Closing Briefing — Monday, August 31, 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 cleanest of the reporting window. Compounding each group's 28 August YTD by Monday's one-day move reproduces the published YTD to within 0.02 percentage points for all eleven groups, with a maximum deviation of 0.020 pp in consumer cyclical. Worked examples: energy 1.3589 × 1.0184 = 1.3839 → +38.39% against a published +38.39%, deviation 0.000 pp; technology 1.2381 × 1.0022 = 1.2408 → +24.08% against +24.09%, deviation 0.008 pp. The two-session real-estate drift this report flagged on Thursday and Friday — 0.195 pp then 0.200 pp — has closed to 0.002 pp, which confirms the diagnosis that it was a constituent or dividend adjustment inside the bucket working through rather than a data error.

The utilities number is the trap, and it is worth a paragraph. Finviz has utilities down only 0.91% on a day the two California names fell 23.07% and 20.06%, because the Finviz bucket is broad and market-cap weighted across a large regulated universe that mostly did not move; CNBC's GICS tally had the S&P 500 utilities sector down about 1.6% intraday, and its month-to-date loss at 5.6%. What the group did produce is the highest relative volume on the board at 1.67, against 1.20 for consumer defensive and 0.79 for technology. When a sector trades at nearly seventy percent above its average volume for a sub-one-percent move, the distribution inside it is what matters, not the mean.

Rotation read: energy at +1.84% was the only group with a real bid, and technology at +0.22% the only other positive, which is the narrowest positive breadth of the reporting window. The composition trap runs the other way from Friday: Alphabet (-2.13% on the A line) and Meta (-0.98%) sit in communication services, which is why the worst group on the board is one a GICS reader files partly under technology, and Amazon's -2.50% is what pushed consumer cyclical down 0.75% on a day the barrel rose 3.42%. On the year the shape is unchanged and extreme: energy +38.39% and technology +24.09% against consumer cyclical -3.48%, communication services -1.98% and utilities -1.86% — the same three negative groups as Friday, with utilities now nearly a point deeper.

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.

Tue 9/1. BMO: Medtronic (MDT). AMC: Palo Alto Networks (PANW), Dell Technologies (DELL).

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 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 (8/28 report):

Source: U.S. Department of the Treasury Daily Treasury Par Yield Curve Rates, Text View for August 2026, read after publication. Changes are versus the 28 August official row (1-day) and the 24 August official row (1-week).

Off-table bills, extracted and reported here because they carry the financing story. 1.5 Mo 3.86% (+3 bp on the day, +8 bp on the week), 2 Mo 3.88% (+2 bp, +8 bp), 4 Mo 3.96% (+2 bp, +6 bp), 6 Mo 3.99% (-3 bp, +3 bp). The 6-month is the only tenor on the entire curve that richened, and it is the shortest one that fully spans the 16 September meeting — on the day the futures strip added eight points of hike probability. See Section 9 block b.

The read: a long-end-led bear steepener, and it is the mirror image of Friday. Walk the curve outward and the sign changes twice: 6-month -3, 1-year +1, 2-year 0, 3-year -1, 5-year +1, 7-year +3, 10-year +2, 20-year +3, 30-year +3. Friday's monotonic decay from +14 at the two-year to +3 at the thirty was policy repricing; Monday's flat-to-negative front and a uniformly cheaper back half is inflation risk premium, and its cause is on the tape — WSJ headlined the session "Long-Term Treasury Yields Rise as Iran War Revives Inflation Worries." 2s30s widened 3 bp to 91 and 2s10s 2 bp to 41, recovering roughly a quarter of Friday's 11 bp and 8 bp flattening in a single session.

The diagnostic that matters is the contradiction inside it. CME's September hike probability rose to 65.4% from 57.0% on its own 1-day column, and Bloomberg reported swaps at "an almost 70% chance" — yet the 2-year did not move and the 6-month richened 3 bp. Those two facts cannot both be about the policy path. The resolution is that Friday's 6-month print of 4.02% was an 8 bp spike off Thursday's 3.94%, the largest one-day move at that tenor in the reporting window, and Monday retraced three of it while the coupon curve went the other way. The bill and the futures strip disagreed, and when they disagree the bill is usually the one carrying month-end cash rather than a policy view — overnight reverse repo take-up jumped to $6.726bn on the 31st (Section 9 block b). Take the futures strip as the policy signal and the 6-month as a calendar artefact, but note that this is the second consecutive session in which the front of the bill curve has failed to confirm the money-market strip.

