| Index / Instrument | Close | Chg | %Chg | Note |
| Dow Jones Industrials | 53,885.10 | −464.02 | −0.85% | The record run ends at five sessions and 24 records (WSJ). Range 53,835.02–54,502.87; closed 1.13% below the intraday high, 41 points off the low |
| S&P 500 | 7,709.96 | −13.59 | −0.18% | Range 7,698.15–7,742.85. A second consecutive red close, and a much narrower range than Wednesday’s failed breakout |
| Nasdaq Composite | 26,348.35 | −15.09 | −0.06% | Range 26,208.43–26,499.42. The best-performing major index on a day the Dow fell 464 points |
| Nasdaq 100 | 29,373.33 | −114.46 | −0.39% | Range 29,123.38–29,569.76; 4.51% below the 30,762.20 record |
| Russell 2000 | 3,001.85 | −17.34 | −0.57% | Investing.com; WSJ marks 3,001.55 (−0.58%) — reconciled in the companion Data Notes. Closed 1.85 points above 3,000 |
| VIX | 15.15 | −0.66 | −4.17% | Range 15.11–16.03. A second consecutive 4%+ decline in equity vol on a second consecutive down day |
| PHLX Semiconductor (SOX) | 12,048.7 | +39.8 | +0.33% | Range 11,707.8–12,288.6. Traded 2.51% below Wednesday’s close at the low and closed green — the single largest reversal in the dashboard |
| UST 10Y (official par) | 4.69% | +6 bp | — | WSJ 8:11 p.m. real-time 4.680%; Bloomberg’s board 4.68%, +7 bp |
| UST 2Y (official par) | 4.25% | +7 bp | — | The front end gave back the entire ADP rally in one session (§6, §8) |
| UST 30Y (official par) | 5.22% | +5 bp | — | 20Y +4 bp to 5.22%; WSJ 30Y 5.226% |
| UST 3M (official par) | 3.90% | +1 bp | — | +8 bp on the week against +1 bp in the 10-year — the financing signal is still there (§6, §9) |
| WTI front month (Sep) | $77.29 | +2.75% | — | Settlement derived and cross-checked (companion Data Notes). CNBC: “gained nearly 3% to close above $77” |
| Brent front month (Oct) | $82.49 | +3.83% | — | CNBC: “jumped almost 4% to close above $82.” Intraday $82.72 |
| Gold (spot) | ~$4,240.7 | −0.15% | — | Touched a seven-week high near $4,271 in the morning (Reuters/CNBC) and closed red. Basis in §11 |
| Silver (spot) | ~$61.51 | −0.94% | — | Gave back a fifth of Wednesday’s +4.35% |
| Copper (Sep) | ~$6.716/lb | −0.17% | — | Printed an all-time high near $6.90/lb intraday and closed lower (CNBC) — the day’s cleanest fade |
| DXY | 99.957 | +0.28% | — | TradingEconomics evening pull; WSJ Dollar Index 96.24. Back above 99.95 |
Sources: CNBC market live blog (“Stocks end Thursday’s session lower”; “Energy and tech are the only 2 S&P 500 sectors in the green”; “Jobless claims tame; productivity better than expected, labor costs in check”; “Oil prices gain after Iran state media publishes restrictive draft plan for Strait of Hormuz”; “Copper jumps to its highest level ever”) plus the CNBC midday and after-hours movers pages; WSJ (“U.S. Stocks Down on Hormuz Deal Doubts”; “Shares of Memory Giants Drop on Soft Guidance”; “SpaceX True Believers Propel Shares Past End of Epic Lockup”) and WSJ Market Data Bonds & Rates and Currencies (8:10–8:11 p.m. ET); Bloomberg /markets and /markets/rates-bonds boards, the HONA/HYG/LQD/TTD quote pages, “Oil Extends Gains as Iran Strikes Targets in Strait of Hormuz”, “AIG Earnings Surpass Expectations After New CEO Takes Over” and “Fed’s Musalem Expresses Preference to Raise Rates”; Investing.com Major Indices, NDX, SOX, Dow-components and S&P-500-component boards (all stamped 15:59:5x); U.S. Treasury Text View; CME FedWatch numeric four-column table (data as of 6 Aug 2026 06:50:47 PM CT) and Investing.com Fed Rate Monitor (Aug 06, 2026 07:45 PM EDT); FRED ICE BofA OAS series, WRESBAL and RRPONTSYD, and NY Fed SOFR/EFFR reference rates; TradingEconomics commodities and currencies boards; Finviz Groups (Performance view, rendered); NY Fed August indicator calendar; Earnings Whispers day pages. Full categorized links are in the companion Data Notes file.
| The tape in one paragraph. The Dow fell 464 points and the Nasdaq fell fifteen. That single sentence is the session. Of 495 S&P constituents on the Investing.com board, 209 rose, 285 fell and one was unchanged, with a median move of −0.27% — a broad, shallow decline in which the index damage was concentrated almost entirely in a handful of very high-priced Dow names. Goldman Sachs −2.62% cost the price-weighted average $27.79 of index price on its own; Caterpillar −1.62% another $14.12; UnitedHealth −2.13% $8.79; Boeing −3.34% $8.02; Honeywell −2.95% $7.32; American Express −1.83% $6.39; Sherwin-Williams −1.69% $6.24 — against Microsoft +2.55%, worth $12.43 the other way. Strip the Dow out and the tape was a rotation, not a sell-off. The morning data was the pivot. Initial claims came in at 199,000 for the week ended 1 August against a 204,000 Dow Jones consensus; Q2 preliminary productivity rose 1.4% against 0.6% expected; and unit labour costs rose just 1.3% against a 2.1% estimate (BLS via CNBC). That is the exact reconciliation this report said was missing on Wednesday: a labour market that is not shedding workers and a cost structure that is not inflating. The bond market took it as a hawkish package and reversed the entire ADP move — the 2-year rose 7 bp to 4.25%, the 3-year and 5-year 7 bp, the 10-year 6 bp to 4.69% and the 30-year 5 bp to 5.22% (§6) — and St. Louis Fed President Alberto Musalem reinforced it from São Paulo, saying it is “crucial that monetary policy put a meaningful restraint on underlying inflation, rather than tolerating somewhat higher inflation today to pursue productivity growth tomorrow,” confirming he favoured a hike at the July meeting (Bloomberg, Kitco/Reuters). CME’s September hold fell from 45.6% to 45.4% and the 2027 strip repriced far harder: the June-2027 contract fell from 95.960 to 95.880, lifting the implied peak from 4.04% to 4.12% (§8). The second big story was the memory reversal. Western Digital fell 13.03% to $451.52 and SanDisk 6.74% to $1,259.54 on Wednesday night’s guidance, and yet Micron traded as low as $827.00 — 7.4% below its previous close — and closed down only 1.31% at $881.47, while the SOX printed 11,707.8 (−2.51%) and closed +0.33% at 12,048.7. A semiconductor index that opens 2.5% down after Korea’s worst session of the year and closes green is the most important single fact in this report — the Kospi fell 4.58%, SK Hynix over 10% and Samsung 6.30%. The third story was oil. Iranian state media published a draft Hormuz plan that would bar U.S. and Israeli shipping and condition transit for other nations on compensation — the opposite of the reopening the market had spent three sessions pricing. WTI settled +2.75% at $77.29 and Brent +3.83% at $82.49, energy was the best S&P sector (+1%+, CNBC) with Exxon +2.13% and Chevron +1.51%, and the whole complex has now round-tripped Monday’s collapse. Against that, the three assets that should have rallied on a Hormuz escalation did not: gold closed −0.15% after a seven-week high, silver −0.94%, and copper made an all-time high near $6.90/lb and closed lower. And VIX fell another 4.17% to 15.15. The tape’s own summary, then: the data was good, the rates market treated it as hawkish, the equity market treated it as fine, and the havens treated it as over. |
| 2 · Market Hot Spots (ranked by tradability) |
1. The semiconductor reversal is the trade of the session.
SOX low 11,707.8 (−2.51% from Wednesday’s close) → close 12,048.7 (+0.33%), a 2.91% rally off the low. The set-up could hardly have been worse: Korea’s Kospi −4.58% with SK Hynix −10%+ and Samsung −6.30%, Japan’s Tokyo Electron −5.5% and Advantest −2.34%, and Western Digital and SanDisk both broken on guidance. Micron’s low of $827.00 was 7.4% below its prior close of $893.19; it closed at $881.47, −1.31%. Fiserv traded to $49.56 against a $54.11 close — an 8.4% intraday range that ended unchanged on the day. The mechanism is worth naming: this was forced selling from Asian hours meeting a domestic bid, not a change in the fundamental view. MACOM +14.49% to $301.64, Arm +4.41%, Astera Labs +4.10% and AMD +1.50% to $489.29 led the recovery, and AMD closing +1.50% one session after −7.04% is the specific evidence. Forward catalyst: Applied Materials reports 8/13 after the close (§5), so the September quarter guides decide whether this reversal was a bottom or a bounce.
