Closing Briefing — Thursday, August 13, 2026

Sources: CNBC market live blogs for 13 August ("Stocks finish in the green"; "PPI comes in flat for July"; "Workday in talks to be acquired by a private equity firm, Reuters says"; "Netflix pops after Bill Ackman discloses a new stake"; "Stocks making big moves: Netflix, Super Micro Computer, Cisco Systems"; "Cleveland Fed President Beth Hammack says the central bank should raise rates immediately"; "Russell 2000 hits all time high"; "Reddit to join the S&P 500, shares rise"; "Number of bullish individual investors decline in latest AAII weekly survey"; "Market factoids heading into Friday's session"); Bloomberg /markets and /markets/rates-bonds boards, "S&P 500 Hits All-Time High on Inflation Cooldown", "Workday Shares Surge After Report Silver Lake Is in Talks to Buy Company", "US Jobless Applications Edged Up to 209,000 Last Week", "US High-Grade Bond Market Sees the Most Issuers Since January"; WSJ Market Data Bonds & Rates (5:04 PM ET) and Consumer Rates; Investing.com Major Indices, NDX, SOX, Russell 2000, S&P-500-component, Dow-component, European-indices and commodity-futures boards (equity boards stamped 15:59:59); U.S. Treasury Text View (official par curve, 13 Aug); BLS Producer Price Indexes — July 2026 (USDL 26-1380); CME FedWatch (live Current and Compare views, ZQU6 mid 96.3275) and Investing.com Fed Rate Monitor (Aug 13, 2026 05:45 PM EDT); FRED ICE BofA OAS series, RRPONTSYD, WRESBAL; NY Fed reference-rates API and the August indicator calendar; TradingEconomics commodities and currencies boards; Finviz Groups (Performance table view, rendered, stamped Thu Aug 13 2026 6:11 PM ET); Earnings Whispers day pages 8/14–8/21. Links in the Source Links appendix.

1. A leveraged buyout re-rated the entire software complex in one afternoon. Reuters reported Silver Lake in talks to acquire Workday; trading was halted with the stock up about 25%, and it closed near +18% at $206.45, lifting market value from roughly $43bn to $51.1bn. What makes this a cross-asset item rather than a single name is the read-across: GoDaddy +9.45%, CoStar +8.36%, Fiserv +7.64%, Global Payments +7.11%, Intuit +7.03%, Fair Isaac +6.33%, Paycom +5.85%, Shopify +5.45%, Fidelity National Information +5.27%, Adobe +4.53%, Paychex +4.18%, Salesforce +4.14%, PTC +3.99%, Tyler Technologies +3.56%, Autodesk +3.39%, FactSet +3.31%, Gen Digital +3.25%, Verisk +3.08%, Fortinet +2.88%. Nineteen software and software-adjacent names up 2.9% to 9.5% on a day the SOX rose 0.46%. The mechanism is a floor: Bloomberg's own Markets Magazine cover this week was "AI Looms Over Software Companies — and the Investors Who Piled Into Them", and a sponsor bidding at a 20% premium tells a market that has de-rated the sector all year that private capital disagrees with the public discount. Forward catalyst: whether a second software situation surfaces, and whether the deal is confirmed or denied. A single bid is a re-rate; two is a regime.

2. The semiconductor complex was the funding source, and Applied Materials is the test. SOX +0.46% against NDX +1.15%, with an intraday high of 12,704.4 and a close of 12,456.0 — a 1.96% high-to-close fade, the widest of any gauge on the board. Underneath: Arista −3.27%, Monolithic Power −4.35%, Amphenol −2.48%, Applied Materials −2.48%, Microchip −2.22%, ON Semiconductor −2.11%, Texas Instruments −1.14%, against Western Digital +7.31%, Seagate +4.91%, Micron +4.23%, Intel +3.58% and Lam Research +3.34%. Memory and storage up, logic and analog down — the dispersion inside the theme is now as wide as the theme's own move. Then Applied Materials reported after the close: a beat, guidance of roughly $8.95bn revenue and $3.36 non-GAAP EPS, calendar-2026 semiconductor equipment growth raised above 30%, management calling demand "unprecedented" — and the stock fell about 3% in extended trading, having already lost 2.48% in the regular session. Forward catalyst: the Friday open, which marks a beat-and-raise that was sold against a SOX that already lagged its index by 69 bp.

3. Cisco lost 8.38% on 10 basis points of gross margin, and it was the largest single Dow drag. The close was $113.50, −8.38%, worth −$10.38 of Dow index price on its own, against a +$69.72 index. The trigger was fiscal-first-quarter adjusted gross-margin guidance of 65–66% against a 66.1% FactSet consensus — Goldman's Michael Ng wrote before the open that "we expect CSCO to trade mixed as the stronger FY27 guide is offset by gross margin concerns." The market did not trade it mixed; it took 8.4% out of a $450bn network incumbent for a guidance midpoint 60 bp light, on a session it simultaneously paid a 20% premium for a software company on a rumour. The asymmetry between how a delivered margin miss and a rumoured take-private are priced is the single most useful positioning fact of the day. Forward catalyst: whether the networking complex follows through Friday, and Arista's −3.27% sympathy move.

4. The producer-price print was flat on the headline and hot in the line that feeds PCE, and nobody traded the second line. Final demand unchanged month-on-month against +0.2% expected; final demand goods −0.7%; final demand less foods, energy and trade services +0.4% after +0.1% in June, with portfolio management +6.5% the largest single contributor and health, beauty and optical goods retailing margins also higher (BLS). Portfolio-management and health-care PPI feed the core PCE deflator almost mechanically, which makes the 26 August print materially more hawkish than the market's reaction implies. Instead the September hike probability fell to 33.0% from 36.7% and the whole 2027 strip rallied 6.5 bp at the terminal (§8). Cleveland Fed's Beth Hammack, speaking the same morning, doubled down: "It's really critical we act now to bring inflation back to the target level. The longer it stays above the target, the harder it is to bring back down." Forward catalyst: advance retail sales 8/14 8:30, consensus +0.1%; the 26 August PCE deflator, which is now substantially knowable from this release.

5. Claims hit an eight-week high and the labour market stopped being a non-event. Initial claims 209,000 against a 202,000 consensus, up 9,000 from 199,000 (Labor Department via Bloomberg). Three weeks below 200,000 ended. The level is nowhere near the ~240,000 that would put a 2026 cut into a strip currently priced at 0.0% for every 2026 meeting, but the direction changed and it changed alongside a core PPI acceleration — the stagflationary pairing rather than the disinflationary one. The equity market's answer was to buy small caps to a new 52-week high and sell gold. Forward catalyst: claims 8/20 8:30, and whether 209,000 is a holiday-week artefact or a trend.

6. Crude broke a run and the whole energy complex went with it, except the refiners. WTI settled $81.21, −2.47%, Brent $87.00, −2.23%, both after opening higher — WTI was +1.15% at $82.31 in Asian hours on the worsening Oman spill and the Gulf of Oman and Red Sea attacks (CNBC). The reversal ran through the day: range 80.10–83.31. And yet Valero +3.85%, Phillips 66 +3.12% and Marathon Petroleum +2.33% against ConocoPhillips −2.17%, EOG −1.21%, Halliburton −1.35%, Baker Hughes −1.29%, Occidental −1.45% and Exxon −0.69%. Seven consecutive sessions of refiners over producers, and this was the first on a materially lower barrel — which is the cleanest confirmation the crack, not the barrel, is the trade (§11, §12). Forward catalyst: the EIA weekly report; any Hormuz headline.

7. Gold ended six sessions of gains with a 2.26% intraday fade, on a day the terminal rate fell. Comex December settled $4,407.10, −1.35%, after printing $4,508.97 intraday — a $101.87 round trip. Silver −1.67%, platinum −2.54%, copper −0.46%, Newmont −3.10%, Freeport −3.45%, and Basic Materials was the worst Finviz group at −1.64% (§3). The configuration this report has flagged for a week — gold rising as an inflation hedge while the dollar rose and the franc weakened — broke on the first session with a soft headline inflation print and a falling terminal rate. That is the correct sign for a rates trade and the wrong sign for an inflation hedge, so one of the two explanations for the six-day run has just been eliminated (§11, §12 idea 5). Forward catalyst: retail sales 8/14; whether the metal holds $4,300 spot.

8. Netflix +5.42% on an activist disclosure, and Reddit joined the index after the bell. Bill Ackman's Pershing Square disclosed a new position in Netflix, which closed +5.42% at $78.23 — the firm's second attempt at the name after a three-month holding in 2022. Separately, S&P confirmed Reddit will join the S&P 500 on 18 August, replacing AvalonBay Communities, which Equity Residential is acquiring with the combined company continuing in the index as Vivmark Residential; Reddit rose about 11% after hours to $175.79 despite being down 31% year to date. Both apartment REITs traded the deal in the regular session — AvalonBay +2.39%, Equity Residential +2.41%, Essex +2.18% — and Communication Services was the best Finviz group at +1.45%. Forward catalyst: the 18 August index rebalance flow; whether the Reddit add marks the bottom of a 31% drawdown or the top of an index-inclusion pop.