The week now has one shape and the long end has taken it back. On five sessions the 1-year is 12 bp cheaper at 4.16%, the 2-year 10 bp, the 3-year 9 bp and the 5-year 8 bp, while the 10-year is 5 bp cheaper, the 20-year 3 bp and the 30-year 2 bp. 2s30s has still flattened 8 bp on the week and 2s10s 5 bp — but on Friday those numbers were 15 bp and 11 bp. One session of oil has undone a quarter of a week of Jackson Hole. The relevant level, per Bloomberg, is that the 10-year at 4.75% is the highest since January 2025, and its live board marks the same 4.75% with a +3 bp change against the official par +2 bp — a zero basis-point level gap and a one basis-point change gap, the tightest agreement of the reporting window and a straightforward artefact of a 4:59 p.m. live mark against a 3:30 p.m. bid-side construct.

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 2.0 percentage-point gap, and a prior-session correction that runs the other way from the last one.

CME FedWatch headline — 16 September 2026 meeting. Data as of 31 Aug 2026, 04:57:43 p.m. CT (5:57 p.m. ET), read from the FedWatch probability table.

Provenance of every column, stated — and Friday's number corrected. CME's NOW column carries a 04:57:43 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 — a near-settlement snapshot on the same basis as Friday's. 1 DAY carries the legend date 28 August and prints 57.0% for the hike, against the 59.7% this report published from CME's own live column on Friday evening. Friday's true close was therefore about 2.7 pp lower than published; the correction is recorded, and it does not change any conclusion drawn on Friday — the direction, the magnitude and the modal shift all stand. Note the pattern, because it is now twice in a row: the live CME read after the CT close has overstated the settled figure by 1.5 pp and then 2.7 pp. Treat the NOW column as indicative to roughly three percentage points. 1 WEEK (24 August) and 1 MONTH (31 July) carry genuine reference dates and are used in calculations. The Investing.com matrix below is timestamped 31 Aug 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 65.4% at 5:57 p.m. ET; Investing.com at 63.4% at 5:45 p.m. ET — a 2.0 percentage-point difference across twelve minutes, against 3.8 pp on Friday. The gap is large in probability and negligible in price. Investing.com publishes the September future at 96.295, down 1.0 bp from Friday's 96.305; because the 16 September meeting sits mid-month, only about 47% of the contract's averaging period is affected by the decision, so one basis point of ZQ price is worth roughly ten percentage points of hike probability. The entire 2.0 pp vendor gap is therefore two-tenths of a basis point of contract price, and the whole 7.5 pp move on Investing.com's card is three-quarters of a basis point. Bloomberg's swaps read of "an almost 70% chance" sits above both vendors and is a third construction; the three-way spread of roughly 6.6 pp is two-thirds of 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 8.4 pp on CME's own columns (57.0% to 65.4%) and 7.5 pp on Investing.com's (55.9% to 63.4%) — a third of Friday's move, on a day with no Fed communication and no U.S. data above Low sensitivity. What moved it was the barrel. On a one-week view it is 65.4% against 41.4% on CME and 63.4% against 40.4% on Investing.com, so 24.0 pp and 23.0 pp in five sessions, essentially all of it in two. The one-month column is now the discipline: on 31 July CME priced 67.0% at 375-400 with nothing above, so Monday's 65.4% is still 1.6 pp below where the market sat a month ago. This is a completed round trip, not a new regime — a point worth holding when the desk is being asked to chase. Further out, every horizon extended: October's cumulative-above went to 73.5% from 70.1%, December's to 88.9% from 88.4%, January 2027's to 91.7% from 91.8%, and the peak of the strip sits at 96.0% across June and July 2027. The probability of a cut at any 2026 meeting remains 0.0%.