2. Hormuz went the wrong way, and only oil noticed.
Iranian state media published a draft plan banning U.S. and Israeli vessels from the strait and requiring compensation before other nations that have “harmed Iran” may transit (CNBC). Brent +3.83% to $82.49, WTI +2.75% to $77.29, and after the close Bloomberg’s wrap ran “Oil Extends Gains as Iran Strikes Targets in Strait of Hormuz.” President Trump said in the Oval Office that he thinks the war “will end pretty soon.” The tradable observation is the non-reaction elsewhere: gold closed −0.15%, silver −0.94%, the VIX fell 4.17%, and the 10-year sold off 6 bp — an escalation in the world’s most important oil chokepoint produced a bid in crude and nothing at all in the assets that hedge geopolitical risk. Either the equity and metals complexes believe the draft is an opening bid rather than a policy, or the hedges are already owned. §13 argues it is the former, and that the residual risk is now two-sided in size.
3. Honeywell Aerospace is the largest single-name dislocation of the week.
HONA −23.16% to $156.47 on 20.6m shares, a fresh 52-week low of $150.03 against a 52-week high of $297.50, market capitalisation now $49.6bn (Bloomberg). The company cut 2026 organic sales growth guidance to 4–5% from 7–9%, blaming a precision-casting shortage that is diverting scarce components to Boeing and Airbus OEM lines and away from higher-margin aftermarket; Q2 revenue of $4.52bn (+5% y/y) came with adjusted EPS of $1.87, down 32% year on year (Reuters, Benzinga). Reuters’ framing — the forecast cut leaves it “starting behind the curve” — is the right one for a first standalone quarter. The read-through matters beyond the name: the parent, Honeywell (HON), fell 2.95% in sympathy and was the fourth-largest Dow drag, and Boeing fell 3.34%. The question is whether a casting bottleneck is idiosyncratic or an industry-wide aftermarket constraint — GE Aerospace and RTX are the paired tests, and neither reports again inside this window.
4. The Dow’s 464-point loss is arithmetic, not information.
Seven high-priced constituents supplied roughly $79 of index price decline between them (Goldman, Caterpillar, UnitedHealth, Boeing, Honeywell, American Express, Sherwin-Williams), against $12.43 of offset from Microsoft. Meanwhile the equal-weight signal — the median S&P 500 move of −0.27% on 209 advancers against 285 decliners — and the Nasdaq’s −0.06% both describe a market that barely moved. The Dow’s five-session record run and 24 records of the year ended on a day when the median stock fell a quarter of a percent. Practically: do not read the Dow print as a risk signal this week; read the median and the SOX.
5. Financials were the quiet casualty, and that is a rates story with a twist.
Goldman Sachs −2.62% to $1,032.58, JPMorgan −0.82%, American Express −1.83%, Interactive Brokers −3.07%, and the KBW Nasdaq Bank Index −1.25% (WSJ) — on a day the 2-year rose 7 bp and the curve bear-flattened. Banks normally like a higher front end. What they do not like is a flatter 2s10s (44 bp, −1 bp on the day and −1 bp on the week) combined with a mortgage rate back up at 6.77% (WSJ consumer rates, +2 bp w/w) and UWM Holdings suspending its dividend and raising capital after a $452m quarterly loss, which took the stock down more than 33% (CNBC). The mortgage-origination complex breaking on the same day the 30-year fixed ticks up is the specific mechanism, and it is why D.R. Horton −3.59%, Lennar −3.48% and Builders FirstSource −4.42% were among the worst S&P performers.
6. The ad-tech complex bifurcated violently.
AppLovin −19.63% to $335.79 on Wednesday-night guidance (Q3 adjusted EBITDA of $1.71–1.74bn against a $1.75bn StreetAccount consensus), The Trade Desk −6.80% to $17.67 on 92.0m shares into its own after-close print — against a 52-week range of $16.70–$91.45, TTD is now 80.7% below its 52-week high — while Magnite rose about 18% (24/7 Wall St.). Zillow fell more than 12% after Evercore ISI cut it to In Line and slashed the target to $40 from $80, citing “a monetization strategy pivot” and a “still weak housing market”; JPMorgan cut LegalZoom to Neutral because “Google-driven traffic has slowed materially.” The common thread is not advertising demand — it is AI-mediated search destroying the traffic funnel, and it is now showing up in three separate downgrades in one session.
7. Insurance and utilities delivered the day’s cleanest earnings beats.
MetLife +3.83% to $99.95, Allstate +3.98% to $275.11, Aflac +0.88%, AIG −0.16% into its own after-close print — Bloomberg: “AIG Earnings Surpass Expectations After New CEO Takes Over.” Parker-Hannifin +7.31% to $1,069.80 and Motorola Solutions +8.20% to $474.07 were the two largest large-cap gainers; Paycom +23.70% to $216.23 was the single best S&P 500 performer. Against that, Constellation Energy fell 1.52% to $261.10 having traded $280.00 — a 6.75% high-to-close fade on the most anticipated print of the week for the AI-power complex — and Vistra closed +0.52% ahead of its own Friday-morning report. The Texas interconnection audit has now had its first earnings adjudication and the verdict was a fade, not a break.
8. Health care did nothing, and the dispersion inside it was enormous.
The Finviz group closed +0.21%, but that conceals Becton Dickinson +3.76%, Zoetis +3.87%, Cencora +2.46% and Eli Lilly +1.89% to $1,191.94 against Viatris −7.79% to $16.27, Moderna −4.28%, DaVita −4.25%, Centene −4.16%, Solventum −5.02%, Align −3.28% and UnitedHealth −2.13%. Moderna is the fade to note: the stock was up 4% pre-market on the FDA approval of its mRNA flu vaccine mFlusiva for adults 50 and over, and closed down 4.28% — an 8-point round trip on unambiguously good regulatory news.
| 3 · Sector Performance — August 6, 2026 (Finviz classification, U.S.-listed) |
| Sector | 1-Day | 1-Week | YTD |
| Energy | +1.56% | −1.82% | +30.03% |
| Healthcare | +0.21% | +0.05% | +6.15% |
| Technology | +0.04% | +5.12% | +23.10% |
| Industrials | −0.19% | +3.68% | +14.18% |
| Consumer Defensive | −0.20% | −0.55% | +8.23% |
| Financial | −0.30% | +1.22% | +8.69% |
| Utilities | −0.48% | −2.43% | +1.79% |
| Communication Services | −0.57% | +5.10% | +0.54% |
| Consumer Cyclical | −0.60% | +5.84% | −2.07% |
| Basic Materials | −0.68% | +3.73% | +13.51% |
| Real Estate | −0.91% | −1.49% | +10.38% |
Three green, eight red, and the entire table sits inside a 2.24-point range — the narrowest sector dispersion in a fortnight. Energy +1.56% is the only group that moved with conviction, and it moved for one reason: Brent +3.83% on the Iranian draft. The internals are unusually broad — Exxon +2.13% to $154.86, Chevron +1.51% to $189.23, APA +5.40%, Occidental +4.14%, Devon +2.42%, ConocoPhillips +1.50% post-print, Diamondback +1.86%, EOG +1.48%, Halliburton +2.17%, Baker Hughes +1.75%, Targa +3.12% post-print — with the refiners lagging the producers for a second session (Marathon Petroleum +0.50%, Valero +0.20%, Phillips 66 +1.47%) because a crude rally compresses the crack. Note the beta: Energy captured 41% of Brent’s move, against Wednesday’s −1.99% on a flat settlement. The sector is trading the barrel again, which it was not on Tuesday or Wednesday. Technology +0.04% is the composition story of the day: the group was flat while the SOX rose 0.33% and Microsoft rose 2.55%, because Salesforce −3.20%, Akamai −3.00%, Microchip −4.41%, Dell −5.41%, Super Micro −3.12%, Fair Isaac −5.47% and EPAM −15.29% offset them; Finviz’s Technology bucket is not a semiconductor index and on a day of violent intra-tech rotation that distinction is the whole number. Communication Services −0.57% masks Alphabet A −1.23% and Alphabet C −0.97% recovering only a quarter of Wednesday’s 4% break, with Netflix −0.70% and Meta +0.19% — the group has been red on both of the last two sessions and is now the second-worst YTD performer in the table at +0.54%. Real Estate −0.91% was the worst group on Host Hotels −7.04%, Iron Mountain −4.07%, Alexandria −3.23%, Federal Realty −3.22% and Healthpeak −3.19% — a rates-sensitive group behaving exactly as a 7 bp back-up in the 2-year would predict, and note the tower REITs went the other way: American Tower +1.86%, Crown Castle +1.46%, SBA Communications +2.44%, reversing Wednesday’s 4–5% declines. Basic Materials −0.68% is precious metals giving back, with Freeport-McMoRan −1.73% on a day copper made an all-time high, which is the sector table’s own second-order tell. Source: Finviz Groups (Performance view, rendered); Finviz buckets are not official GICS/S&P sector indices, and CNBC’s S&P 500 GICS read (energy and information technology the only two green) differs on health care because Finviz screens all U.S.-listed names.