Six green, five red, and a 3.09-point dispersion — wider than Wednesday's 2.67 and the widest of the week. CNBC's own GICS-based tally reads seven of eleven higher with Communication Services +1.6% leading and Materials −0.7% lagging; the Finviz buckets rank the same two at the same two ends, which is the cross-check that matters. Communication Services +1.45% is a genuine breadth move rather than a megacap one: Netflix +5.42%, Paramount Skydance +4.89%, Charter +4.19%, Fox A and B +3.82%, T-Mobile +3.53%, Omnicom +3.33%, Comcast +2.79%, Verizon +2.64%, Meta +2.74% recovering Wednesday's −3.38%, Alphabet A +0.82%. The group is still −0.60% year to date, the worst on the board, so this is a laggard bounce, not leadership. Technology +1.27% is entirely software, and that is the inversion of Wednesday. Twenty-four hours ago the group's leadership was Super Micro, Dell, Seagate and the semis; today it is GoDaddy +9.45%, Fiserv +7.64%, Intuit +7.03%, Fair Isaac +6.33%, Paycom +5.85%, Fidelity National Information +5.27%, Adobe +4.53%, Paychex +4.18%, Salesforce +4.14%, PTC +3.99% — against Cisco −8.38%, Corning −5.32%, Monolithic Power −4.35%, Arista −3.27%, Applied Materials −2.48%, Amphenol −2.48%, Microchip −2.22%. The same group led on two consecutive sessions with almost no overlap in the names that did it. Real Estate +1.18% on a 5 bp fall in the 10-year had an idiosyncratic driver as well as a duration one: AvalonBay +2.39% and Equity Residential +2.41% on the merger that clears an S&P 500 seat for Reddit, with Essex +2.18%, Crown Castle +2.39% and American Tower +2.21% alongside. Basic Materials −1.64% was the worst group and it is the commodity complex in equity form: Mosaic −3.63%, Freeport-McMoRan −3.45%, Newmont −3.10%, CF Industries −2.47%, Martin Marietta, Steel Dynamics −1.01%, Linde −0.26%, against gold −1.35%, silver −1.67% and copper −0.46% (§11). Energy −0.11% on a 2.47% crude decline is the number to notice: the group barely moved because Valero +3.85%, Phillips 66 +3.12% and Marathon Petroleum +2.33% offset ConocoPhillips −2.17%, EOG −1.21%, Halliburton −1.35%, Baker Hughes −1.29% and Exxon −0.69%. Industrials −0.53% was dragged by the electrical-equipment complex — Eaton −1.44%, Rockwell −1.30%, Hubbell −1.13%, Johnson Controls −0.88% — and by Amentum −3.80%, with FedEx +3.85% and JB Hunt +2.33% on the other side. Source: Finviz Groups via rendered page (?g=sector&v=140&o=name); Finviz buckets are not official GICS/S&P sector indices — they screen all U.S.-listed names, which is why Alphabet, Meta and Netflix sit in Communication Services and Amazon and Home Depot in Consumer Cyclical.

Reconciliation. All eleven groups were checked against Wednesday's published YTD compounded by Thursday's 1-day move. Ten reconcile to within 0.04 point: Communication Services 0.9798 × 1.0145 = −0.60% vs −0.60% (exact); Consumer Cyclical 0.9763 × 1.0016 = −2.21% vs −2.21% (exact); Consumer Defensive 1.0794 × 1.0086 = +8.87% vs 8.87% (exact); Healthcare 1.0854 × 0.9975 = +8.27% vs 8.27% (exact); Basic Materials 1.1652 × 0.9836 = +14.61% vs 14.60%; Financial 1.0857 × 1.0052 = +9.13% vs 9.14%; Real Estate 1.0979 × 1.0118 = +11.09% vs 11.08%; Technology 1.2533 × 1.0127 = +26.92% vs 26.93%; Utilities 1.0237 × 0.9999 = +2.36% vs 2.37%; Energy 1.3407 × 0.9989 = +33.92% vs 33.88%, the largest of the ten at 0.04 point. Industrials fails again and in the same direction: 1.1622 × 0.9947 = +15.60% against the +16.91% shown, a 1.31-point gap on a day the group fell 0.53% and its year-to-date rose 0.69 points. That is arithmetically impossible in a fixed basket, and it is the second consecutive session the same group has broken — so the Finviz Industrials constituent set is being revised rather than mis-stamped. The 1-day and 1-week figures remain usable; the Industrials YTD is not comparable to the prior edition's and is flagged rather than reconciled. Four groups reconcile exactly; maximum deviation excluding Industrials is 0.04 point (Energy).

Up

•  Workday (WDAY) +18% to $206.45 (not carried by the constituent board used here — see Data Notes) — the move of the day. Reuters reported Silver Lake is in talks to take the company private in what would rank among the largest software buyouts ever; shares were halted after jumping about 25% and closed near +18%, taking market value from roughly $43bn to $51.1bn. Talks are ongoing and no deal is assured. The stock was down 18% year to date before the report.

•  GoDaddy (GDDY) +9.45% to $100.50, CoStar (CSGP) +8.36% to $33.05, Fiserv (FI) +7.64% to $55.50, Global Payments (GPN) +7.11% to $94.83, Intuit (INTU) +7.03% to $358.24, Fair Isaac (FICO) +6.33% to $1,110.32, Paycom (PAYC) +5.85% to $223.63, Shopify (SHOP) +5.45% to $158.61, Fidelity National Information (FIS) +5.27% to $44.13, Adobe (ADBE) +4.53% to $270.46, Paychex (PAYX) +4.18% to $125.40, Salesforce (CRM) +4.14% to $201.32, PTC (PTC) +3.99% to $153.89, CDW (CDW) +3.94% to $142.05, Tyler Technologies (TYL) +3.56%, Autodesk (ADSK) +3.39%, FactSet (FDS) +3.31%, Gen Digital (GEN) +3.25%, Verisk (VRSK) +3.08%, AppLovin (APP) +2.93% to $312.67, Fortinet (FTNT) +2.88%, Equifax (EFX) +2.84%, Palo Alto Networks (PANW) +2.32% — the software and payments read-across from the Workday report. The iShares Expanded Tech-Software ETF (IGV) rose more than 3% and is up about 28% over six months after falling more than 24% in the first quarter (CNBC).

•  Netflix (NFLX) +5.42% to $78.23 — Bill Ackman's Pershing Square disclosed a new stake, its first since a three-month position in early 2022 that was exited after the company's first subscriber decline in a decade.

•  Western Digital (WDC) +7.31% to $487.28, Seagate (STX) +4.91% to $921.37, Micron (MU) +4.23% to $949.83, Lam Research (LRCX) +3.34% to $337.01, Intel (INTC) +3.58% to $104.56 — the memory-and-storage leg of the semiconductor complex kept working while logic and analog did not (§2 item 2). Intel is now +3.7% over two sessions and through $104.

•  Robinhood (HOOD) +4.70% to $99.37, Palantir (PLTR) +4.65% to $179.00, KKR (KKR) +3.97% to $115.30, Corpay (CPAY) +3.80%, Keurig Dr Pepper (KDP) +5.28% to $31.12, Paramount Skydance (PSKY) +4.89%, Charter Communications (CHTR) +4.19% to $156.51 (reversing Wednesday's −4.78%), Super Micro (SMCI) +4.09% to $39.15 (a second post-earnings advance, +23.9% over two sessions).

•  Valero (VLO) +3.85% to $342.92, Phillips 66 (PSX) +3.12% to $232.61, Marathon Petroleum (MPC) +2.33% to $356.37 — the refiners rose 2.3% to 3.9% on a barrel that fell 2.47%. Seventh consecutive session of refiners over producers (§11, §12).

•  FedEx (FDX) +3.85% to $339.35, Fox Corp A (FOXA) +3.82%, Fox Corp B (FOX) +3.82%, Tesla (TSLA) +3.80% to $339.96, Kraft Heinz (KHC) +3.55%, T-Mobile (TMUS) +3.53% to $183.38, Norwegian Cruise Line (NCLH) +3.41%, Omnicom (OMC) +3.33%, Zoetis (ZTS) +3.13% to $75.84 (reversing Wednesday's −2.40%), CBRE (CBRE) +3.05%, S&P Global (SPGI) +3.04% to $422.56, Campbell's (CPB) +3.01%, Conagra (CAG) +2.84%, Best Buy (BBY) +2.83%, DR Horton (DHI) +2.81% to $149.92 (reversing Wednesday's −3.30%), Comcast (CMCSA) +2.79%, Lennar (LEN) +2.75% to $87.54, Meta (META) +2.74% to $594.73, Carnival (CCL) +2.71%, Verizon (VZ) +2.64% to $48.22, Axon (AXON) +2.62% to $615.50 (recovering 46% of Wednesday's −5.74%), ICE (ICE) +2.58%, ADP (ADP) +2.58%, JB Hunt (JBHT) +2.33% to $282.39, PayPal (PYPL) +2.33%, Deckers (DECK) +2.30%, eBay (EBAY) +2.22% to $104.58, Equity Residential (EQR) +2.41%, AvalonBay (AVB) +2.39%, Crown Castle (CCI) +2.39%, American Tower (AMT) +2.21%, Essex Property (ESS) +2.18%, Moody's (MCO) +2.21%, Merck (MRK) +1.98% to $135.55, Oracle (ORCL) +1.97% to $156.30, Nike (NKE) +1.78% to $41.23, Hewlett Packard Enterprise (HPE) +1.74% to $59.81, Visa (V) +1.59% to $365.14, Walt Disney (DIS) +1.53% to $104.80, Dell (DELL) +2.07% to $494.51, Albemarle (ALB) +1.67%, Kinder Morgan (KMI) +1.10%, Lowe's (LOW) +1.04% to $218.22 (reports 8/19 BMO), Qualcomm (QCOM) +1.07%, PulteGroup (PHM) +1.08%, Apple (AAPL) +1.00% to $305.26.

•  Reddit (RDDT) +11.18% to $175.79 after hours (not an S&P 500 constituent until 18 August) — S&P confirmed the company will join the S&P 500 on 18 August, replacing AvalonBay Communities. Reddit is down 31% year to date going into the add.

•  Birkenstock (BIRK) +10% premarket (not an S&P 500 constituent) — quarterly revenue and adjusted EBITDA above expectations, with full-year guidance seen at the high end of prior. Maersk +7% in Copenhagen after raising 2026 EBITDA guidance for the second time this year, reporting preliminary second-quarter underlying EBITDA of $3.0bn against a $2.04bn LSEG consensus — the clearest single read on what the Hormuz blockade is doing to freight economics. Samsonite +7.2% in Hong Kong on an agreement to buy 85% of BÉIS for $178.5m.

