYG fell 0.89% against a 4 bp move in the 10-year, so about 0.75% of that loss is credit and not duration, and the same-day cash market agrees with the lagged index. Action: back to a half. Catalyst: ADP 9/2 and payrolls 9/4; the September IG calendar clearing from Wednesday into a $215bn month; Broadcom, HPE and NetApp 9/2. Invalidation: the differential back through 750 bp, or IG widening beyond 90 bp, which would mean the aggregate has joined the tail and the pair no longer isolates anything. Sizing: a half.

5. Short the debasement complex against long the dollar — the funding leg finally worked; take a third off

Expression: short an equal-weighted basket of Comex gold, Comex silver and a bitcoin proxy against long the dollar index, dollar-notional matched. Mark: the short basket returned gold -2.37%, silver -3.51%, bitcoin -2.00% for an average of -2.63%, against DXY +0.22% — a 2.85-point gain on the quarter, and the first session in which both legs worked. The honest reading: Monday's version had the thesis leg working and the funding leg failing, and this report flagged that expressing the short unfunded might be better. That change would have cost 22 basis points of the gain. The dollar converted because the alternatives broke — sterling fell on a 16 bp gilt selloff, the euro on a 3.3% inflation print — rather than because the Fed repriced, so the funding leg is working for a reason the thesis does not own. Action: take a third off, run the rest, because gold at $4,375.40 is now up only 0.21% on the year and the easy part of the move is behind it. Catalyst: ADP 9/2, payrolls 9/4, CPI 9/11; Chinese physical demand on the Shanghai open. Invalidation: gold reclaiming $4,499, Monday's prior-edition settle; or DXY back below 99.11, last Thursday's close. Sizing: two-thirds of a quarter after the trim.

6. Short the AI-halo basket against long the name that actually reported — close it, at a profit, into the hardware prints

Expression: short an equal-weighted basket of Synopsys, Palo Alto Networks, ServiceNow, Fortinet and Adobe against long Nvidia, beta-adjusted, quarter size. Mark: the basket returned -5.63%, -5.25%, -1.13%, -5.31%, -2.29% for an average of -3.92%, against Nvidia -1.51% — a 2.41-point gain, the largest single-session gain this pair has produced. The honest reading: the thesis was that names rallying on somebody else's numbers give it back on their own, and Tuesday delivered it in one session — Palo Alto fell 5.25% into its print, then beat on revenue, beat on EPS, guided fiscal 2027 above consensus and fell a further 2% after hours. Cumulatively the basket now underperforms Nvidia by 0.51 pp against the 6 pp threshold, so the position has recovered its entire drawdown and is marginally profitable. Action: close it. The catalyst it was constructed around has passed, the largest short-basket name has reported, and holding through Broadcom, HPE and NetApp on Wednesday evening would be a different trade with a different thesis — three hardware balance sheets against a semiconductor complex down 2.14%, which is a long-vol expression rather than a pairs one. Closed at +2.41 pp on the session, +0.51 pp cumulative.

Prior closes, marked forward. The gasoline-versus-distillate crack pair, closed a week ago, would have lost a further $9.63 on Tuesday — $15.40 across two sessions — which makes that exit the most valuable single decision in this book. The long-silver-against-short-gold pair, closed on thesis invalidation, would have lost a further 1.14 points: silver -3.51% against gold -2.37%. The on-balance-sheet-versus-off-balance-sheet AI funding pair would have gained again, with Oracle -5.23% the largest loss in the financing complex. Three of the last four exits have been vindicated within two sessions; the CCC cut in idea 4 is the one that was not, and it is marked as such above.

The vol note. VIX closed 16.34, up 9.52%, having traded 16.80, on a session the index lost 0.71% — a thirteen-to-one ratio of volatility gain to index loss, after Monday's four-to-one and a week of roughly one-to-one. This is the first 16 handle of the reporting window and the first session in which equity volatility led rather than followed. The calendar it is now paying for: ADP Wednesday 08:15, EIA petroleum Wednesday 10:30, claims and ISM services Thursday, payrolls Friday 08:30, then Labor Day, then PPI on the 10th and CPI on the 11th, into a 16 September meeting priced at roughly two-thirds. A 16.34 handle asks for about a 1.02% daily move, against a market that delivered 0.33% and 0.71% in the last two sessions — so the gamma bought last week is marked at a fair price rather than a cheap one. Take profit on half the index gamma dated 2 through 4 September and roll the balance into the 10-11 September inflation block, where the surface has not yet repriced and where the barrel actually lands in the data. Rates volatility remains unmarkable for the current session — the MOVE series is one day stale at 75.32 — but Bloomberg's positioning data says the hedging is happening in Treasury options rather than in the equity index, with skew favouring puts on long-bond futures.

Crowded consensuses to stress-test, with the numbers.

The two-sided geopolitical tape. Escalation: two supertankers exiting Hormuz were struck by projectiles late Monday, the Sidr and the Senegal Prosperity, with no claim of responsibility; U.S. Central Command confirmed strikes on Islamic Revolutionary Guard Corps targets; Bloomberg reports Hormuz flows had recovered to only half of pre-war levels before this week and that current and former officials expect the conflict to run for months; Canada's retaliation on $20bn of U.S. goods lands 8 September. De-escalation, or at least containment: there were no reports of strikes on Tuesday; Qatar's foreign ministry says mediation is ongoing and Pakistan's prime minister still backs the Islamabad framework; and Bessent publicly downplayed the strait as something pipelines would make "irrelevant," which is an administration signalling it does not intend to escalate. Domestically the fastest-moving policy risk stayed sub-federal in both directions: California's Assembly adjourned without a vote, handing back 8.93% and 5.92% of what it took on Monday, with a special session possible and autumn hearings promised.

Structural watch items. SOFR printed 3 bp through IORB at 3.68% on the month-end effective date with a 3.77% 99th percentile and window-high $3,056bn volume, before reverse repo take-up drained to $725m; reserve balances at $2.9249tn with the print covering the turn still to come; $215bn of expected September corporate supply into a 4.79% ten-year and a 30-year that has round-tripped a buyback announcement; Japan's 10-year at 3.001%, its first 3% print since 1996, and the 2-year at a 31-year high of 1.81%, with the Bank of Japan deciding 18 September and Bessent saying he has "information that the market doesn't have"; euro-area inflation at 3.3% with a 14.3% energy component and an ECB hike near-fully priced; Nike at a 52-week low and its lowest in more than twenty years; and construction spending at its weakest since October 2023 with mortgage rates at a one-year high.

Full source links and the complete Data Notes & Conflicts appendix are in the companion file US_CrossAsset_Daily_2026-09-01_DataNotes.txt, alongside the canonical Markdown report of record.