(a) Current-year meeting distributions

Investing.com Fed Rate Monitor, updated 31 Aug 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 has fallen 23.0 points in a week, from 59.6% to 36.6%, and 7.5 in a session. Second, December's modal outcome is now two hikes for a second consecutive session, at 39.8% for 4.00-4.25% against 39.0% for one, and the gap has widened from 0.4 points to 0.8. Third, and least remarked, the tail thinned while the body thickened: December's +75 bp bucket fell from 11.0% to 10.1% and October's +50 bucket from 18.0% to 17.5%, even as both meetings' cumulative-above rose. The market added confidence to the first two hikes and took a little away from the third — which is what an inflation-supply shock does to a reaction function that is already committed at the front.

(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 barely moved and that is the finding. Every meeting from January to December 2027 stays modal 4.00-4.25%, exactly as on Friday, and the contract prices are unchanged to a tenth of a basis point at seven of eight points — 95.985, 95.885, 95.835, 95.780, 95.770, 95.780, 95.775, 95.810 against Friday's 95.985, 95.880, 95.830, 95.780, 95.770, 95.770, 95.775, 95.810. The trough is still 95.770 in July 2027. So an eight-point rise in the September hike probability and a three-percent move in crude produced nothing at all beyond eighteen months. The strip has decided the oil shock pulls tightening forward without changing the destination — the definition of a supply shock priced as transitory in level and permanent in timing. The first non-trivial cut probability, 0.8% at 3.25-3.50%, still appears only at December 2027, unchanged from Friday.

(c) Year-end probability ladders

Year-end 2026 — the 9 December meeting.

Cumulative above the current range: 88.9%. Cumulative below: 0.0%. Sum: 100.1%.

Year-end 2027 — the 8 December meeting.

Cumulative above the current range: 92.7%. Cumulative below: 0.8%. Sum: 100.0%.

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 and March 2027 rows sum to 100.1% and the June, July and September 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 28 August 2026, not the 31 August close. Same-day direction is cross-checked against the cash-market proxies underneath and against Bloomberg's single-name credit coverage.

CDX — the full six-step ladder was worked again and is reported so the gap stays auditable. (1) Bloomberg in Chrome: /markets, /markets/fixed-income and /markets/rates-bonds were all rendered; the Fixed Income Indices table and the global government-yield boards populate, and no CDX line appears on any of them. (2) WSJ Market Data bonds page: rendered, with the credit tables loading Treasurys, consumer rates and government bonds but no CDX row; a full-text scan of the page returned zero occurrences of the string. (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 and methodology documentation, not the daily running spread. (4) FT Markets Data and Reuters credit wraps returned no dated CDX quote for 31 August. (5) TradingView and Barchart symbol searches for CDX resolve to an unrelated ETF. (6) Cash-market proxies, labelled as proxies: HYG closed $79.81, +0.09%, and LQD $106.21, -0.13%, against 28 August closes of $79.74 and $106.35. The HY vehicle rose and the IG vehicle fell on a day the 10-year cheapened 2 bp — duration taking the loss and credit risk being bought, which is the same message the FRED series carries through 28 August. No CDX level is published here, because an undated third-party digest number is not a CDX level.

The interesting credit story this session was not in the index, it was in one sector. IG at 79 bp is unchanged on the day and exactly where 2026 started. HY tightened 3 bp to 260 bp, its tightest of the year and 21 bp through the January level. CCC tightened 5 bp to 1,026 bp, which narrowed the CCC-minus-HY differential to 766 bp from 768 — the second narrowing in four sessions, and a small step back from the record. So the aggregate market did nothing and the tail improved slightly. Against that, Bloomberg reported single-name widening of a size the index cannot see: PG&E's 6.3% 2056 bonds widened 15 bp to 138 bp and Edison International's 4% 2047 notes 10 bp to 133 bp on Trace, both by 8:43 a.m. New York time, on a legislative outcome Impax's Tony Trzcinka called a "material credit negative" with rating-agency downgrades likely. A regulated utility's long bond widening fifteen basis points in a morning while the IG index does not move is the clearest available demonstration of what an aggregate at a cycle tight actually conceals.

(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 28 August 2026 effective date. Reverse repo take-up is the same-day 31 August operation.