•  Analyst action: UBS reiterated Buy on Brinker International (EAT) and raised its target to $285 from $260, about 16% above Wednesday's close (not an S&P 500 constituent). Dennis Geiger: "We're encouraged by Chili's sales momentum into F1Q, with QTD sss meaningfully accelerating from F4Q levels (5.6%) and strength sustaining from July into early August." The stock had already risen more than 11% on Wednesday's print.

Down

•  Tapestry (TPR) −16.49% to $128.39 — the worst S&P 500 performer by a factor of two. Fiscal fourth-quarter EPS $1.32 against a $1.28 FactSet consensus on revenue of $1.88bn against $1.87bn — a double beat — with the quarterly dividend raised to 46.25 cents from 40 cents. The stock fell 7% premarket and then more than doubled the decline through the session. It had already fallen 6.7% over the two sessions into the print, so the two-day drawdown is now about 22%. A beat, a raise and a dividend increase, sold to the tune of one-sixth of the equity, is the sharpest single-name statement on consumer positioning this month.

•  Cisco (CSCO) −8.38% to $113.50 — first-quarter adjusted gross-margin guidance 65–66% against a 66.1% FactSet consensus, overshadowing what Goldman called a stronger fiscal-2027 guide. −$10.38 of Dow index price, the largest single drag (§2 item 3).

•  Corning (GLW) −5.32% to $158.54 — reversing Wednesday's +5.18% and then some, with no company catalyst in the reviewed sources; the optical-components read-across from the Cisco guide is the most plausible transmission.

•  Monolithic Power (MPWR) −4.35% to $1,363.00, Arista Networks (ANET) −3.27% to $203.62 (reversing Wednesday's +6.39%), Amphenol (APH) −2.48% to $165.75, Applied Materials (AMAT) −2.48% to $534.54 (reported 8/13 AMC — see §5), Microchip (MCHP) −2.22% to $77.68, ON Semiconductor (ON) −2.11% to $81.56, Texas Instruments (TXN) −1.14% to $273.43 — the logic-and-analog leg.

•  Amentum (AMTM) −3.80% to $21.67 — a second consecutive decline after its 8/11 print.

•  Mosaic (MOS) −3.63% to $21.78, Freeport-McMoRan (FCX) −3.45% to $66.83, Newmont (NEM) −3.10% to $114.19, CF Industries (CF) −2.47% to $117.05, Martin Marietta (MLM) −1.27%, Steel Dynamics (STLD) −1.01% — the materials complex tracked gold −1.35%, silver −1.67% and copper −0.46% (§11).

•  Ulta Beauty (ULTA) −3.07% to $515.57, McKesson (MCK) −2.51% to $856.69 (a second consecutive decline, −5.1% over two sessions), Generac (GNRC) −2.36%, Ralph Lauren (RL) −2.26% to $387.39 (the Tapestry read-across), ConocoPhillips (COP) −2.17% to $124.54, Teledyne (TDY) −1.87%, Willis Towers Watson (WTW) −1.85%, Darden (DRI) −1.84% to $223.50 (reversing Wednesday's +4.01%), Vertex (VRTX) −1.80%, Stryker (SYK) −1.76%, West Pharmaceutical (WST) −1.74%, Fastenal (FAST) −1.69%, F5 Networks (FFIV) −1.64% to $416.00, UnitedHealth (UNH) −1.58% to $399.18 (−$6.41 of Dow price, the second-largest drag), Centene (CNC) −1.54%, Occidental (OXY) −1.45% to $57.70, Eaton (ETN) −1.44% to $453.33, Halliburton (HAL) −1.35%, Rockwell Automation (ROK) −1.30% to $445.51, McDonald's (MCD) −1.27% to $272.20 (−$3.50 of Dow price), Baker Hughes (BKR) −1.29%, EOG Resources (EOG) −1.21% to $141.41, Hubbell (HUBB) −1.13% to $507.10, Eli Lilly (LLY) −0.92% to $1,209.00, Johnson Controls (JCI) −0.88%, Williams (WMB) −0.88%, Amazon (AMZN) −0.80% to $265.13 (−$2.14 of Dow price), Exxon Mobil (XOM) −0.69% to $158.65, 3M (MMM) −0.62%, JPMorgan (JPM) −0.57% to $363.11, Honeywell (HON) −0.57% to $234.00, Home Depot (HD) −0.50% to $341.70 (reports 8/18 BMO), Hilton (HLT) −0.50% to $320.92, Boeing (BA) −0.38% to $230.32, Walmart (WMT) −0.23% to $115.74 (reports 8/20 BMO), Caterpillar (CAT) −0.12% to $854.60.

•  Bank of America (BAC) −1.10% to $64.10 — the largest decliner among the money-centre banks on a day the KBW Nasdaq Bank Index slipped 0.10% to 193.01 (WSJ board) while the Finviz Financial group rose 0.52%. Citigroup +0.84% and Goldman +0.52% went the other way.

•  After the close: Applied Materials (AMAT) −~3% despite a beat, guidance of roughly $8.95bn revenue and $3.36 non-GAAP EPS and calendar-2026 equipment growth raised above 30% on "unprecedented" AI demand — the stock is up close to 100% year to date, and that is the most-cited explanation for the reaction.

Times are ET. Every day page from Friday 8/14 through Friday 8/21 was independently re-pulled from Earnings Whispers this session (/1 = before open, /2 = after close) and screened against the Investing.com S&P-500-component board. The 8/10–8/13 rosters are carried from prior verified pulls with reactions added from this session's boards. Re-verify times and membership against company IR before trading any date.

Mon 8/10 — completed. BMO: Berkshire Hathaway B (BRK.B) 8:00 — closed −0.53% at $507.18 on Thursday, a fourth consecutive decline and now −4.3% from the post-print close. AMC: Simon Property Group (SPG) 4:05 — closed +0.41% at $221.47.

Tue 8/11 — completed. BMO: Cardinal Health (CAH) 6:45; Amentum (AMTM) 8:00 — closed −3.80% at $21.67, a second consecutive decline. AMC: Lumentum (LITE) 4:00; Super Micro Computer (SMCI) 4:05 — +4.09% to $39.15, a second post-print advance and +23.9% over two sessions.

Wed 8/12 — completed. BMO: Amcor (AMCR) 6:00; Trimble (TRMB) 6:55. AMC: Cisco (CSCO) 4:05 — closed −8.38% at $113.50, the largest single drag on the Dow at −$10.38 of index price, on first-quarter adjusted gross-margin guidance of 65–66% against a 66.1% FactSet consensus; Coherent (COHR) 4:05 — the photonics read-across showed up in Corning −5.32% rather than in the name itself.

Thu 8/13 — completed. BMO: Tapestry (TPR) 6:45 — −16.49% to $128.39, the worst performer in the index, on a $1.32 EPS beat against $1.28 FactSet, revenue $1.88bn against $1.87bn and a dividend raised to 46.25 cents from 40 cents; roughly −22% over two sessions. AMC: Applied Materials (AMAT) 4:00 — closed −2.48% at $534.54, then fell about 3% after the bell despite a beat, guidance of roughly $8.95bn revenue and $3.36 non-GAAP EPS, and calendar-2026 equipment growth raised above 30% on "unprecedented" demand.

Fri 8/14. Neither page lists an S&P 500 reporter. The before-open page is entirely micro-cap, biotech and materials; the after-close page is micro-cap and gaming. None of the reviewed names is carried by the constituent board.

Mon 8/17. Neither page lists an S&P 500 reporter (ten names screened across both buckets).

Tue 8/18. BMO: Home Depot (HD) 6:00. AMC: Keysight Technologies (KEYS) 4:05, Jack Henry & Associates (JKHY) 4:15.

Wed 8/19. BMO: Lowe's (LOW) 6:00, Estée Lauder (EL) 6:00, Target (TGT) 6:30, Analog Devices (ADI) 7:00, TJX Companies (TJX) 7:30. AMC: Nordson (NDSN) 4:30.

Thu 8/20. BMO: Deere & Company (DE) 6:20, Walmart (WMT) 7:00. AMC: Ross Stores (ROST) 4:00.

Fri 8/21. Neither page lists an S&P 500 reporter (four names screened before the open; the after-close page returned "NONE").

Changes vs. the prior calendar (8/12 report):

•  No additions and no removals. Every 8/14–8/21 name carried by the prior edition re-appeared on this session's independent re-pull with an identical timestamp: HD 6:00, KEYS 4:05, JKHY 4:15, LOW 6:00, EL 6:00, TGT 6:30, ADI 7:00, TJX 7:30, NDSN 4:30, DE 6:20, WMT 7:00, ROST 4:00. TPR 6:45 and AMAT 4:00 both reported as scheduled.

•  Index membership change to diarise, not an earnings item: Reddit (RDDT) joins the S&P 500 on 18 August, replacing AvalonBay Communities (AVB), which Equity Residential is acquiring; the combined company remains in the index as Vivmark Residential. Neither name reports in this window, but the 18 August rebalance sits on the same morning as the Home Depot print.

•  Membership caveat, restated rather than buried: the constituent board used as this report's screening proxy does not carry Coherent (COHR) — retained for continuity across six editions — and does not carry Workday (WDAY), which is why the day's largest single-name move is flagged in §4 rather than treated as an index constituent. Confirm both with company IR. The same board does carry HD, KEYS, JKHY, LOW, EL, TGT, ADI, TJX, NDSN, DE, WMT, ROST, TPR and AMAT.