Reconciliation. All eleven groups reconcile against Wednesday’s published YTD compounded by Thursday’s 1-day move: Basic Materials 1.1428 × 0.9932 = +13.50% vs. 13.51% shown; Communication Services 1.0109 × 0.9943 = +0.51% vs. 0.54%; Consumer Cyclical 0.9858 × 0.9940 = −2.01% vs. −2.07%; Consumer Defensive 1.0845 × 0.9980 = +8.23% vs. 8.23%; Energy 1.2802 × 1.0156 = +30.02% vs. 30.03%; Financial 1.0900 × 0.9970 = +8.67% vs. 8.69%; Healthcare 1.0586 × 1.0021 = +6.08% vs. 6.15%; Industrials 1.1453 × 0.9981 = +14.31% vs. 14.18%; Real Estate 1.1138 × 0.9909 = +10.37% vs. 10.38%; Technology 1.2307 × 1.0004 = +23.12% vs. 23.10%; Utilities 1.0229 × 0.9952 = +1.80% vs. 1.79%. Maximum deviation: 0.13 pt (Industrials); median deviation 0.02 pt; two groups exact. Energy’s deviation collapsed from Wednesday’s 0.23 pt to 0.01 pt once producers and refiners stopped moving in opposite directions. Industrials is the new outlier and it follows the Honeywell Aerospace repricing — precisely the kind of single-name capitalisation shock that makes a market-cap-weighted group aggregation drift. Flagged rather than smoothed.
| 4 · Movers & Single-Name Catalysts |
Up
• Paycom Software +23.70% to $216.23 on 3.65m shares — the best S&P 500 performer of the session. Traded $208.00–$220.48 and closed 1.9% below its high; a clean post-print gap that held all day, which is the opposite of Wednesday’s pattern.
• Motorola Solutions +8.20% to $474.07 and Parker-Hannifin +7.31% to $1,069.80 (post-print, reported 7:30 a.m.) — the two largest large-cap gainers, both on results. Parker traded $1,057.00–$1,099.94.
• Leidos +6.63% to $135.26, closing within 24 cents of its session high of $135.50 — the strongest close-to-high ratio of any large gainer.
• Fox Corporation B +6.07% to $55.37 and Fox Corporation A +5.30% to $61.79 on 8.31m shares (A) — both reported before the open.
• PTC +5.90% to $147.66, Albemarle +5.57% to $125.46, APA +5.40% to $36.53 (post-print, and the best energy name), Occidental +4.14% to $56.04 on 16.04m shares, LyondellBasell +4.08% to $61.44.
• Insulet +4.53% to $139.30, recovering just 22% of Wednesday’s −20.15% — the classic partial retrace after a guidance cut, and a useful calibration for how little dip-buying a lowered outlook attracts.
• Allstate +3.98% to $275.11 and MetLife +3.83% to $99.95 — the insurance complex was the most reliable earnings cohort of the week. Ralph Lauren +3.95% to $395.83 (reported 8:00 a.m.), having traded as low as $371.93 — a 6.4% rally off the session low.
• Zoetis +3.87% to $77.27 on 16.29m shares (7:00 a.m.) and Becton Dickinson +3.76% to $177.07 (6:30 a.m.) — the two best health-care prints of the day. T-Mobile US +3.77% to $180.00, closing 38 cents below the high.
• Uber +3.36% to $70.47 on 27.14m shares, recovering 63% of Wednesday’s −5.31% guide-down. Targa Resources +3.12% to $268.23 post-print. Paramount Skydance +3.42%.
• Exxon Mobil +2.13% to $154.86 on 8.98m shares — closed one cent below its session high of $154.87, the tightest close-to-high of any megacap. Chevron +1.51% to $189.23. Both were named by President Trump earlier in the week as “making too much money” off elevated oil prices (CNBC).
• Microsoft +2.55% to $499.87 on 28.35m shares — the largest positive contributor to the Dow and the only megacap that mattered. Bloomberg: “Microsoft Opens Fourth Data Center Region in India for Azure.” Traded $488.52–$501.56. Walt Disney +2.87% to $104.68, a second consecutive advance and now +7.7% from Tuesday’s close, closing five cents below its high.
• SpaceX +6.10% to $114.87 on 244.92m shares (not an S&P 500 constituent) — the highest-volume name on the U.S. tape, on the day its first insider lock-up expired and up to 911.5m shares became eligible for sale. WSJ: “SpaceX True Believers Propel Shares Past End of Epic Lockup.” A 6% rally into the release of 20% of eligible locked-up stock is the most surprising supply-absorption event of the quarter, and it recovered 45% of Wednesday’s −13.61%.
• MACOM Technology +14.49% to $301.64, Arm +4.41% to $286.68, Astera Labs +4.10% to $331.50, Credo Technology (all SOX constituents; MACOM and Credo are not S&P 500 members) — the semiconductor names that led the reversal. Cloudflare +17% and Magnite +18% (neither an S&P 500 constituent) on guidance and ad-tech share gains respectively.
• Analyst action: JPMorgan raised its SharkNinja price target to $207 from $170 after a top- and bottom-line beat (+14% upside to Wednesday’s close), analyst Andrea Teixeira arguing investors “are viewing SN as a high-quality, compounding growth company vs. a better-than-average consumer durables company, along with a higher multiple.” The stock hit a 52-week high intraday and closed down about 1% — a target raise sold into strength, and the third such fade in three sessions.
Down
• AppLovin −19.63% to $335.79 on 14.51m shares — the worst S&P 500 performer. Q3 adjusted EBITDA guided to $1.71–1.74bn against a $1.75bn StreetAccount consensus, with Q2 revenue also narrowly short. Traded $332.19–$352.00 and closed 4.6% below the high.
• EPAM Systems −15.29% to $93.07 and Axon Enterprise −14.28% to $522.46 — Axon’s $628.22 high against a $516.15 low is a 21.7% intraday range, the widest on the board.
• Western Digital −13.03% to $451.52 on 16.06m shares, delivering Wednesday night’s after-hours guide-down into the cash session; SanDisk −6.74% to $1,259.54 on 18.82m shares. WSJ: “Shares of Memory Giants Drop on Soft Guidance.”
• Honeywell Aerospace −23.16% to $156.47 (not an S&P 500 constituent; NASDAQ GS, $49.6bn market capitalisation), a fresh 52-week low of $150.03.
• Datadog −19.05% to $229.24, range $225.26–$243.00, +0.79% to $231.05 in after-hours trade. Q3 revenue guided to $1.135–1.145bn against a $1.11bn FactSet consensus — a beat on the quarter and only a narrow raise on the guide, which is the same pattern that broke the memory names.
• Viatris −7.79% to $16.27 on 17.14m shares (6:55 a.m.), traded $18.39 and closed 11.5% below the high — the widest high-to-close fade of any S&P 500 name.