The turn arrived, and it arrived in the facility rather than in the rate. Overnight reverse repo take-up was $6.726bn on 31 August, against $175m on the 28th, $456m on the 27th, $702m on the 26th and $380m on the 24th — a thirty-eight-fold jump on the month-end date and the largest print of the reporting window by more than an order of magnitude. Read it with the rate: SOFR rose a basis point to 3.65% on the 28 August effective date and is now level with the 3.65% IORB for the first time in the window, having sat a basis point below it all month, with the 99th percentile up a basis point to 3.73% and volume $28bn lighter at $2,808bn. Repo followed a tick — TGCR and BGCR both to 3.63% — while EFFR and OBFR were unchanged at 3.63% and EFFR volume rose to $123bn from $111bn. That combination is a benign turn: cash looking for a home found the facility rather than bidding the market, which is the opposite of a reserve-scarcity turn in which the facility empties and SOFR prints through IORB. 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, and the next observation covers the week containing the turn itself.

The bill strip disagreed with the futures strip, and the fulcrum is again the six-month. The 6-month bill richened 3 bp to 3.99% — the only tenor on the entire par curve to fall — while the 1-month rose 1 bp to 3.85%, the 1.5-month 3 bp to 3.86%, the 2-month 2 bp to 3.88%, the 3-month 1 bp to 3.91% and the 4-month 2 bp to 3.96% (Section 6, off-table tenors). The six-month is the shortest tenor that fully spans the 16 September meeting and it richened on the day CME added 8.4 points to that meeting's hike probability. On the week the ordering is intact — 1.5-month +8 bp, 2-month +8 bp, 1-month +6 bp, 4-month +6 bp, 3-month +4 bp, 6-month +3 bp — so the weekly story is still a general cheapening of the front with the six-month lagging. The honest reading is that Friday's 8 bp spike at the six-month overshot and Monday took three back into a month-end cash glut, not that the bill market has a different Fed view. Watch whether the facility drains back under $1bn on 1 September and whether SOFR goes back below IORB, with the 9 September buyback operation the next scheduled event.

(c) Rates volatility and swap spreads

The MOVE series did update, for the first time in four sessions, and it moved the right way. Investing.com now marks 70.97 with a 28/08 timestamp, a day range of 69.86-70.97 and an open of 69.86 — internally consistent within the day and reconciling exactly to the 69.86 this report published for 27 August, which makes the +1.11 point move on Friday publishable. What remains disqualifying is the stated "previous close" of 95.74, twenty-five points outside the series' own day range for a fourth consecutive session, so the 31 August change is still withheld and the level carries its vintage. The cross-market fact that can be stated without the vendor is the one that matters: equity volatility rose 3.40% on a session in which the S&P fell 0.33%, a four-to-one ratio, while the two-year did not move at all. Rates volatility should have been the calmer of the two on that curve, and equity volatility is the one that bid — which reverses the decoupling this report flagged on Friday and puts the equity market on the correct side of the calendar for the first time in a week. The swap-spread basis is unchanged in substance: Treasuries have outperformed equivalent-maturity swaps since the buyback announcement, and Fed researchers put hedge-fund swap-spread positions at a record $305bn last year against under $50bn in 2022. Jefferies' David Zervos put the structural case for that basis plainly: "the Treasury has become the dominant policy actor in shaping financial conditions," with buybacks and potential curve-twisting operations providing "a powerful market backstop."

(d) Issuance, leveraged loans and private credit

Levels from the TradingEconomics currency board taken after the U.S. close, with the vendor's own date column reading Aug/31, so the published %Chg measures the completed Monday session. Investing.com instrument pages are used for USD/CNH and USD/TWD, which TradingEconomics does not carry on the majors board; both had rolled into the Tuesday Asian session by the time of the pull, so their %Chg columns are not reproduced and the 24-hour move is computed against the prior edition's levels for the same vendor (Data Notes). 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 failed the easiest test it will get all year. Monday handed the dollar a hawkish repricing — CME's September hike probability up 8.4 points to 65.4% — and a 3.42% move in crude, which is a terms-of-trade gain for a net energy exporter and an inflation input for its central bank. DXY fell 0.29% to 99.415 anyway, surrendering more than half of Friday's 0.52% gain, and Bloomberg's own coverage framed the month as dollar weakness on the Treasury buyback programme rather than strength on the Fed. Twenty-four hours ago this report wrote that a dollar rally underwritten by a communication style "survives exactly as long as the next data point allows." It did not last one session, and there was no data point. The currency that did rally is the one with the oil: USD/CAD fell 0.35% to 1.38550, the best G10 performer, which is the orthodox response and makes the dollar's own failure more conspicuous, not less.