•  Conservative exclusions on this pull: Fabrinet (FN) 4:15 and Flexsteel (FLXS) 4:15 on 8/17; Toll Brothers (TOL) 4:30 and Mercury Systems (MRCY) 4:00 on 8/18; Coty (COTY) 4:30, BILL Holdings (BILL) 4:00, Wolfspeed (WOLF) 4:05 and John B. Sanfilippo (JBSS) 4:10 on 8/19; Flowers Foods (FLO) 4:05, Advance Auto Parts (AAP) 6:30 and OSI Systems (OSIS) 4:00 on 8/20; BJ's Wholesale (BJ) 6:45 and Buckle (BKE) 6:50 on 8/21. Toll Brothers remains the most borderline of the group and all are listed in Data Notes.

•  Timing bucket still unpublished: none. Every name above carries a specific clock time.

•  What the forward calendar hands the desk. The current week is now closed with the two most consequential prints already marked: Applied Materials sold a beat-and-raise and Tapestry sold a beat, a revenue beat and a dividend increase by 16.5%. Those are the two data points to carry into next week's block, because next week is six U.S. consumer reporters — Home Depot, Lowe's, Target, TJX, Walmart and Ross Stores — compressed into three sessions, and the market has just demonstrated twice that it is marking consumer and equipment beats down rather than up. Analog Devices (8/19 BMO) is the analog-semiconductor read against a session in which analog and logic were sold while memory and storage were bought (§2 item 2), and Keysight (8/18 AMC) the test-and-measurement one. There is no S&P 500 reporter on either Friday, so the two four-day gaps in the calendar both belong to macro (§7).

•  Names on the reviewed pages whose S&P 500 membership could not be conservatively verified — Outlook Therapeutics (OTLK), Acurx (ACXP), LanzaTech (LNZA), Creative Media & Community Trust (CMCT), PAVmed (PAVM), Suncrete (RMIX), Sinda (SIND), Americas Gold and Silver (USAS), HIVE Digital (HIVE), Duos (DUOT), Flexible Solutions (FSI), Cloudastructure (CSAI), New Era Energy (NUAI), Parabolic Technologies (PARA), Bally's (BALY), Barfresh (BRFH), H World (HTHT), Freightos (CRGO), BitFuFu (FUFU), InspireMD (NSPR), AXE Compute (AGPU), Flexsteel (FLXS), Fabrinet (FN), XP (XP), DocGo (DCGO), Yalla (YALA), Baidu (BIDU), Klarna (KLAR), Amer Sports (AS), Prenetics (PRE), Hesai (HSAI), VNET, Corporación América Airports (CAAP), iQIYI (IQ), Pony AI (PONY), ReNew Energy (RNW), uCloudlink (UCL), Eltek (ELTK), Elauwit (ELWT), Einride (ENRD), Auna (AUNA), SQM, La-Z-Boy (LZB), ZTO Express (ZTO), Mercury Systems (MRCY), Toll Brothers (TOL), AXIL Brands (AXIL), Iridex (IRIX), Full Truck Alliance (YMM), Viking (VIK), ZIM, Datavault AI (DVLT), Kingsoft Cloud (KC), TOYO, Weibo (WB), BILL Holdings (BILL), Webull (BULL), Coty (COTY), Alvotech (ALVO), Carlyle Credit Income Fund (CCIF), ClearSign (CLIR), John B. Sanfilippo (JBSS), Telix (TLX), Unifi (UFI), Accuray (ARAY), Ionic Digital (IOND), Wolfspeed (WOLF), Alibaba (BABA), Advance Auto Parts (AAP), Autohome (ATHM), Futu (FUTU), NetEase (NTES), Aegon (AEG), Atour (ATAT), Daqo (DQ), LSI Industries (LYTS), ATRenew (RERE), Flowers Foods (FLO), Flux Power (FLUX), OSI Systems (OSIS), eXoZymes (EXOZ), Intchains (ICG), BJ's Wholesale (BJ), KE Holdings (BEKE), Buckle (BKE), ZKH Group (ZKH) — are excluded; see Data Notes.

One-week column measured against the 5 August par curve, the same weekday one week prior.

The shape and the diagnostic: a belly-led bull steepener with the bill anchored. The 5-year fell most at −6 bp; the 2-year, 3-year, 7-year and 10-year all fell 5 bp; the 20-year fell 4 and the 30-year 3. At the very front, the 3-month bill did not move at all and the 1-month and 2-month actually rose a basis point. That combination is not a general duration rally, it is a specific statement, and the statement is about when rather than whether. A market that pushes a hike out of September but keeps it in the strip leaves the bill — which matures before the December meeting can bite — exactly where it was, and rallies the two-to-five-year sector where the deferred hike actually sits. 3M10Y flattened 5 bp to 76 bp precisely because the numerator moved and the denominator did not. Meanwhile 2s30s steepened 2 bp to 106 bp: the long end participated least in a rally driven by policy timing, because policy timing is not what the 30-year is priced off.

The one-week frame now says something different from what it said on Wednesday, and it is worth stating plainly. A week ago this report described a bear steepener at the back sitting on a bull flattener at the front, with everything from five years out higher and the front lower. That has half-reversed. On the week the 5-year is now only −1 bp, the 7-year and 10-year are flat, and the 20-year and 30-year are +2 and +4 bp — so the term-premium build of the past fortnight has been trimmed but not undone, while the front has extended its rally (1-year −6 bp, 6-month −4 bp, 2-month −3 bp). 2s30s at +7 bp on the week is still the dominant weekly signal. The market has spent five sessions buying the front, has now spent one session buying the belly as well, and has still not bought the long bond. For a real-money book the trade-relevant fact is unchanged in direction and smaller in magnitude: policy expectations are moving, term premium is not.

Bills separately, because they are a funding signal rather than a policy one (§9b). The 3-month par yield held at 3.87% for a second consecutive session while the 1-month and 2-month rose a basis point each — so the very front of the bill curve flattened between one and three months even as everything from six months out rallied 3 to 6 bp. That is what a settling funding market looks like when policy expectations are falling around it: SOFR printed 3.62% for 12 August, two basis points lower, and the 99th-minus-1st percentile span narrowed back from 15 bp to 13 bp (§9b). The bill did not follow the coupon curve because the bill is not trading the Fed card. Vendor cross-check on the coupon curve: Bloomberg's board marks the U.S. 10-year at 4.64%; WSJ real-time at 5:04 PM ET has 10Y 4.647% (−4.3 bp), 30Y 5.216% (−4.5 bp), 5Y 4.317% (−6.6 bp), 3Y 4.215% (−7.6 bp) and 2Y 4.155% (−5.2 bp). WSJ's marks are taken two hours after the official 3 p.m. par fixing and are directionally identical everywhere; the 3-year's 2.6 bp discrepancy is the largest and is a timing artefact, not a level dispute. WSJ also has the 30-year fixed mortgage at 6.74%, down from 6.76% a week ago — two basis points of relief against a 5-year that has fallen only one on the week, which is why the housing complex traded better on Thursday than the duration move alone would justify (§4).

Current week (Aug 10–14) — released

Current week — remaining

Next week (Aug 17–21)

Current target range: 3.50–3.75% (WSJ Consumer Rates table, unchanged for the period shown; IORB 3.65%). The market prices hikes, not cuts: the probability of a target range below 3.50–3.75% is 0.0% at every 2026 meeting on both vendors.

The headline — CME FedWatch, 16 September meeting

Meeting information from CME's Current view: contract ZQU6, expiring 30 September, mid price 96.3275, prior volume 47,006, prior open interest 239,864.

Provenance of every column, stated. CME's Current view published the EASE / NO CHANGE / HIKE headline numerically and it was read live at approximately 6:20 PM ET — per this report's own standing caveat, a live read after 5:00 p.m. ET is indicative rather than an end-of-day settlement snapshot, and if CME's settled figure later differs it will be disclosed and corrected in the next edition rather than quietly dropped. The Compare view plots the four series but publishes no numeric table on this run, so: 1 DAY reproduces the prior edition's CME figures (59.9 / 40.1 / 0.0 for 12 August), and CME's plotted 1D bars match those heights. 1 WEEK is taken from Investing.com's independent previous-week column (hold 44.9%, +25 55.1%) and corroborated against CME's 1W bars at roughly 45% and 55%. 1 MONTH cannot be carried: CME's legend shows a reference date of 13 July 2026 against the prior edition's 10 July 2026, a different date, so those cells are marked as chart-read approximations to about a point rather than presented as vendor figures. No estimated cell is published here as though it were tabulated.

The cross-check — Investing.com Fed Rate Monitor, and the gap reconciled

Investing.com's card for 16 September, updated 5:45 PM EDT on 13 August, shows hold 67.0% (previous day 63.3%, previous week 44.9%) and +25 bp 33.0% (36.7%, 55.1%), on a September fed funds contract price of 96.330.

The gap is 1.8 percentage points on the hold, and it reconciles almost exactly to the quoted contract price. CME's mid is 96.3275; Investing.com's is 96.330 — a difference of 0.25 basis points. The 16 September meeting falls on day 16 of a 30-day contract month, so a certain 25 bp hike moves the September average effective rate by roughly 25 × (14/30) = 11.7 bp, which makes one basis point of ZQU6 worth about 8.6 points of headline probability. Multiply: 0.25 bp × 8.6 = 2.1 points, against an observed gap of 1.8. The two vendors are not disagreeing about anything; they are quoting the same contract 0.25 bp apart. Note also that both vendors' previous-week columns now agree to the decimal at 44.9%, which is a stronger corroboration of the 1 WEEK provenance than this section has usually been able to offer.

(a) Current-year meeting distributions — current / [prev day] / [prev week]

Modal ranges in bold. Contract prices: ZQU6 96.330, ZQV6 96.280, ZQZ6 96.165.