• Host Hotels & Resorts −7.04% to $23.36 on 16.10m shares, Fair Isaac −5.47% to $1,031.89, Dell Technologies −5.41% to $437.65, Best Buy −5.30%, Solventum −5.02%, Norwegian Cruise Line −4.58%, C.H. Robinson −4.52%, Builders FirstSource −4.42%, Microchip Technology −4.41% to $74.36 (reported after the close), Moderna −4.28%, DaVita −4.25%, United Parcel Service −4.18%, Centene −4.16%, Expedia −4.10%.
• UWM Holdings −33% or more (not an S&P 500 constituent) after the largest U.S. mortgage lender suspended its dividend for the first time and raised fresh capital following a $452m quarterly net loss (CNBC). Shares were down about 40% in late-afternoon trade before closing off the low.
• Zillow −12%+ (not an S&P 500 constituent) after Evercore ISI cut it to In Line from Outperform and took the price target to $40 from $80: “With softening fundamentals, a somewhat surprising monetization pivot, and a still weak housing market, we step to the sidelines.” Peloton −14% (not an S&P 500 constituent) after guiding to falling fiscal-2027 sales despite matching on earnings and beating on revenue.
• The Trade Desk −6.80% to $17.67 on 92.05m shares, then down roughly a further 6% after the close (24/7 Wall St.). The 52-week range is $16.70–$91.45.
• Goldman Sachs −2.62% to $1,032.58, American Express −1.83%, Interactive Brokers −3.07%, JPMorgan −0.82%; KBW Nasdaq Bank Index −1.25% (WSJ). D.R. Horton −3.59%, Lennar −3.48%, Builders FirstSource −4.42% — the housing complex on a 30-year fixed back at 6.77%.
• Boeing −3.34% to $232.17, Caterpillar −1.62% to $856.96, Honeywell −2.95% to $240.80, Sherwin-Williams −1.69%, UnitedHealth −2.13% to $403.97, 3M −0.75% — the Dow’s drag list.
• Constellation Energy −1.52% to $261.10 having traded $280.00 (a 6.75% high-to-close fade), Howmet −0.58% from a $310.00 high (−6.5%), Keurig Dr Pepper −1.24% from a $33.14 high (−8.4%) — three post-print fades of 6%+ from the intraday high on the same morning, which is what a market that is selling good news looks like.
• International Paper −4.07%, Iron Mountain −4.07%, Match Group −4.04%, IFF −4.00% (giving back half of Wednesday’s +8.88%), Southwest Airlines −3.79%, Emerson −3.42%, Kimberly-Clark −3.30%, Align −3.28%, Alexandria Real Estate −3.23%, Federal Realty −3.22%, Salesforce −3.20%, Healthpeak −3.19%, Super Micro −3.12%, Kraft Heinz −3.03%, Akamai −3.00%, Kenvue −2.77%, CVS Health −2.93%.
• Megacap check: Microsoft +2.55%, Apple +0.45%, Meta +0.19%, Amazon −0.13%, Nvidia −0.22% on 98.18m shares, Tesla −0.63%, Netflix −0.70%, Broadcom +0.53%, Alphabet A −1.23%. Eight of the nine largest names moved less than 1.3% and six moved less than 0.7%. The megacap complex was inert; the damage was entirely in the Dow’s price-weighted tail.
| 5 · S&P 500 Earnings Calendar — Current & Next Week |
S&P 500 components only; times are ET. Every day page from Friday 8/7 through Friday 8/14 was re-pulled from Earnings Whispers this session and screened against the current constituent list. Re-verify times and membership against company IR before trading any date.
Mon 8/3 — completed. BMO: Loews (L), Marriott (MAR), Tyson (TSN). AMC: SBA Communications (SBAC), Vertex (VRTX), Diamondback (FANG), Palantir (PLTR), ON Semiconductor (ON), Alexandria Real Estate (ARE), Clorox (CLX), ONEOK (OKE), Williams (WMB).
Tue 8/4 — completed. BMO: ADM, Ball (BALL), DuPont (DD), Gartner (IT), Henry Schein (HSIC), Leidos (LDOS), Revvity (RVTY), Waters (WAT), Apollo (APO), Caterpillar (CAT), IDEXX (IDXX), Kimberly-Clark (KMB), Merck (MRK), Zebra (ZBRA), Aptiv (APTV), Marathon Petroleum (MPC), Pfizer (PFE), Kimco (KIM), AMETEK (AME), Broadridge (BR), Duke (DUK), McDonald’s (MCD), NRG, Rockwell (ROK), TransDigm (TDG), Cummins (CMI), FIS, PSEG (PEG), Sysco (SYY), W.W. Grainger (GWW), Progressive (PGR), Expeditors (EXPD), Pinnacle West (PNW). AMC: Amgen (AMGN), Booking (BKNG), Gilead (GILD), Wynn (WYNN), Arista (ANET), DaVita (DVA), Devon (DVN), Emerson (EMR), Jacobs (J), Match (MTCH), AMD, Celanese (CE), Healthpeak (DOC), IFF, Mosaic (MOS), Prudential (PRU), Assurant (AIZ).
Wed 8/5 — completed. BMO: EOG Resources (EOG), Cencora (COR), CVS Health (CVS), NiSource (NI), Zimmer Biomet (ZBH), Eli Lilly (LLY), Iron Mountain (IRM), Global Payments (GPN), Uber (UBER), CDW, Charles River Labs (CRL), Insulet (PODD), Kraft Heinz (KHC), Phillips 66 (PSX), Honeywell Aerospace (HONA), Walt Disney (DIS). AMC: Western Digital (WDC) — −13.03% Thursday, SanDisk (SNDK) — −6.74%, AppLovin (APP) — −19.63%, the worst S&P performer, Axon (AXON) — −14.28%, Expedia (EXPE) — −4.10%, Block (XYZ), Corpay (CPAY), DoorDash (DASH), eBay (EBAY), Realty Income (O), Solventum (SOLV) — −5.02%, McKesson (MCK), Motorola Solutions (MSI) — +8.20%, Albemarle (ALB) — +5.57%, MetLife (MET) — +3.83%, News Corp B (NWS), Occidental (OXY) — +4.14%, Texas Pacific Land (TPL), News Corp A (NWSA), CF Industries (CF), Host Hotels (HST) — −7.04%, Steris (STE), Atmos Energy (ATO), Allstate (ALL) — +3.98%.
Thu 8/6 — completed. Initial claims 199k and Q2 preliminary productivity / unit labour costs at 8:30 (§7). BMO: Targa Resources (TRGP) 6:00 — +3.12%, Becton Dickinson (BDX) 6:30 — +3.76%, Kenvue (KVUE) 6:30 — −2.77%, Molson Coors (TAP) 6:30 — +1.29%, Viatris (VTRS) 6:55 — −7.79%, ConocoPhillips (COP) 7:00 — +1.50%, Datadog (DDOG) 7:00 — −19.05%, Evergy (EVRG) 7:00 — +0.01%, Fiserv (FISV) 7:00 — unchanged after trading $49.56, Howmet (HWM) 7:00 — −0.58%, 6.5% off its high, Keurig Dr Pepper (KDP) 7:00 — −1.24%, 8.4% off its high, Warner Bros. Discovery (WBD) 7:00 — +1.77%, Zoetis (ZTS) 7:00 — +3.87%, Constellation Energy (CEG) 7:05 — −1.52%, 6.75% off its high, Parker-Hannifin (PH) 7:30 — +7.31%, Sempra (SRE) 7:55 — −0.37%, APA 8:00 — +5.40%, Fox Class B (FOX) 8:00 — +6.07%, Ralph Lauren (RL) 8:00 — +3.95%, Fox Corporation (FOXA) — +5.30%; no clock time was published, so confirm with Fox IR. AMC: Airbnb (ABNB) 4:00 — closed −0.56%, up about 7% after hours on Q2 EPS of $1.37 on $3.61bn against $1.25 on $3.58bn (LSEG), Akamai (AKAM) 4:00 — −3.00%, The Trade Desk (TTD) 4:00 — −6.80% and roughly a further 6% after hours, Aflac (AFL) 4:05 — +0.88%, Gen Digital (GEN) 4:05 — +0.50%, ResMed (RMD) 4:05 — −0.35%, Monster Beverage (MNST) 4:10, Republic Services (RSG) 4:10 — +1.24%, AIG 4:15 — −0.16%; Bloomberg: “AIG Earnings Surpass Expectations After New CEO Takes Over”, Microchip Technology (MCHP) 4:15 — −4.41%, Consolidated Edison (ED) 4:30 — +0.58%.