The Swiss franc is the tell. USD/CHF rose only 0.05% to 0.80825 — meaning the franc weakened fractionally — on a day the United States struck targets inside the Strait of Hormuz, Iran attacked the United Arab Emirates and Jordan, and a supertanker was reported mined. Put that beside gold -0.67% and silver -0.78% (Section 11) and the message is unambiguous: three independent haven assets were offered into a live military escalation. The market is not pricing this war as a risk event any more. It is pricing it as a supply event, and supply events are bought in crude and paid for in duration, which is precisely what the curve did (Section 6).

The Asian crosses inverted for a fourth consecutive session, and the mechanism has now been named. USD/KRW fell 0.60% to 1368.49, its largest single-session won gain of the reporting window, on the day the Kospi opened down 3.5% and closed up 0.46%. Bloomberg reported BNP Paribas's explanation: Korean chip firms are bringing capital in from overseas to fund production expansion, so the bid is corporate repatriation rather than portfolio inflow — which is why it survives an equity gap down. USD/CNH at 6.7188 is a 0.19% stronger offshore yuan against Friday's 6.7313, reclaiming Thursday's extreme, and USD/CNY fell 0.18% to 6.71791 on a day China's manufacturing PMI beat at 49.8. Jefferies' note that a 6% annual yuan appreciation is now large enough to move Apple's iPhone pricing model by five to eight percentage points (Section 4) is the clearest sign yet that this is a macro variable rather than a rates footnote.

Settlement basis, stated, and it needs care this session because four contracts rolled. WTI, RBOB, heating oil and natural gas are quoted on the October contract; copper and silver on December; gold on December. The prior edition quoted September RBOB, September heating oil, September natural gas, September copper and front-month silver, so five of eight rows are not comparable to Friday's published levels. Day changes below are computed strictly within the Investing.com per-contract historical series as read on 31 August, so every change is like-for-like even where the level is not comparable to the prior edition. Brent is quoted on the TradingEconomics front-month basis because the Investing.com per-contract page rolled to December. Weekly and year-to-date columns are TradingEconomics spot returns on the vendor's Aug/31 date stamp. Reconciliations are in Data Notes.

\*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 did not do their job, and it was not close. A live military escalation inside the world's most important oil chokepoint, an unconfirmed report of a mined supertanker, a Fed that has just been repriced eight points more hawkish — and gold fell 0.67% to $4,499.40, silver 0.78% to $67.255, with the gold-silver ratio drifting out to 66.90 from 66.83. Bitcoin was roughly the same story in the other direction, little changed at $78,856.35 with ether at $2,473.98 (Bloomberg), so the whole non-yielding complex was inert on a day it should have been the trade. The mechanism is the one Friday established and Monday extended: the expected real policy rate is now the dominant term in the discount factor for every asset that does not pay a coupon, and an oil shock raises it because the Fed's stated standard is underlying inflation. Note what this does to the geopolitical premium argument — Bloomberg's own explainer this session was about gold bulls turning to exotic options and spreads in an "orderly" rally, which is a description of positioning being managed rather than added to.

The crack spreads inverted, and the size of the move is the story. On a consistent October basis against October WTI:

That is a clean regime change and it has a physical cause: middle distillate — diesel, jet, marine gasoil — is the product most exposed to Strait of Hormuz transit and to bunker demand on rerouted tonnage, while gasoline is a domestic-demand product heading into the post-Labor-Day driving-season roll-off. This report closed the long-gasoline-crack-against-distillate position on Friday on the grounds that its second derivative had turned; that exit avoided $5.77 of loss in the first session after it (Section 12). Add Bloomberg's report that refiners have been granted the highest level of biofuel waivers since 2017, which is a direct margin transfer into the same complex, and the refining trade has two independent tailwinds and one seasonal headwind.

Equity is still not marking any of it evenly. Finviz energy closed +1.84%, the only group with a real bid, on SLB +4.83%, Exxon +2.71%, Devon +2.45%, Kinder Morgan +2.17% and Chevron +2.12% — a producer-and-services rally, not a refiner rally, on the day the refining margin made its largest move of the month. Meanwhile the cost side was marked immediately and hard: Royal Caribbean -3.82%, Uber -4.02%, Carnival -3.51%, Airbnb -3.27%, Booking -3.19%, Southwest -2.81%. The market bought the barrel and sold everyone who burns it, and left the people who refine it alone.