The one-day and one-week changes, and the momentum has changed shape. In one session the September hike probability fell 3.7 points (36.7% → 33.0%), the October cumulative 4.3 points (52.7% → 48.4%) and the December cumulative 6.5 points (72.1% → 65.6%). That ordering is the inverse of Wednesday's. Twenty-four hours ago the removal decayed with tenor — 9.7 points out of September, 8.6 out of October, 5.2 out of December — which is the signature of deferral: the hike moves later rather than away. Today the removal increased with tenor, and the further out the meeting the more was taken out. Deferral has become cancellation at the margin, and it happened on a session whose core producer print accelerated to +0.4% (§7). Over one week the three fell 22.1, 20.2 and 18.4 points — still a large front-loaded move, but the weekly gradient has flattened from a 8.5-point spread between September and December to 3.7 points. The multi-day read: CME's own columns put the September hold at roughly 25% a month ago, 44.9% a week ago, 59.9% a day ago and 65.2% now. The hold has gone from a minority outcome to a two-in-three outcome inside four weeks, and the +50 bp September tail, which the chart shows at roughly 24% in mid-July, is 0.0% on both vendors.

(b) Next-year meeting path

The contract prices moved a long way, and this is the single most important number in the section. The lowest price on the strip is still 96.015 at the 28 July 2027 meeting, but it now implies 3.985% against 4.050% on Wednesday — the terminal rate fell 6.5 basis points in one session, the largest single-day move in the terminal this report has recorded. Every 2027 contract rallied: December 2027 from 95.995 to 96.060, June 2027 from 95.955 to 96.020, January 2027 from 96.100 to 96.140. The modal range remains 3.75–4.00% at every single 2027 meeting and the modal probability still decays monotonically, from 42.7% in January to 35.3% in December, so the shape is unchanged and the whole distribution has shifted down. The first non-zero probability of a target range below 3.50–3.75% now appears at the 28 July 2027 meeting at 0.4%, rising to 4.2% by December 2027 — one meeting earlier than Wednesday, when the first non-zero cut sat in October 2027. Eleven months of futures with no easing priced at all, down from fourteen.

(c) Year-end probability ladders

Year-end 2026 (9 December 2026 meeting), relative to the current 3.50–3.75%:

Year-end 2027 (8 December 2027 meeting), relative to the current 3.50–3.75%:

The week-over-week arc, and the divergence this report has tracked for a fortnight closed today. Wednesday's edition described a front end pricing hikes out while the back end priced them in: the cumulative probability of sitting above 3.50–3.75% at the end of 2027 had risen from 73.9% to 78.3% over five sessions even as 16 points came out of September. That reversed in a single day. The 2027 year-end cumulative-above fell from 79.0% on Wednesday to 73.2%, a 5.8-point one-day move, and it is 8.7 points below the 81.9% of a week ago. The cumulative-below rose to 4.2% from 3.0% on Wednesday and 3.5% a week ago. Both ends of the strip priced hikes out on the same session for the first time in this report's window — hold at year-end 2026 up 6.5 points to 34.4%, hold at year-end 2027 up 4.7 points to 22.6%, +50 and +75 outcomes down 3.1 and 3.4 points at the back. That is a parallel dovish shift rather than a re-timing, and it is precisely what a market does when it reads the goods line of a producer print and stops there.

Rounding, stated transparently. Investing.com's cards do not always sum to exactly 100.0%: the September, October, December 2026, January 2027, September 2027, October 2027 and December 2027 cards each sum to 100.0, while March 2027, April 2027, June 2027 and July 2027 each sum to 99.9. CME's September headline sums to 100.0. Cumulative figures above are computed by summing the vendor's published cells without re-normalising, so a cumulative may carry the same ±0.1 rounding as its parent card. The 1 MONTH column of the CME table is read off plotted bar heights and is accurate to about one point; it is not used in any cumulative or change calculation anywhere in this report.

(a) IG and HY credit spreads

As-of date, stated rather than implied: the ICE BofA option-adjusted spread series below are FRED's 12 August prints, published on the morning of 13 August. FRED carries these with a one-business-day lag, so there is no 13 August index credit spread available at the time of writing; the same-day direction is cross-checked against Bloomberg and WSJ coverage below.

CDX retrieval note — the six-step ladder was worked again and the index level remains publicly unobtainable. (1) Bloomberg in Chrome: /markets and /markets/rates-bonds were both rendered; the latter publishes the Bloomberg Fixed Income Indices grid — Global Aggregate 500.67, +0.98 — but contains no CDX string anywhere in the page text, verified programmatically. (2) WSJ Market Data Bonds & Rates: Treasurys, consumer rates, government-bond and money-rate tables — no CDX. (3) Cbonds / IHS Markit index pages: both the CDX.NA.IG 5Y and CDX.NA.HY 5Y pages were rendered directly and each displays the level as *** bps with the previous value masked to *** on 11/08/2026 — entitlement-gated on both. (4) FT Markets Data / Reuters credit wraps: no CDX level in the reviewed 13 August material. (5) Search across ICE and S&P Dow Jones product pages: the product and options pages carry specifications, not levels; one third-party digest quoted "investment grade trades at 81 bps" with no dated attribution and it is withheld rather than published, because a single undated digest figure is exactly the kind of number this report has been wrong to trust before. (6) Cash-market proxy, labelled as such: the Bloomberg Global Aggregate rose 0.98 points on a session when coupon Treasury yields fell 3–6 bp — a duration-only explanation covers essentially all of it, leaving no measurable residual tightening to report. All six steps are named so the gap is auditable rather than asserted. Quoting convention reminder: CDX IG 5y is quoted in basis points of spread and CDX HY 5y in price points, where a rising price means tightening credit spreads.

What the numbers say: the tail stopped widening, and the index has stopped moving at all. The IG credit spread was unchanged at 79 bp, exactly where it started the year and inside a 1 bp range for eight consecutive prints. The HY credit spread tightened 1 bp to 271 bp — a new 2026 tight, 10 bp inside January. The CCC credit spread tightened 3 bp to 1,020 bp, and the CCC-minus-HY differential narrowed 2 bp to 749 bp, which partially reverses Wednesday's 7 bp widening but leaves the differential 1 bp wider on the week and 145 bp wider than January's 604 bp. The mechanism worth naming: the tail's Wednesday widening did not extend, so the four-print widening trend this report flagged is now a three-print widening and one retracement — not yet a break, and specifically not yet the confirmation §12's protection trade needs. Same-day direction cross-check: neither Bloomberg's 13 August markets coverage nor WSJ's bonds page reported a credit-spread move of any size on the session, which is consistent with a 0–3 bp index drift being invisible at the headline level. The one credit-relevant equity tell is that Bank of America fell 1.10% and the KBW Nasdaq Bank Index slipped 0.10% to 193.01 (WSJ board) on a day the S&P set a record — bank equity did not participate in a record close, which is a mild but real dissent from the credit picture.

(b) Money-market and funding plumbing

Basis, stated: the NY Fed publishes reference rates at approximately 8:00 a.m. ET for the prior business day, so the SOFR, EFFR, OBFR and repo figures below are 12 August effective rates published on 13 August. ON RRP is same-day (13 August); reserve balances are the weekly average for the week ended 12 August.

The distribution relaxed and the buffer kept shrinking, and those two facts point in opposite directions. The good news first: the secured mean fell 2 bp to 3.62%, tri-party and GC repo both fell 3 bp to 3.60%, and the SOFR 99th-minus-1st span narrowed from 15 bp to 13 bp after two consecutive sessions of widening. Dispersion between cash-rich and cash-poor counterparties eased, which is what the two-session pattern flagged on Wednesday needed in order not to become a trend. The bad news is structural rather than daily: ON RRP printed $0.450bn, a new record low, on the same day reserve balances fell $49.3bn to $2,944.1bn — the first sub-$2.95tn week since late June and $198.7bn off the July peak. The facility that used to absorb the marginal dollar now holds less than half a billion; the reserve stock that replaced it is falling at roughly $50bn a week. SOFR printing three basis points through IORB while repo volumes exceed $1.2tn a day says there is no scarcity today, and that is the correct reading — but a system draining reserves at this pace with no RRP cushion is a system whose September quarter-end has no shock absorber. The NY Fed's Reserve Demand Elasticity release on 20 August is the one scheduled number that speaks to it directly, and it is upgraded to Medium sensitivity in §7 for that reason.

(c) Rates volatility and swap spreads

The vol tell of the session is small and it has the wrong sign. The VIX rose 0.55% to 14.63 on the day the S&P 500 set a record intraday high and a record close. That is not a large move and it is well inside the noise band — but the index closed 0.22% below its own high, the SOX faded 1.96%, and the AAII survey published the same afternoon showed bulls falling to 34.7% from 37.0% with bears at 37.9% (§13). Three independent measures of the marginal buyer said the same thing on a record day. At 14.63 one-month implied prices a daily S&P move of about 0.92%, against realised of +0.26%, −0.06% and +0.65% over the last three sessions — implied is still above realised, so the short-vol carry is intact and the position stays crowded. The honest gap in this table is unchanged and it matters this week: §6 shows the 30-year down only 3 bp against a 5-year down 6, which is a term-premium statement that would show up first in the 30-year swap spread, and no public source publishes it.

(d) Issuance, leveraged loans and private credit

•  The primary market is running at a record annual pace, and Bloomberg has now put a number on it. There have been $1.4tn of U.S. IG notes sold in 2026 to date, 9% above the pace of 2020, the year that ended at a record $1.75tn. On 10 August, nineteen issuers came to market — the most in seven months — ranging from utilities to overseas banks to Tyson Foods, raising capital ahead of the week's inflation prints. For context on the seasonal, August has averaged $95bn of IG sales since 2019 and typically empties out in the last ten days before a Labor Day surge. A record year-to-date pace and the busiest single day since January, in the slowest month of the calendar, is the clearest available evidence that IG issuers see no reason to wait — consistent with an index credit spread that has not moved a basis point in eight prints.