Fri 8/7 — payrolls day. BMO: Vistra (VST) — listed “Before Open”; the reviewed calendar published no specific time, so confirm with Vistra IR — the print that attaches an earnings number to the Texas interconnection moratorium, landing 90 minutes after the payroll. Take-Two Interactive (TTWO) 7:00, PPL 7:30. AMC: the reviewed after-close page lists three names (HE 4:05, GLBS, GLXZ), none an S&P 500 constituent — no S&P 500 after-close reporters are published for that date, unchanged for a sixth consecutive edition.
| Next week (Aug 10–14) — twelve S&P 500 reporters across five sessions |
Mon 8/10. BMO: the reviewed before-open page lists no S&P 500 reporters (50 names screened). AMC: Simon Property Group (SPG) 4:05.
Tue 8/11. BMO: Cardinal Health (CAH) 6:45, Amentum (AMTM) 8:00 (new to this pull). AMC: Lumentum (LITE) 4:00, Super Micro Computer (SMCI) 4:05.
Wed 8/12 — CPI day (8:30). BMO: Amcor (AMCR) 6:00, Trimble (TRMB) 6:55. AMC: Cisco (CSCO) 4:05, Coherent (COHR) 4:05.
Thu 8/13 — PPI day (8:30). BMO: Tapestry (TPR) 6:45. AMC: Applied Materials (AMAT) 4:00.
Fri 8/14 — retail sales, business inventories and Michigan preliminary. Neither page lists an S&P 500 reporter (twelve names screened across both).
Changes vs. the prior calendar (8/5 report). Addition: Amentum (AMTM) 8:00 on 8/11 BMO — the only genuine new S&P 500 entry across the eight day pages re-pulled, taking next week’s count from eleven to twelve. No removals: every 8/7 and next-week name carried by the prior edition re-appeared with an identical timestamp. Membership caveat, stated rather than buried: the Investing.com U.S.-500 constituent board used as this report’s screening proxy does not carry Lumentum (LITE) or Coherent (COHR). Both were carried by the prior edition and are retained here for continuity, but they are the two least certain names in next week’s list — confirm with company IR before trading either date. The same board does carry Amentum, which is why it is added. Ferguson Enterprises (FERG) 6:45 appeared again on the 8/10 before-open page and is excluded, consistent with every prior edition. Timing bucket still unpublished: VST (8/7 BMO). The forward calendar remains thin but is no longer empty — Cisco (8/12 AMC) and Applied Materials (8/13 AMC) are the only two that move an index, and both land on inflation-print days. Applied Materials on PPI day is now the single most important scheduled equity event in the next seven sessions. Conservatively excluded names are listed in the companion Data Notes.
| 6 · U.S. Treasury Yields — Official Par Curve (Treasury.gov, 3:30 PM ET) |
| Tenor | 6 Aug | 5 Aug | 1-day chg | 30 Jul | 1-week chg |
| 1 Mo | 3.80% | 3.77% | +3 bp | 3.79% | +1 bp |
| 1.5 Mo | 3.80% | 3.79% | +1 bp | 3.80% | 0 bp |
| 2 Mo | 3.84% | 3.84% | 0 bp | 3.84% | 0 bp |
| 3 Mo | 3.90% | 3.89% | +1 bp | 3.82% | +8 bp |
| 4 Mo | 3.92% | 3.91% | +1 bp | 3.92% | 0 bp |
| 6 Mo | 3.99% | 3.98% | +1 bp | 3.98% | +1 bp |
| 1 Yr | 4.06% | 4.03% | +3 bp | 4.04% | +2 bp |
| 2 Yr | 4.25% | 4.18% | +7 bp | 4.23% | +2 bp |
| 3 Yr | 4.31% | 4.24% | +7 bp | 4.30% | +1 bp |
| 5 Yr | 4.40% | 4.33% | +7 bp | 4.38% | +2 bp |
| 7 Yr | 4.53% | 4.47% | +6 bp | 4.52% | +1 bp |
| 10 Yr | 4.69% | 4.63% | +6 bp | 4.68% | +1 bp |
| 20 Yr | 5.22% | 5.18% | +4 bp | 5.22% | 0 bp |
| 30 Yr | 5.22% | 5.17% | +5 bp | 5.21% | +1 bp |
| Spread | 6 Aug | 5 Aug | 1-day | 30 Jul | 1-week |
| 2s10s | +44 bp | +45 bp | −1 bp | +45 bp | −1 bp |
| 3M10Y | +79 bp | +74 bp | +5 bp | +86 bp | −7 bp |
| 2s30s | +97 bp | +99 bp | −2 bp | +98 bp | −1 bp |
The shape: a belly-led bear-flattener, and the diagnostic is that this was a policy-path repricing with no term-premium component whatsoever. The signature is unusually clean and it is the exact mirror of Wednesday’s. The 2-year, 3-year and 5-year all rose 7 bp; the 7-year and 10-year 6 bp; the 20-year 4 bp and the 30-year 5 bp; the 2-month, 4-month and 6-month barely moved. A term-premium event steepens 2s30s; 2s30s flattened 2 bp to 97 bp. A growth scare rallies the belly; the belly led the selloff. What moves 2s/3s/5s by 7 bp and leaves 20s at 4 bp is a market marking up the expected level of the funds rate over the next eighteen to thirty months and nothing further out — precisely what a 199k claims print, a 1.4% productivity number, a 1.3% unit-labour-cost number and a St. Louis Fed president arguing for “meaningful restraint” should do. Wednesday’s front-end-only bull-flattener has been completely retraced: the 2-year is now 2 bp above where it stood a week ago, having been 4 bp below it 24 hours earlier.
The week-on-week picture is where the real information is. Over five sessions the entire coupon curve has moved between 0 and +2 bp — 2Y +2, 5Y +2, 10Y +1, 20Y 0, 30Y +1 — while the 3-month bill is +8 bp. That is not a policy curve; that is a financing curve. 3M10Y has compressed from 86 bp to 79 bp in a week, all of it from the bill end, and the 6-month is +1 bp and the 1-year +2 bp on the week, which they could not be if the market were pricing materially more near-term tightening. §9 pairs this with ON RRP take-up at $1.429bn — a new low for the series — SOFR back below IORB, and reserve balances at $2.993tn. The coupon curve has done nothing in a week. The bill curve has cheapened 8 bp. Watch the second one.
Real-time versus official par. WSJ’s 8:11 p.m. marks read 30-year 5.226%, 10-year 4.680%, 7-year 4.532%, 5-year 4.397%, 3-year 4.310%, 2-year 4.256%, 1-year 4.052%, 6-month 3.958%, 3-month bill 3.827%, 1-month bill 3.670%; Bloomberg’s board shows the U.S. 10-year at 4.68%, +7 bp. Against the official par 4.69% that is a 1.0 bp baseline-and-timing artefact — Treasury strikes par yields at 3:30 p.m. off a bid-side curve, WSJ quotes dealer marks four and a half hours later — not a disagreement about the level, and every tenor agrees on direction. The 3-month bill gap is structural: WSJ’s 3.827% is a discount-basis secondary-market quote and the official 3.90% is a coupon-equivalent par yield. One-week comparisons use 30 July, the same weekday one week prior.
| 7 · U.S. Macroeconomic Calendar |
Current week — released
| Date | Release | Actual | Consensus | Prior | Sensitivity / take |
| Mon 8/3 | ISM Manufacturing | 48.9 | 49.2 | 49.0 | Medium — prices paid fell 1.9 to 71.1 |
| Tue 8/4 | JOLTS job openings (June) | 7.36m | 7.6m | 7.54m | High — −178k; openings rate 4.4%. Trade balance −$73.3bn (BEA) |
| Wed 8/5 | ADP private payrolls (July) | +44,000 | +75,000 | +95,000 | High — weakest since January; most of the gain in health care |
| Wed 8/5 | ISM Services PMI (July) | 54.1 | 54.5 | 54.0 | High — employment 47.4 (lowest since March); prices paid 70.3 |
| Thu 8/6 | Initial jobless claims (w/e 1 Aug) | 199,000 | 204,000 | 198,000 | High — up just 1,000 and below consensus. The tie-breaker after ADP came down on the hawkish side |
| Thu 8/6 | Nonfarm productivity (Q2 prelim) | +1.4% | +0.6% | +0.8% | High — more than double consensus |
| Thu 8/6 | Unit labour costs (Q2 prelim) | +1.3% | +2.1% | — | High — the most under-watched number of the week, 0.8 pt below expectations (BLS) |
Take on the Thursday triplet — this is the reconciliation Wednesday’s report said was missing, and it resolved the wrong way for duration. Wednesday produced a stagflationary pair: a labour input that missed by 41% and a services prices index at a three-year high in its trailing average. Thursday supplied the third leg and dissolved that reading: unit labour costs at +1.3% against +2.1% say the cost push is not coming from wages, and productivity at +1.4% against +0.6% says output per hour is doing the work. Combined with claims at 199k, the package describes an economy with slowing hiring but no firing and no wage-cost pressure — which is a good economy, and therefore one that does not need a rate cut and can absorb a hike. The one caution: ADP at 44k and ISM services employment at 47.4 are not refuted by a low claims number — claims measure separations, ADP measures hires. Both can be true, and Friday’s payroll is the only release that arbitrates them.