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 gave back a basis point; hold the quarter, the stop is still four away

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 nine 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 raised to 23.0. Monday's mark: ZQU6 96.295, ZQZ6 96.035 — a spread of 26.0 bp. That is -1.0 bp on the day, worth -$41.67 per contract pair on the quarter, and it leaves the trade +9.5 bp from entry.

Why it gave back, and the mechanism is the reason to keep holding. The spread needs December to absorb more of every repricing than September. Monday delivered the opposite in miniature: ZQU6 fell 1.0 bp while ZQZ6 was unchanged at 96.035, because the oil shock is being priced as a pull-forward rather than an extension. Look at the distributions and it is unmistakable — September's cumulative-above rose 7.5 points to 63.4% while December's rose only 0.5 points to 88.9%, and December's +75 bp bucket actually fell, from 11.0% to 10.1%. A supply shock that raises the odds of the first hike without raising the odds of the third is precisely the configuration that compresses a September-versus-December spread. The trade has now made money on a policy repricing and lost money on a supply repricing, which is a clean characterisation of what it owns.

The modal path, the base case and the tails. Modal path: a 25 bp hike on 16 September is modal at 63.4% on Investing.com and 65.4% on CME, with Bloomberg reporting swaps near 70% and ease at 0.0%; a hike is modal at 28 October too, at 56.0% against 73.5% cumulative above; and two hikes by 9 December remains the single most likely year-end state at 39.8% against 39.0% for one and 11.2% for none. The 2027 strip is modal 4.00-4.25% at every meeting and did not move at all, troughing at 95.770 in July 2027. Base case: the market prices two hikes into year-end and nothing incremental thereafter — tightening pulled forward, terminal rate unchanged. Tail one, and it is Tuesday: ISM manufacturing at a 55.3 consensus against Chicago's 47.1 and Dallas's 11.6. A sub-50 print unwinds a 65.4% September hike and the spread compresses violently — at ten percentage points of probability per basis point of ZQ price, this can move 3 bp in an hour. Tail two: payrolls Friday at +55,000; a hot print takes September above 80% and compresses the spread from the front instead. Practical implication: the position is a quarter with 9.5 bp of profit and a 3.0 bp stop distance, into two dated events that both compress it. Hold the quarter, do not add, and leave the stop at 23.0 bp — the asymmetry no longer justifies size, but the carry to 16 September does justify presence.

Expression: long ZQU6 / short ZQZ6, DV01-matched one-for-one, quarter size. Catalyst: ISM manufacturing and JOLTS 9/1 10:00; ADP 9/2 08:15; claims and ISM services 9/3; payrolls 9/4 08:30; the buyback operation 9 September; CPI 11 September; the 16 September FOMC. Invalidation: the spread through 23.0 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. New — long the October distillate crack against short the October gasoline crack, the reverse of the position closed on Friday

Expression: long the heating-oil crack against short the RBOB crack, both on the standard 42-gallon basis against October WTI, barrel-for-barrel, half size. Thesis: this is a deliberate reversal and the reversal is the point. On Friday this book took off the last third of a long-gasoline-crack position because its daily gain had decayed from $3.34 to $2.33 to $1.01 and Bloomberg had just reported funds adding bullish gasoline bets at the fastest pace in six months. In the first session after that exit the differential moved $5.77 the other way — distillate crack +$5.53 to $100.59, gasoline crack -$0.24 to $44.47. The mechanism is physical rather than positional: middle distillate is the Hormuz barrel. Diesel, jet and marine gasoil are what transits the strait, what reroutes when it is mined, and what rerouted tonnage burns; gasoline is a domestic-demand product entering the post-Labor-Day roll-off with a winter-grade spec change already in the October contract. Two independent tailwinds sit underneath the long leg: the U.S. has just granted refiners the highest level of biofuel waivers since 2017 (Bloomberg), and heating oil rose 4.70% against RBOB's 2.04% on the day. Catalyst: EIA weekly petroleum inventories Wednesday 10:30, where a distillate draw confirms; any Hormuz transit headline; EIA natural gas storage Thursday as a heating-fuel substitution cross-check. Invalidation: the differential back through $95 on the distillate crack, or a credible de-escalation headline out of Hormuz — this is a geopolitical supply trade and it will be invalidated by diplomacy, not by data. Sizing: a half, and note the obvious risk: three-quarters of the move happened in one session, so the entry is chasing.