•  The take-private bid escalated from a restaurant chain to a $51bn software platform in twenty-four hours. Wednesday's leveraged-finance signal was the Financial Times reporting a Trian-led bid for Wendy's. Thursday's was Reuters reporting Silver Lake in talks to buy Workday, which would rank among the largest software buyouts ever; the stock closed near +18% at $206.45 for a $51.1bn market value, and Silver Lake could bring in additional investors to finance it. Two sponsor situations in two sessions, one of them a mega-cap, is the clustering this report said would distinguish a re-opened LBO market from a single opportunistic bid. The financing condition behind it is in the table above: an HY credit spread at 271 bp, a new 2026 tight. That is not a coincidence; it is the arithmetic.

•  The private-credit and AI-infrastructure concentration is now a stated public-market theme rather than an inferred one. Bloomberg's Markets Magazine cover feature this week is "AI Looms Over Software Companies — and the Investors Who Piled Into Them", on buyout funds and lenders drawn to reliable software revenue and low costs, with the framing that the entire business model is at risk; a companion piece is headlined "'SaaSpocalypse' Risk From AI Reaches Beyond Private Equity". Thursday's tape traded directly against that thesis — a sponsor bid for Workday and a 3%-plus move in the software ETF. When the sell-side narrative and the sponsor bid point in opposite directions on the same asset, the spread between public and private marks is the position, and it is exactly the kind of gap that gets closed violently in one direction.

•  IG and HY primary volumes for the specific session, the Morningstar LSTA leveraged loan index and bank CDS: not obtainable this session. Sources attempted and named so the gap is auditable: Bloomberg /markets and /markets/rates-bonds and the 13 August markets coverage, WSJ Market Data Bonds & Rates, the Cbonds CDX.NA.IG and CDX.NA.HY pages, and the Investing.com and TradingEconomics boards. None published a session-level primary volume, a loan-index level or a bank CDS quote for 13 August. The KBW Nasdaq Bank Index at 193.01, −0.10% (WSJ board) and the Finviz Financial group at +0.52% are the only bank-risk proxies this report can source, and both are equity proxies rather than credit ones.

Quote basis — read this before the table. All pairs are spot in the market convention shown; a positive move on a USD/XXX pair means the dollar strengthened, and on EUR/USD, GBP/USD, AUD/USD and NZD/USD it means the dollar weakened. The TradingEconomics pull landed at 18:14 ET on 13 August, after that vendor's daily boundary had rolled — the DXY and metals rows carry a fresh-session clock stamp and the vendor's own %Chg column prints between −0.35% and +0.08% for the entire board. That column is therefore NOT reproduced. The "24h" column below is computed by this report against the same vendor's levels published in the 12 August edition, taken at 21:04 ET — an interval of roughly 21.2 hours rather than exactly 24, which is disclosed here and worked in Data Notes. YTD figures are the vendor's own column.

The take — the dollar did not move on the largest terminal-rate move of the month, and that is the FX story. Run the inputs: the 1-year Treasury yield fell 3 bp, the 2-year and 3-year 5 bp each, the 5-year 6 bp, Investing.com's September hike probability fell 3.7 points, and the 2027 terminal contract rallied 6.5 bp to imply 3.985% against 4.050% (§8). Every one of those is dollar-negative, and the dollar index closed +0.01% over 21 hours — effectively unchanged. Compare that with Wednesday, when the index rose 0.13% on a smaller set of dollar-negative inputs and this report attributed the move to a firming terminal rate. That explanation is now unavailable, because the terminal fell. What is left is the least satisfying and probably the correct reading: the dollar is not trading U.S. rates at all this week. It is trading the absence of a credible alternative — a euro whose bourses faded an opening gain, a yen its own central bank has not defended past 159.5, and a sterling that fell on a positive investment surprise. A currency market that will not move on 6.5 bp of terminal is a currency market waiting for something else, and the something else is 8:30 on Friday.

The contrarian cross is the won for a third consecutive session, and the Taiwan dollar is now the control experiment. USD/KRW rose 0.34% — the won weakened, and by the largest margin on the board — on a session when the Kospi gained 3.6% to 6,813.34 and confirmed a technical bull market, with SK Hynix +7% and Samsung +5.7% and the index up roughly 23% from its 30 July low. Meanwhile USD/TWD fell 0.30% — the Taiwan dollar strengthened most on the board — on a day TSMC rose only 0.62%. Put the two together and the inference is much stronger than either alone: where the equity rally is largest the currency is weakest, and where the equity rally is smallest the currency is strongest. Foreign buying pulls the currency with it; domestic leveraged buying does not, because the money is already onshore. The Kospi's 23% ten-day rebound is being funded in Seoul, and the Taiwan bid is the one with an offshore footprint. For anyone running Asian tech beta, that is the difference between a flow you can follow and a flow you are the exit for.

The franc broke its own story. USD/CHF rose 0.14%, a fourth consecutive session of franc weakness — but for the first time in that run gold fell, and it fell 1.35% with a 2.26% intraday fade (§11). The configuration this report has described for a week was "buy the non-yielding real asset, sell the low-yielding haven currency", which is an inflation-hedge trade. On Thursday one leg carried on and the other reversed, on the session with the softest headline inflation print and the largest terminal-rate decline of the month. A pair that survives a soft CPI, a rising dollar and a VIX collapse but not a soft producer print has told you which of its two explanations was load-bearing — it was the inflation hedge, and it just got weaker. §12 cuts the expression accordingly.

And the one that keeps having the wrong sign: USD/CAD at −0.08% on a crude settle 2.47% lower. The loonie strengthened on the largest one-day decline in the barrel since 5 August. Over three sessions the pair has now printed −0.10% on a 1.30% crude gain, +0.15% on a flat barrel and −0.08% on a 2.47% decline — three sessions, three signs, and none of them the petro-currency sign. The Canadian dollar is trading the U.S. rate differential and nothing else; anyone still using CAD as a crude proxy is running an unhedged rates position, and this is the third consecutive edition in which that has been true.

Basis, stated before the table. The settlement column is the 13 August front-month settle taken from the Investing.com futures historical boards, which publish the official close per contract month; each level was independently corroborated against the 18:03 ET electronic quote and its change-versus-settle field, and the derivation is worked in Data Notes. The week / month / YTD columns are TradingEconomics' own spot-basis series and are NOT on the same basis as the settlement column; the two are not arithmetically reconcilable and the vendor columns are reproduced with attribution rather than recomputed. Contract months are stated in every row.

The take — the whole complex sold, and the composition says it was a dollar-and-rates session, not a supply session. WTI settled $81.21, −2.47%, having been +1.15% at $82.31 in Asian hours on genuinely bearish-for-supply news: the 800,000-barrel Russian-oil tanker aground off Oman since 30 June is now leaking onto the Omani coastline near a nature reserve, and the Gulf of Oman and Bab el-Mandeb attacks of the prior 48 hours had not been resolved (CNBC/Reuters). A barrel that cannot hold a gain on a worsening spill and unresolved shipping attacks is a barrel whose marginal buyer has left. Brent −2.23%, natural gas −2.60%, gasoline −1.30%, heating oil −1.62%, gold −1.35%, silver −1.67%, platinum −2.54%, copper −0.46% — every energy and metals row on the board was lower on the same session. That uniformity is the tell: an idiosyncratic oil story does not sell platinum, and a growth scare does not sell gold. What sells all of them together is a flat headline producer print read as disinflation, and it did.

The positioning read, and the crack is now the only thing working. Distillate is up 99.58% year to date against crude's 41.13% on the same vendor's board, and heating oil rose 9.07% on the week while crude rose 4.85% — the margin is still widening even as the barrel falls. The equity expression confirmed it in the strongest form yet: Valero +3.85%, Phillips 66 +3.12% and Marathon Petroleum +2.33% against ConocoPhillips −2.17%, EOG −1.21%, Occidental −1.45%, Exxon −0.69%, Halliburton −1.35% and Baker Hughes −1.29% — a 3.1% average for the refiners against a −1.4% average for the producers and services, on a session the barrel fell 2.47%. Seven consecutive sessions of refiners over producers, and this is the first on a materially lower crude price, which is the version that actually tests the thesis. A logistics disruption widens the crack while a falling wellhead price reduces the refiner's feedstock cost; the two effects compound rather than offset, and Thursday was the first day the market priced both at once.

Gold's reversal is the most informative price on the board, and it eliminates an explanation. The metal settled $4,407.10, −1.35%, after printing $4,508.97 intraday — a $101.87 round trip and a 2.26% fade from the high — ending a six-session run. For a week this report has argued gold was being held as an inflation hedge rather than a rate-differential trade, on the evidence that it rose through a soft CPI, a rising dollar and a VIX collapse. On Thursday the terminal rate fell 6.5 bp, which is unambiguously good for a non-yielding asset, and gold fell 1.35%. A rates explanation predicts the opposite; an inflation-hedge explanation predicts a decline on a headline print read as disinflation — and that is what happened. So the inflation-hedge reading survives and the rates reading does not, which is useful precisely because the position is now smaller and the thesis narrower. Vendor bases, because they diverge widely today: the Comex December settle is $4,407.10, TradingEconomics' spot is $4,355.05, Bloomberg's evening board reads $4,420.40 and WSJ's ticker $4,410.70. The $52.05 futures-over-spot gap is the December contract's carry; the Bloomberg and WSJ readings are live electronic prints taken at different minutes. All four are correct on their own basis and none should be compared with another without adjusting. Silver −1.67% against copper −0.46% inverts Wednesday's tell: the precious side was sold harder than the industrial side, which is a monetary-demand unwind rather than a growth signal, and Newmont −3.10% against Freeport −3.45% says the equity market split it the same way.