Current week — remaining, and next week
| Date / time | Release | Sensitivity |
| Fri 8/7 08:30 | Employment Situation (July) | Very high. Consensus not verified in the reviewed sources; June nonfarm payrolls were +57k |
| Fri 8/7 11:00 | NY Fed Survey of Consumer Expectations | Medium — live given ISM prices at 70.3 and Musalem’s warning that conditions are “fertile” for expectations to unanchor |
| Tue 8/11 10:00 | NAR Existing Home Sales; Consumer Credit Panel (11:00) | Low–medium — less low with the 30-year fixed back at 6.77% and UWM suspending its dividend |
| Wed 8/12 08:30 | Consumer Price Index (July) | Very high — still the only release before 16 September able to move the September card by more than 10 points on its own |
| Thu 8/13 08:30 | Initial claims; Producer Price Index (July) | High |
| Fri 8/14 | Advance Retail Sales (08:30); Business Inventories, Michigan preliminary and SPF (10:00) | High / medium |
| Look-ahead framing — the asymmetry flipped back in a single session, and it is now the mirror of Wednesday’s. On Wednesday this report argued a weak payroll was the higher-probability surprise and the market was only two-thirds positioned for it. Thursday’s data removed the premise. With claims at 199k, productivity at 1.4% and unit labour costs at 1.3%, a payroll below 50k on Friday now has to fight three corroborating datapoints rather than joining two. The asymmetry: a payroll of 100k or better validates the hawkish package, takes the September hike back through 60% and pushes the 2-year toward 4.35%; a payroll below 50k creates the genuinely uncomfortable configuration — soft hiring, no firing, no wage costs — which the committee’s hawks will read as a supply-side improvement rather than a demand slowdown, and which therefore may not buy as much dovish repricing as Wednesday’s 5-point move suggested. The path nobody is hedged for remains a soft payroll followed by a hot CPI on 8/12, and it survived Thursday intact: unit labour costs falling does not stop tariff and energy pass-through, and Brent just went up 3.83%. |
| 8 · Fed Funds Futures & Rate Path |
Current target range: 3.50–3.75% (held 9–3 on 29 July; Hammack, Kashkari and Logan dissented for +25 bp; IORB 3.65%). Chair: Kevin Warsh. Next FOMC: Wednesday, September 16, 2026, 2:00 PM ET — 41 days away. CME FedWatch September card read live: contract ZQU6, expiry 30 Sep 2026, mid price 96.2975, prior volume 26,569, prior open interest 232,697, data as of 6 Aug 2026 06:50:47 PM CT. Investing.com Fed Rate Monitor cards stamped Aug 06, 2026 07:45 PM EDT.
CME FedWatch headline — September 16, 2026 meeting
| Target rate (bps) | NOW | 1 DAY (5 Aug) | 1 WEEK (30 Jul) | 1 MONTH (6 Jul) |
| Ease (below 350) | 0.0% | 0.0% | 0.0% | 0.0% |
| 350–375 — hold (current) | 45.4% | 45.6% | 36.6% | 42.9% |
| 375–400 (+25 bp) | 54.6% | 54.4% | 63.4% | 46.2% |
| 400–425 (+50 bp) | 0.0% | 0.0% | 0.0% | 10.8% |
| Cumulative hike (≥375) | 54.6% | 54.4% | 63.4% | 57.0% |
Column sums: 100.0 / 100.0 / 100.0 / 99.9 — the 1-MONTH column rounds one-tenth light; the other three sum exactly. Provenance of every column. CME published its complete numeric four-column table again, dated 1 DAY = 5 Aug, 1 WEEK = 30 Jul, 1 MONTH = 6 Jul, so all four columns are read directly off CME; none is carried, estimated or reconstructed. The 1-DAY column reads 45.6% / 54.4%, which is precisely what this report published as NOW on Wednesday — the third consecutive session in which a live read reconciles to CME’s subsequent settlement snapshot to the tenth of a point. Note the 1-WEEK reference date has rolled from 29 Jul to 30 Jul, so the 63.4% shown is not comparable with the 58.3% the prior edition printed for 29 July — 29 July was FOMC day and the distribution rebuilt over the following session. Standing caveat: CME’s historical columns are end-of-day settlement snapshots, so a live read after ~5:00 p.m. ET is indicative only.
Reconciling CME against Investing.com — the gap has collapsed to 0.5 points from 2.0. For September, CME puts the hold at 45.4% and the +25 bp at 54.6%; Investing.com puts them at 44.9% and 55.1%. CME’s mid price is 96.2975 and Investing.com’s future price is 96.300 — a difference of 0.25 bp of contract price, with Investing.com again quoting the higher (more dovish) price. On the ~3-points-of-headline-probability-per-basis-point sensitivity this report calibrated in late July, that price gap alone should give Investing.com a hold roughly 0.75 points above CME’s; instead it prints 0.5 points below, so the methodology residual is ~1.25 points and it is working in the opposite direction to Wednesday’s. The magnitude is identical to Wednesday’s 1.25-point residual, which is the useful part: the day-weighting effect is stable and it is the sign of the price gap that flipped. The source is unchanged — CME day-weights the September contract (roughly 14 of the month’s 30 days carry the post-meeting rate) and anchors on the realised effective rate, with EFFR fixing at 3.63%, while Investing.com maps price to outcomes on a simpler basis. The two snapshots are five minutes apart, so timing explains none of it. Both vendors agree on direction and on the sign of every change; do not mix their columns.
(1) 2026 meeting distributions — Investing.com, current / [prev-day] / [prev-week]
| Meeting | 3.50–3.75 (hold) | 3.75–4.00 (+25) | 4.00–4.25 (+50) | 4.25–4.50 (+75) | Cumulative hike |
| Sep 16, 2026 | 44.9 / [46.6] / [37.6] | 55.1 / [53.4] / [62.4] | 0.0 / [0.0] / [0.0] | 0.0 / [0.0] / [0.0] | 55.1 / [53.4] / [62.4] |
| Oct 28, 2026 | 31.5 / [33.7] / [27.2] | 52.1 / [51.5] / [55.6] | 16.5 / [14.7] / [17.2] | 0.0 / [0.0] / [0.0] | 68.6 / [66.2] / [72.8] |
| Dec 9, 2026 | 15.9 / [19.5] / [16.2] | 41.9 / [44.0] / [44.1] | 34.0 / [30.2] / [32.8] | 8.1 / [6.2] / [7.0] | 84.0 / [80.4] / [83.9] |
Row sums: September 100.0 / 100.0 / 100.0; October 100.1 / 99.9 / 100.0; December 99.9 / 99.9 / 100.1 — five of the nine columns round exactly. No easing is priced at any 2026 meeting in any column — 0.0% below 3.50% throughout, for the fourteenth consecutive session. Contract prices behind the cards: Sep 96.300, Oct 96.220, Dec 96.065 (implied 3.700%, 3.780%, 3.935%), against 96.305 / 96.225 / 96.090 on Wednesday — the October contract is 0.5 bp cheaper and the December contract 2.5 bp cheaper in a single session.