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 1 bp, unchanged on the day — the 30-year at 5.25% and the 20-year at 5.24%, both 3 bp higher — and 1 bp flatter on the week. Flat again, for a fourth session. What changed: nothing in the spread, but the environment moved in the position's favour. An oil-led bear steepener that cheapens the 20-year and the 30-year by exactly the same three basis points, on a day the two-year does not move, says the long end is being repriced as a block rather than by maturity — which is the condition under which a supply-driven dislocation between the two points can appear when the calendar arrives. Citi has pushed its forecast for larger auctions to 2028 and raised the tail risk that Treasury eliminates the 20-year. 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 — cut back to a quarter; two narrowings in four sessions is a signal

Expression: long CCC-exposed credit protection (or short a levered-loan / CCC-heavy vehicle) against long IG cash. Mark: CCC 1,026 bp, -5 bp; HY 260 bp, -3 bp; IG 79 bp, unchanged on the 28 August FRED update, taking the CCC-minus-HY differential to 766 bp from 768 — a 2 bp loss and the second narrowing in four sessions. The reason to cut rather than hold: the thesis was that the tail dissents while the aggregate tightens. The aggregate is still tightening — HY at 260 bp is the tightest of the year — but the tail has now tightened alongside it twice in four sessions, and CCC at 1,026 bp is 11 bp inside its weekly level. One counter-signal is noise; two in four sessions with the differential 2 bp off its record is a position paying to wait. Set against that, the case for keeping some got stronger in an unexpected place: PG&E's 2056 bonds widened 15 bp and Edison's 2047 notes 10 bp in a single morning on a legislative vote, with downgrades flagged as likely, and none of it registered in an index at 79 bp. That is the mechanism this trade owns — idiosyncratic repricing invisible in the aggregate — arriving in investment grade rather than in CCC. Action: back to a quarter. Catalyst: ISM manufacturing 9/1; the September IG calendar clearing from Tuesday; the California utilities vote on Tuesday; Broadcom 9/2. Invalidation: the differential back through 750 bp, or IG widening beyond 85 bp. Sizing: a quarter.

5. Short the debasement complex against long the dollar — hold the quarter, but the dollar leg has failed twice

Expression: short an equal-weighted basket of Comex gold, Comex silver and a bitcoin proxy against long the dollar index, quarter size, dollar-notional matched. Mark: the short basket returned gold -0.67%, silver -0.78%, bitcoin +1.74% for an average of +0.10%, against DXY -0.29% — a 0.39-point loss on day one. The honest reading: the thesis leg worked and the funding leg did not. Gold and silver both fell on a live Middle East escalation, which is the strongest possible confirmation that the metals are trading the real policy rate rather than the geopolitical premium. The dollar, which was supposed to be the other side of that same rate, fell 0.29% on a day the September hike probability rose 8.4 points. That is now two consecutive sessions in which the dollar has failed to convert a hawkish repricing, and it points at the Treasury buyback programme Bloomberg has been writing about rather than at the Fed. Action: hold the quarter, but consider expressing the short leg unfunded rather than against the dollar if the DXY fails again. Catalyst: ISM 9/1, payrolls 9/4, CPI 9/11; Chinese physical demand on the Shanghai open. Invalidation, unchanged: gold reclaiming $4,664, its 27 August settle — it is at $4,499.40; or DXY back below 99.11, Thursday's close — it is at 99.415, so this one is four-tenths of a percent away and is the live risk. Sizing: a quarter.

6. Short the AI-halo basket against long the name that actually reported — hold, the mark is flat and the invalidation budget is intact

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 -0.68%, +2.84%, +2.29%, +2.95%, +0.43% for an average of +1.57%, against Nvidia +1.48% — a 0.09-point loss, essentially flat. The honest reading: the halo names outran the principal again, but by nine basis points rather than by three and a third points. Cumulatively the basket has outperformed Nvidia by 3.41 pp against a 6 pp threshold, so roughly 57% of the budget is spent across two sessions. Nvidia at $220.78 remains above the $209.66 pre-print level, so that condition is intact, and it now has its own catalyst in the $3.5bn MediaTek investment. Action: hold the quarter; do not add ahead of Tuesday. Catalyst: Palo Alto Networks reports after the close on Tuesday 1 September — the largest name in the short basket, which has risen 12.83% and 2.84% on two of the last three sessions without reporting anything; Broadcom, HPE and NetApp Wednesday evening. Invalidation: the basket outperforming Nvidia by 6 pp cumulatively, or Nvidia through $209.66.