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

1. The rates trade — the reversal was stopped out in one session; re-enter one meeting further back

Mark on the prior edition's rates idea, first, and it is a clean loss. The book was long ZQV6 against short ZQZ7, DV01-matched one-for-one at half size, from 96.265 / 95.995 — an entry spread of 27.0 bp. Thursday: ZQV6 rose to 96.280 (+1.5 bp) while ZQZ7 rose to 96.060 (+6.5 bp), so the spread narrowed to 22.0 bp and went straight through the stated 23.0 bp invalidation. The position is stopped out at −5.0 bp in a single session, its entire life. The reasoning for the reversal was that the 2027 hawkish drift had cost the previous structure money for five sessions and the short-the-back leg was finally on the right side of it. It was on the wrong side of it within twenty-four hours, because the drift ended the same day: §8 shows the 2027 year-end cumulative-above falling 5.8 points and the terminal contract rallying 6.5 bp. Two rates structures stopped out in two sessions is a signal about the framework, not the entry — the honest conclusion is that this desk has been trading the slope of a strip whose level is doing all the work.

Modal path, base case and the tails — this is what the replacement expresses. Modal path: hold on 16 September (Investing.com 67.0%, CME 65.2%, ease 0.0%); hold on 28 October at 51.6% modal with the cumulative hike at 48.4%; one 25 bp hike delivered by 9 December, 3.75–4.00% modal at 44.4% with the cumulative at 65.6%; terminal 3.985% drawn by the 28 July 2027 contract at 96.015; year-end 2027 modal 3.75–4.00% at 35.3%, with the first non-zero easing probability anywhere on the strip at 0.4% in July 2027. Base case: goods disinflation is real and mostly crude unwinding through the chain, core services are not disinflating (+0.4% ex food, energy and trade services, portfolio management +6.5%), and the labour market has produced its first crack at 209,000 — so the committee waits, and hikes once late in the year if core services persist. Hawkish tail: a hot advance consumer print and a jump in the Michigan inflation-expectations component on Friday, which at roughly 8.6 points of probability per basis point of ZQU6 rebuilds 10–20 points of near-term hike risk and drags the whole 2027 strip back down. Dovish tail: a soft consumer print alongside a second week of claims above 205,000, which would produce the first non-zero 2026 cut probability of the year — currently 0.0% at every 2026 meeting for a nineteenth consecutive session.

Expression: long ZQZ6 (December 2026 fed funds) against short ZQZ7 (December 2027 fed funds), DV01-matched one-for-one, both legs 30-day fed funds futures at $41.67 of DV01 per basis point per contract. Entry spread 10.5 bp (96.165 less 96.060). What it says: hikes come later in 2026 than the market thought a week ago — which the data supports — but the destination is not lower, and Thursday priced the destination lower by 6.5 bp on a release whose core line accelerated fourfold. This is deliberately not a front-versus-back trade: both legs sit behind the September meeting that is now 65–67% priced for a hold, so the position is not exposed to the one outcome the market has already decided. Rationale for the wider tenor: the spread was 13.5 bp on Wednesday and 10.5 bp on Thursday; the entire 3.0 bp move came from the back leg, and the back leg is the one whose repricing contradicts the release's own core services line. Catalyst: advance retail sales 8/14 8:30 and the Michigan preliminary 8/14 10:00 (§7); the 26 August PCE deflator, which Thursday's producer print has largely written. Invalidation: the spread through 7.5 bp; or the core PCE deflator printing at or below +0.2% m/m on 26 August; or any 2026 meeting showing a non-zero cut probability, which would mean the entire framing is wrong. Sizing: small — half the size of the position it replaces, which was itself half size. Two stops in two sessions earns less risk, not a bigger swing.

2. Own the crack, not the barrel — long refiners against short crude. Upgrade.

Expression: long a basket of Valero, Phillips 66 and Marathon Petroleum against a short in September WTI, sized so the futures leg matches the basket's crude beta. Mark: the best session the idea has had, and the first genuine test of it. VLO +3.85%, PSX +3.12%, MPC +2.33% — a 3.10% average — against COP −2.17%, EOG −1.21%, OXY −1.45%, XOM −0.69%, HAL −1.35%, BKR −1.29% and a crude settle 2.47% lower. Seven consecutive sessions of refiners over producers, and this is the first on a materially lower barrel rather than a flat one. Thesis, strengthened: the prior six sessions were consistent with the crack widening and with refiners simply having more beta to a rising barrel. Thursday separated the two: the barrel fell 2.47% and the refiners rose 3.10%, which only the crack explains. The product-level confirmation is in §11 — heating oil is +99.58% year to date against crude's +41.13%, and rose 9.07% on the week against crude's 4.85%. A logistics disruption keeps the product bid while a falling wellhead cuts the feedstock cost; on Thursday those compounded. Catalyst: the EIA weekly inventory report; the next OPEC monthly; any Hormuz reopening headline. Invalidation, unchanged: a confirmed Hormuz reopening, which compresses the crack faster than the barrel and loses on both legs; or a distillate build above 3m barrels. Sizing: medium, upgraded from medium — this is now the highest-conviction idea in the section by a clear margin.

3. New — sell the rumour premium, buy the delivered miss

Expression: long an equal-weight pair of Cisco and Applied Materials against short an equal-dollar basket of the highest-multiple software names that re-rated purely on read-across — GoDaddy, CoStar, Paycom and Fair Isaac — beta-adjusted rather than dollar-matched. Thesis: Thursday priced two things in opposite directions with no cash-flow difference between them. Cisco fell 8.38% because first-quarter gross-margin guidance came in at 65–66% against a 66.1% consensus — a 60 bp miss on a midpoint. Applied Materials fell 2.48% and then about 3% more after the bell having beaten, guided to roughly $8.95bn of revenue and $3.36 of non-GAAP EPS, and raised calendar-2026 equipment growth above 30% on "unprecedented" demand. Meanwhile four software names with no news of their own rose 6.3% to 9.5% because a sponsor is reported to be in talks — talks Reuters explicitly notes may not produce a deal — for a different company. A market that pays nine percent for a rumour and takes eight percent for sixty basis points is mispricing information content, not fundamentals. Catalyst: confirmation or denial of the Silver Lake–Workday talks; the Friday open, which marks the Applied Materials reaction. Invalidation: a second confirmed sponsor bid in enterprise software, which would make the read-across a fundamental re-rating rather than a rumour premium and should retire the idea immediately; or the short basket underperforming the long by more than 4% within five sessions, which would mean the mispricing was the other way. Sizing: small — this is a mean-reversion trade against a live M&A catalyst, which is the single worst kind of short to be caught in.

4. Own downside in the consumer block, not in the index

Honest mark on the prior edition's index put spread: it lost, and it was invalidated twice on the same day. The stated invalidations were "a PPI at or below +0.1% m/m" and "the S&P closing above 7,766.01". The producer print came in at 0.0% and the index closed at 7,798.99, a record. Both triggered; the structure is a full premium write-off and should be closed rather than rolled. Two consecutive editions have now paid for index-level downside and lost, which is the market telling this desk that the index is not where the fragility is. Replacement expression: a one- to three-week put spread on a consumer-discretionary basket or the sector ETF, struck beneath Thursday's level and financed by selling the wing — not an index structure, and not an outright. Thesis: the fragility is demonstrably in the consumer names rather than the index. Tapestry beat on EPS ($1.32 vs $1.28 FactSet), beat on revenue ($1.88bn vs $1.87bn), raised its dividend 15.6% — and fell 16.49%. Ralph Lauren fell 2.26% and Ulta 3.07% in sympathy. That reaction function goes into a block of six U.S. consumer reporters in three sessions — Home Depot, Lowe's, Target, TJX, Walmart and Ross Stores, 8/18 to 8/20 (§5), with the advance consumer print at 8:30 on Friday first (§7). Catalyst: advance retail sales 8/14 8:30; Home Depot 8/18 BMO; Lowe's, Target and TJX 8/19 BMO; Walmart and Ross Stores 8/20. Invalidation: a retail-sales control group above +0.5% m/m, which would say Tapestry was idiosyncratic; or Home Depot rising more than 3% on its print. Sizing: small, defined-risk, premium-at-risk only.

5. Long gold against the Swiss franc — cut to a quarter

Expression: long spot gold funded in CHF, notional-matched, held at half size since Tuesday. Mark: the trade lost about 1.2%. Gold settled $4,407.10, −1.35%, ending a six-session run with a 2.26% fade from a $4,508.97 high; USD/CHF rose 0.14% over 21 hours, so the short-franc leg contributed +0.14% and the long-gold leg −1.35%. Neither stated invalidation triggered — the franc weakened rather than strengthened, and spot gold at $4,355.05 is well above the $4,300 trigger. Thesis, narrowed rather than broken: the position had two possible explanations and Thursday eliminated one of them. A 6.5 bp fall in the terminal rate is unambiguously good for a non-yielding asset, and gold fell anyway — so this was never a real-rate trade. What is left is the inflation-hedge reading, which correctly predicts a decline on a headline print the market read as disinflation. A thesis that survives its own test but loses money on it deserves less capital, not the same capital. Catalyst: the Michigan inflation-expectations component 8/14 10:00, which is now the cleanest single test of the remaining leg; the 26 August PCE deflator. Invalidation, tightened: spot gold below $4,300; or the franc strengthening more than 0.75% against the dollar in a session where gold is flat or lower; or a second consecutive session of gold falling while the terminal rate falls, which would mean the inflation-hedge leg has gone too. Sizing: cut from half to a quarter.