Multi-day momentum — the four-session dovish drift stopped dead, and it stopped at the back of the curve first. On CME’s cumulative basis the September hike arc reads 57.0% a month ago → 63.4% a week ago (30 Jul) → 54.4% a day ago → 54.6% now: up 0.2 points on the day, down 8.8 points on the week, down 2.4 points on the month. The front card barely moved. What moved was everything behind it: December’s ≥+50 bp cohort went from 36.4% to 42.1% in one session, a 5.7-point jump, while December’s hold fell 3.6 points from 19.5% to 15.9%. On Investing.com’s basis the cumulative hike rose monotonically with maturity: September +1.7 points, October +2.4, December +3.6. That distribution is the diagnostic: Wednesday the market moved a hike from September to October; Thursday it did not move one back — it added conviction to the hikes already sitting further out, and the further out they sat the more conviction it added. The mechanism is the data package in §7 plus Musalem, who said inflation is “well above the 2% target,” that “the balance of risks is tilted toward inflation remaining above target a year or more from now,” that it is “crucial that monetary policy put a meaningful restraint on underlying inflation, rather than tolerating somewhat higher inflation today to pursue productivity growth tomorrow,” and who confirmed he favoured a hike at the July meeting, seeing underlying inflation at 2.5–3% (Bloomberg, Kitco/Reuters). That is a fourth named hawk in three sessions after Kashkari, Cook and the three July dissents.
(2) 2027 meeting path — modal range, cumulative above/below current 3.50–3.75%
| Meeting | Modal range | Prob. | Cum. above 3.75 | Cum. below 3.50 | Price | Implied |
| Jan 27, 2027 | 3.75–4.00 | 37.6% | 86.7% | 0.0% | 96.025 | 3.98% |
| Mar 17, 2027 | 4.00–4.25 | 35.9% | 90.3% | 0.0% | 95.960 | 4.04% |
| Apr 28, 2027 | 4.00–4.25 | 35.5% | 91.1% | 0.0% | 95.920 | 4.08% |
| Jun 9, 2027 | 4.00–4.25 | 35.1% | 91.7% | 0.0% | 95.880 | 4.12% |
| Jul 28, 2027 | 4.00–4.25 | 34.8% | 91.1% | 0.2% | 95.885 | 4.12% |
| Sep 15, 2027 | 4.00–4.25 | 33.9% | 89.3% | 0.7% | 95.900 | 4.10% |
| Oct 27, 2027 | 4.00–4.25 | 32.5% | 86.6% | 1.7% | 95.910 | 4.09% |
| Dec 8, 2027 | 4.00–4.25 | 30.3% | 81.9% | 3.5% | 95.955 | 4.05% |
The terminal rate the strip draws is 4.12%, reached in June 2027 — up 8 bp in a single session and 12 bp in a week, the largest one-day move in the terminal rate since the July FOMC. The contract prices show it plainly: 95.880 at June 2027 against 95.960 on Wednesday, with every 2027 contract cheaper: Jan 96.060 → 96.025, Mar 96.005 → 95.960, Apr 95.975 → 95.920, Jul 95.960 → 95.885, Dec 96.040 → 95.955. The structural change is the modal range. On Wednesday, 4.00–4.25% was modal at exactly one meeting on the entire strip (June 2027). Tonight it is modal at seven of the eight 2027 meetings — every one from March onward. That is a wholesale re-rating of where the funds rate sits through 2027, executed in one session, and it did not come from the September card. The easing tail also thinned everywhere: 3.5% below 3.50% by December 2027 against 6.5% on Wednesday, 1.7% by October 2027 against 3.4%, and 0.7% by September 2027 against 1.7%. Cumulative tightening probability peaks at 91.7% in June 2027 against 87.8% on Wednesday: the market is now ~92% certain the Fed is higher than today at some point in the next year, and ~96.5% certain it is not lower by end-2027.
(3) Year-end probability ladders — Investing.com, current / [prev-day] / [prev-week]
| Outcome | Range | Year-end 2026 (Dec 9, 2026) | Year-end 2027 (Dec 8, 2027) |
| −100 bp or more | ≤2.75% | 0.0% / [0.0] / [0.0] | 0.0% / [0.0] / [0.0] |
| −75 bp | 2.75–3.00% | 0.0% / [0.0] / [0.0] | 0.0% / [0.0] / [0.0] |
| −50 bp | 3.00–3.25% | 0.0% / [0.0] / [0.0] | 0.3% / [0.7] / [0.5] |
| −25 bp | 3.25–3.50% | 0.0% / [0.0] / [0.0] | 3.2% / [5.6] / [4.3] |
| Hold | 3.50–3.75% | 15.9% / [19.5] / [16.2] | 14.5% / [19.8] / [16.3] |
| +25 bp | 3.75–4.00% | 41.9% / [44.0] / [44.1] | 29.5% / [32.7] / [30.1] |
| +50 bp | 4.00–4.25% | 34.0% / [30.2] / [32.8] | 30.3% / [26.8] / [28.7] |
| +75 bp | 4.25–4.50% | 8.1% / [6.2] / [7.0] | 16.5% / [11.4] / [14.9] |
| +100 bp | 4.50–4.75% | 0.0% / [0.0] / [0.0] | 4.8% / [2.5] / [4.4] |
| +125 bp | 4.75–5.00% | 0.0% / [0.0] / [0.0] | 0.7% / [0.3] / [0.7] |
| +150 bp or more | ≥5.00% | 0.0% / [0.0] / [0.0] | 0.1% / [0.0] / [0.1] |
Ladder sums: 2026 = 99.9 / 99.9 / 100.1; 2027 = 99.9 (current). 2027 cards summing to exactly 100.0: January, March, June, September, October; summing to 99.9: April, July, December. Stating it explicitly because the ladder invites the question: the probability of any 2026 rate cut is 0.0%, and the probability of a cumulative 100 bp of tightening by end-2026 is also 0.0%. The modal 2026 outcome is one hike; the modal 2027 outcome is now two.
| (4) Rate-path interpretation. One day. The September card was the least interesting part of the session: CME’s hold moved 0.2 points (45.6% → 45.4%) and Investing.com’s 1.7 points (46.6% → 44.9%). Everything happened behind it — December-2026’s ≥+50 bp cohort rose 5.7 points, the June-2027 implied peak rose 8 bp to 4.12%, and 4.00–4.25% went from modal at one 2027 meeting to modal at seven. The trigger was a package, not a print: claims at 199k against 204k, productivity at +1.4% against +0.6%, unit labour costs at +1.3% against +2.1%, and Musalem arguing for “meaningful restraint.” The transmission is worth naming precisely: unit labour costs undershooting does not make a September hike more likely — it makes the economy’s capacity to tolerate a higher rate for longer more credible, which is a 2027 story, and that is exactly where the curve moved. §6 corroborates it: 2s, 3s and 5s all +7 bp against 20s +4 bp. One week. From 30 July, the September hike is down 8.8 points on CME (63.4% → 54.6%) and 7.3 points on Investing.com’s hold-equivalent basis; October cumulative is down 4.2 points (72.8% → 68.6%); December cumulative is up 0.1 points (83.9% → 84.0%) — statistically unchanged. That is the week’s whole arc in three numbers: the market has removed a September hike, half an October hike, and nothing at all from December. The cross-asset tension has migrated: the coupon curve is flat on the week (10Y +1 bp) while the 3-month bill is +8 bp and the 2027 terminal rate is +12 bp. Those are the two live markets. The named macro hooks, in order. (i) Friday 8/7, 8:30 — July payrolls against a June print of just 57k; after Thursday it needs to be genuinely weak to reset the September card. (ii) Friday 8/7, 11:00 — NY Fed Survey of Consumer Expectations, elevated by Musalem’s explicit warning that conditions are “fertile” for expectations to unanchor. (iii) Wednesday 8/12, 8:30 — July CPI, which decides whether Thursday’s benign unit-labour-cost read survives contact with tariff and energy pass-through. (iv) Thursday 8/13 PPI. (v) Friday 8/14 retail sales. Against that stands a committee whose hawkish wing is now at least five voices — the three July dissenters plus Cook and Musalem. Base case and tails. Modal path: one 25 bp hike in 2026, still most likely in October (cumulative 68.6%) rather than September (54.6%), with year-end 2026 at 3.75–4.00% (41.9%) and a second hike now modal by mid-2027, taking the peak to 4.12% around June 2027. Upside tail: a payroll of 150k or better followed by a firm CPI takes the September hike through 65%, puts the December 4.00–4.25% bucket above 40% and sends the 30-year through 5.30%. Downside tail: a payroll below 50k — but the asymmetry has weakened, because with claims at 199k a soft payroll now reads as a hiring pause rather than a labour-market break. What is still not a tail: a 2026 cut. Zero percent at every 2026 meeting, for the fourteenth consecutive session. Practical trading implication — and an honest mark on Wednesday’s idea first. The long-September/short-October calendar spread this report put on for the payroll did not work: on Investing.com’s cumulative-hike basis the two meetings were 13.2 points apart on Wednesday (52.4% vs. 65.6%) and are 13.5 points apart tonight (55.1% vs. 68.6%), so the “hike delayed rather than cancelled” premium widened 0.3 points rather than compressing. The position is roughly flat into Friday’s print and its stated invalidation — a payroll above 150k — has not triggered, so it stands, but it has earned nothing. The expression that matches tonight’s read sits much further out: fade the 2027 re-rating. With the probability of a sub-3.50% funds rate in December 2027 down from 6.5% to 3.5% in one session and 4.00–4.25% now modal at seven of eight 2027 meetings, the strip has already done most of the hawkish work available to it — so the asymmetric trade is to buy ZQZ7 against a short in ZQV6, owning the back of the strip against the meeting where a hike is most likely to be delivered. Catalyst: the 8/7 payroll and the 8/12 CPI. Invalidation: July core CPI at or above +0.4% m/m, which validates the 2027 re-rating rather than fades it. Size on DV01, not notional; horizon two to three weeks. Not personalized investment advice. |
(a) IG and HY credit spreads
FRED publishes the ICE BofA OAS series with a one-business-day lag. The table below is as of Wednesday 5 August 2026 — it is not a same-day mark. Same-day direction is cross-checked against the cash proxies and against Bloomberg/WSJ credit coverage.