Prior closes, marked forward. The gasoline-versus-distillate crack pair, closed on Friday, would have lost $5.77 on Monday — the largest single-session vindication of an exit in the history of this book, and the reason idea 2 above re-enters on the other side rather than pretending the original thesis was right. The long-silver-against-short-gold pair, closed on Friday on thesis invalidation rather than a stop, would have lost a further 0.11 points: silver -0.78% against gold -0.67%. The on-balance-sheet-versus-off-balance-sheet AI funding pair would have gained on Monday, with Marvell rebounding less than Broadcom's +0.42%. Two of the last three exits have now been vindicated within a session, against four that were punished, and both records are kept.

The vol note. VIX closed 14.92, up 3.40%, having traded 15.48 intraday, on a session the index lost 0.33%. A four-to-one ratio of volatility gain to index loss is the market finally paying for the calendar rather than for the tape, and the calendar is dense: ISM manufacturing and JOLTS Tuesday 10:00, ADP Wednesday 08:15, EIA petroleum Wednesday 10:30, claims and ISM services Thursday 08:30 and 10:00, payrolls Friday 08:30, then Labor Day, then PPI on the 10th and CPI on the 11th — nine dated events in nine sessions into a 16 September meeting the strip assigns a 65.4% chance of hiking. A 14.92 handle asks for roughly a 0.93% daily move. Rates volatility remains unmarkable for the current session — the MOVE series is one day stale at 70.97 — but the direction of the equity-rates gap has reversed: equity vol bid on a day the two-year did not move, which puts VIX on the right side of the calendar for the first time in a week. Own gamma dated 1 through 4 September, in the index rather than in single names, and prefer it to vega. The catalysts are dense, dated and clustered, which is a gamma configuration; and there is now a second, undated source of it in a war that reopened over a weekend and can reopen again over the next one.

Crowded consensuses to stress-test, with the numbers.

The two-sided geopolitical tape. Escalation: the U.S.-Iran war reopened on Sunday with strikes on Larak Island and Iranian attacks on the UAE and Jordan; Tehran reported a supertanker mined on Hormuz's southern route; Hormuz vessel traffic is severely disrupted into the war's seventh month; Canada's retaliation on $20bn of U.S. goods lands 8 September and Wells Fargo has already called the new Canadian steel duties "for show"; Black Sea grain disruption persists. De-escalation, or at least containment: Asia opened between 1% and 3.5% lower and closed within half a percent of unchanged, China's manufacturing PMI beat at 49.8, and Bloomberg reported the U.S. granting refiners the highest biofuel waivers since 2017, which is an administration acting on fuel costs rather than on the conflict. Domestically, California's legislature repriced two utilities by a fifth of their equity value in a session and votes again on Tuesday — a reminder that in this cycle the fastest-moving policy risk has been sub-federal.

Structural watch items. Overnight reverse repo take-up at $6.726bn on the month-end turn against $175m on Friday, with SOFR level with IORB at 3.65% for the first time in the window and reserve balances at $2.9249tn with no new print until the week containing the turn; a record ~$1.4tn of 2026 IG issuance opening a September calendar into a 4.75% ten-year, the highest since January 2025; the 6-month bill richening 3 bp on the day the September hike probability rose 8.4 points, a second consecutive session of the bill curve refusing to confirm the futures strip; Japan's 2-year at a 31-year high and the yen at 159.75 with UBP warning intervention will not hold without a fundamental shift; Germany's 2-year at its highest since July 2024 on a day the U.K. was closed; and a momentum unwind in AI names that JPMorgan's Andrew Tyler now lists alongside rate uncertainty and September seasonality as the reason his desk went "tactically cautious."

Full source links and the complete Data Notes & Conflicts appendix are in the companion file US_CrossAsset_Daily_2026-08-31_DataNotes.txt, alongside the canonical Markdown report of record.