6. Protection on the CCC cohort funded in IG — hold, unchanged trigger

Expression: long CCC-exposed credit protection (or short a levered-loan/CCC-heavy vehicle) against long IG cash. Mark: the trade gave back part of Wednesday's gain. The CCC credit spread tightened 3 bp to 1,020 bp, the HY credit spread tightened 1 bp to 271 bp — a new 2026 tight — and the CCC-minus-HY differential narrowed 2 bp to 749 bp. It remains 1 bp wider on the week and 145 bp wider than January's 604 bp. Thesis: unchanged. The tail borrower's funding access is the one credit series still moving; the IG index has not moved a basis point in eight prints, which makes it an efficient funding leg and a useless signal. What weakened: the four-print widening trend is now three prints and a retracement, so the confirmation this idea needed has not arrived. Catalyst: the 26 August PCE deflator; the September quarter-end funding test, with ON RRP at a record-low $0.450bn and reserves $198.7bn off the July peak (§9b). Invalidation, unchanged: the differential back through 735 bp, which would restore the narrowing trend and retire the idea. Sizing: small, held at a half — and note the FRED series lags a business day, so the position should not be sized on a number that is 36 hours old.

7. Close the electrical-equipment-versus-semiconductor pair — the invalidation triggered

Expression: long a basket of Eaton, Emerson, Hubbell, Rockwell Automation and Johnson Controls against short an equal-dollar SOX exposure. Honest mark: the stated invalidation has now triggered and the position is closed. Thursday: the basket averaged −0.90% (ETN −1.44%, ROK −1.30%, HUBB −1.13%, JCI −0.88%, EMR +0.25%) against the SOX +0.46%, a loss of 1.36 points. Over the three sessions since inception the pair is +2.15, −2.85, −1.36 = −2.06 points net, and the SOX has outperformed the basket by 4.21 points over the last two sessions against a stated invalidation of "more than 4% over five sessions". The uncomfortable detail: the pair lost on a session when the SOX rose only 0.46% and faded 1.96% from its high — the basket underperformed a semiconductor index that itself lagged its own market by 69 bp. A hedge that loses when the thing it is short barely rises is not a hedge; it is two separate shorts. Sizing: to zero.

8. The consumer split by ticket size — hold, but it is on the ropes

Expression: long Darden, Yum! Brands and Hilton against short Home Depot, Lowe's, DR Horton, Lennar and PulteGroup, beta-adjusted. Mark: the pair lost 2.11 points. The long basket averaged −0.67% (DRI −1.84%, YUM +0.34%, HLT −0.50%) against a short basket that averaged +1.44% (HD −0.50%, LOW +1.04%, DHI +2.81%, LEN +2.75%, PHM +1.08%). The stated invalidation — the short basket outperforming the long by more than 3% before 8/18 — is not triggered at 2.11 points, but it is two-thirds of the way there in one session. What went wrong, precisely: the homebuilders rallied on a 5–6 bp fall in the belly of the curve and a 30-year mortgage at 6.74%, down 2 bp on the week (§6). Duration worked for the short leg and the restaurants gave back Wednesday's gains. What still argues for it: Tapestry −16.49% on a double beat and a dividend raise is the strongest single evidence this month that the consumer reaction function is asymmetric to the downside, and the block of six reporters starts on 8/18. Catalyst: advance retail sales 8/14 8:30; Home Depot 8/18 BMO. Invalidation, unchanged: the short basket outperforming the long by more than 3% before 8/18; or a retail-sales control group above +0.6% m/m. Sizing: small, unchanged — and it will be closed on the invalidation rather than argued with.

Volatility note. VIX 14.63 (+0.55%), range 14.39–14.80 — equity vol rose on a record close, which is small and has the wrong sign. MOVE is 72.09 on a 12 August vintage; the vendor's delayed series did not update for Thursday, so MOVE/VIX at 4.93× pairs two dates and should not be traded on (§9c). At 14.63 one-month S&P implied prices a daily move of about 0.92% against three-session realised of +0.26%, −0.06% and +0.65%, so the short-vol carry is still positive and the position is still crowded for a good reason. The specific structural argument has changed, and it is why idea 4 moved from the index to the consumer complex. Two consecutive editions bought index-level downside into an event and both were invalidated by the event itself. The dispersion is where the vol is: on a session the index rose 0.65%, Tapestry fell 16.5%, Cisco 8.4%, Corning 5.3%, GoDaddy rose 9.5% and Workday 18%. Single-name realised vol is doing something index implied vol is not, and 14.63 is the price of ignoring it.

The crowded consensuses, with the numbers that would break them.

1. "The Fed is done hiking for now, and there is no chance of a cut." The strip prices a 65.2–67.0% September hold, a 48.4% cumulative October hike, a 65.6% cumulative December hike and 0.0% probability of any 2026 cut for a nineteenth consecutive session (§8). Both tails are therefore effectively unpriced, and Thursday's release gave each of them an argument: core producer prices ex food, energy and trade services rose 0.4% with portfolio management +6.5%, which is the hawkish one, and initial claims printed 209,000 against 202,000 expected, an eight-week high, which is the dovish one. The stress test: a Michigan inflation-expectations jump on Friday morning reprices the hawkish tail at roughly 8.6 points of probability per basis point of ZQU6; a second week of claims above 205,000 on 20 August produces the first non-zero 2026 cut of the year. Cleveland's Beth Hammack said on Thursday that the committee should raise immediately. A distribution with 0.0% at one end is not a forecast; it is an absence of hedging.

2. "Credit is fine." IG at 79 bp, unmoved for eight prints and exactly flat on the year; HY at 271 bp, a new 2026 tight; $1.4tn of IG issued year to date, 9% above a record year's pace; two leveraged buyouts reported in two sessions, the second a $51bn software platform (§9). The stress test: the CCC credit spread is 1,020 bp against 885 bp in January, and the CCC-minus-HY differential is 749 bp against 604 bp. The index is priced for no defaults and the tail is priced for 145 bp more of them than it was on New Year's Eve. Those cannot both be right for much longer.

3. "The AI trade is broad." It is not, and Thursday proved it in the least ambiguous way available: the Nasdaq 100 rose 1.15% while the SOX rose 0.46% and Nvidia 0.54% — the leader underperformed its own index by 69 bp — and the SOX faded 1.96% from its intraday high. Inside the complex, memory and storage rose (WDC +7.31%, STX +4.91%, MU +4.23%) while logic, analog and networking fell (ANET −3.27%, MPWR −4.35%, MCHP −2.22%, ON −2.11%, TXN −1.14%, CSCO −8.38%). Applied Materials beat, raised equipment growth above 30%, called demand "unprecedented" and fell about 3% after the bell. The stress test: if the marginal buyer will not pay for a beat-and-raise from the largest equipment supplier, the theme is being funded by rotation rather than inflow, and rotation has an end.

4. "Software has bottomed." Nineteen names rose 2.9–9.5% on a reported and unconfirmed private-equity approach to a company that is not one of them. Reuters explicitly notes the talks are ongoing and a deal may not materialise. Against that, Bloomberg's own Markets Magazine cover this week argues the entire software business model is at risk from AI, with a companion piece on "SaaSpocalypse" risk reaching beyond private equity. The stress test: a denial, or six weeks of silence. The IGV is up about 28% in six months after falling more than 24% in the first quarter — the round trip is nearly complete and the fundamental question is unanswered.

5. "Volatility is cheap because nothing happens." VIX 14.63, and it rose on a record close; AAII bulls fell to 34.7% from 37.0% with bears at 37.9% and neutrals at 27.4%, so bears exceed bulls by 3.2 points on the day the index printed an all-time high. The stress test: implied at 14.63 prices a 0.92% daily move; Tapestry moved 16.5%, Cisco 8.4% and Workday 18% in the same session. Index vol is cheap because dispersion is doing the work, and dispersion is not a hedge.

The two-sided geopolitical tape. Escalation: the 800,000-barrel tanker aground off Oman since 30 June is now leaking onto the coastline near a nature reserve home to Arabian Sea humpback whales; the Bab el-Mandeb and Gulf of Oman attacks of the prior 48 hours remain unresolved; Iran announced it will join the BRICS New Development Bank as the war with the U.S. and Israel approaches its sixth month, with the central bank governor saying Tehran seeks "bilateral and trilateral monetary cooperation with member states" — a financial-plumbing move rather than a military one, and the first of its kind in this conflict. De-escalation: crude fell 2.47% into all of it, and Maersk raised 2026 guidance for the second time this year on second-quarter underlying EBITDA of $3.0bn against $2.04bn expected. The freight market is making money from the blockade while the oil market has stopped pricing it. The risk is that these reconverge, and there is no reason to assume they reconverge downwards.

Structural watch items. (1) The funding buffer is gone. ON RRP take-up printed $0.450bn, a record low, the second record low in two sessions; reserve balances fell $49.3bn to $2,944.1bn, the first sub-$2.95tn week since June and $198.7bn below the July peak. There is no daily stress — SOFR trades 3 bp through IORB and repo clears $1.2tn a day — but the September quarter-end will be the first one in years with no cushion. (2) Index-level supply. Anthropic could go public as soon as October at a targeted valuation above $2tn (Financial Times), which would be the largest IPO in history, surpassing SpaceX's £1.77tn June listing — SpaceX shares have since traded volatilely and only closed above their IPO price for the first time in weeks on Monday. A $2tn float lands into an equity market at a record high with reserve balances falling $50bn a week. (3) Index composition churn. Reddit joins the S&P 500 on 18 August, replacing AvalonBay, which Equity Residential is acquiring; Reddit is down 31% year to date going into the add. The rebalance flow lands on the same morning as the Home Depot print. (4) Earnings arithmetic. With more than 90% of the S&P 500 reported, second-quarter earnings growth is tracking around 50% year on year (FactSet via CNBC) — a base effect that will not repeat, into a third-quarter comparison that is materially harder.

Full Source Links, the Overnight / Asia & Europe read-through, and the complete Data Notes & Conflicts appendix are in the companion file US_CrossAsset_Daily_2026-08-13_DataNotes.txt.