| Series | Level (5 Aug) | 1-day | 1-week | YTD | Basis |
| IG — ICE BofA US Corporate OAS (BAMLC0A0CM) | 78 bp | 0 bp | −3 bp | −1 bp | vs. 78 bp on 4 Aug, 81 bp on 29 Jul, 79 bp on 31 Dec 2025 |
| HY — ICE BofA US High Yield OAS (BAMLH0A0HYM2) | 275 bp | +2 bp | −12 bp | −6 bp | vs. 273 bp on 4 Aug, 287 bp on 29 Jul, 281 bp on 31 Dec 2025 |
| CCC & lower OAS (BAMLH0A3HYC) | 1,023 bp | +4 bp | +10 bp | +138 bp | vs. 1,019 bp on 4 Aug, 1,013 bp on 29 Jul, 885 bp on 31 Dec 2025 |
| CDX IG 5y / CDX HY 5y | see note | — | — | — | Six-step retrieval ladder worked; level not publicly obtainable |
| HYG (HY cash proxy) | $79.46 | −0.08% | — | — | 6 Aug close, 25.8m shares; prev. $79.52; range $79.41–79.51 (Bloomberg) |
| LQD (IG cash proxy) | $106.36 | −0.35% | — | — | 6 Aug close, 32.1m shares; prev. $106.74; range $106.245–106.51 (Bloomberg) |
CDX retrieval note — the six-step ladder was worked again and the level remains publicly unobtainable. (1) Bloomberg in Chrome: /markets/rates-bonds publishes the Bloomberg Fixed Income Indices (Global Aggregate 500.12, U.S. Aggregate 2,342.22, Asian-Pacific 192.74, Pan-Euro 227.75, EM USD 1,405.72) and the sovereign grid, but no CDX quote. (2) WSJ Market Data Bonds & Rates: no CDX. (3) ICE / S&P Global index pages: product and methodology pages public, levels behind entitlement. (4) FT Markets Data / Reuters credit wraps: no CDX level for 6 August; the only public reference recovered was a secondary source quoting IG “at 81 bps”, which is an ICE BofA cash figure rather than a CDX level and is therefore not usable. (5) cbonds: CDX.NA.IG 5Y and CDX.NA.HY 5Y carried but masked. (6) Cash-market proxy, labelled as such: HYG −0.08%, LQD −0.35%. The proxy read is informative tonight precisely because the two diverged: on a 6–7 bp rise in the 10-year, LQD fell 0.35% and HYG only 0.08%. That 27 bp gap is duration, not credit — LQD’s index duration is roughly double HYG’s — which means the day’s move in the cash proxies was a rates event with no credit component at all. Quoting convention: 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.
The direction cross-check. The IG credit spread at 78 bp is 1 bp inside where it started the year and 3 bp tighter on the week; the HY credit spread at 275 bp is 6 bp inside its 31 December level and 12 bp tighter on the week, having widened 2 bp on 5 August itself — the first daily widening in the HY series in four sessions, and it happened on the day the S&P failed at a record. The series still not participating is the low-quality tail: the CCC credit spread at 1,023 bp is 10 bp wider on the week and 138 bp wider year-to-date. The CCC-minus-HY differential is 748 bp against 604 bp at the start of the year, and it widened 2 bp on 5 August after compressing 4 bp on 4 August — so Wednesday’s compression was a pause, not a turn. On the U.S. Aggregate, Bloomberg’s index fell 0.30% on Thursday (2,349.35 → 2,342.22), which is the 6 bp coupon-curve selloff arriving in total-return terms.
(b) Money-market and funding plumbing
NY Fed publishes reference rates at ~8:00 a.m. ET for the prior business day; SOFR and EFFR below are as of Wednesday 5 August 2026. ON RRP is same-day (6 August).
| Metric | Latest (5 Aug) | Prior (4 Aug) | Note |
| SOFR | 3.64% | 3.66% | −2 bp |
| SOFR volume | $2,989bn | $3,036bn | −$47bn; first sub-$3tn print since 28 July |
| SOFR 1st / 25th / 75th / 99th pctile | 3.60 / 3.63 / 3.70 / 3.73 | 3.60 / 3.64 / 3.71 / 3.74 | 25th, 75th and 99th all −1 bp — the whole distribution shifted back down, exactly reversing 4 August |
| EFFR | 3.63% | 3.63% | Unchanged for a ninth consecutive session; volume $114bn vs $117bn |
| SOFR − IORB | −1 bp | +1 bp | Secured funding is back below the administered floor — the +1 bp flagged on Wednesday lasted exactly one session (IORB 3.65%) |
| SOFR − EFFR | +1 bp | +3 bp | The secured-unsecured gap narrowed 2 bp |
| ON RRP take-up (RRPONTSYD) | $1.429bn (6 Aug) | $1.650bn (5 Aug) | A new low for the series this year, for the second consecutive session |
| Reserve balances (WRESBAL, wk avg) | $2.993tn (w/e 5 Aug) | $2.985tn (w/e 29 Jul) | +$8bn — the first weekly increase since 15 July, but still −$150bn from the $3.143tn peak |
| Standing Repo Facility | — | — | No usage reported in the reviewed NY Fed material |
| Mortgage rates (WSJ, 8/5) | 6.77% 30Y fixed | 6.75% w/w | +2 bp; a fresh 52-week high. Jumbo 6.84%; 15-year 6.18% (also a 52-week high); 5/1 ARM 6.12% vs. 6.02% w/w. Fed funds target 3.50–3.75%; WSJ prime 6.75% |
The plumbing eased and the bill curve did not, and that combination is the tell. SOFR fell 2 bp to 3.64%, the 25th, 75th and 99th percentiles all fell a basis point together, and SOFR–IORB flipped straight back to −1 bp — a clean, symmetrical reversal of Tuesday’s parallel shift, which retires the idea that Wednesday’s +1 bp print was the start of something. Reserve balances also rose for the first time in three weeks, +$8bn to $2.993tn. Against that, ON RRP take-up made another new low at $1.429bn and the 3-month bill is +8 bp on the week while the whole coupon curve is +0 to +2 bp (§6). Read together: overnight funding is comfortable and term bill supply is not being absorbed. Those are different problems. The overnight market says there is no stress today; the bill curve says there is no marginal cash left to buy the next auction. None of this is stress at these levels — but the sequence (reserves $150bn off their peak, RRP effectively empty two days running, bills cheapening 8 bp in a week while the 10-year does nothing) remains the first three steps of the 2019 pattern, and the September quarter-end is the test, not August. The genuinely new datapoint is at the household end: the 30-year fixed at 6.77% and the 15-year at 6.18% are both 52-week highs, and UWM’s dividend suspension and capital raise is the first corporate casualty of it.
(c) Rates volatility and swap spreads
| Metric | Level | Change | Vintage |
| MOVE index | 76.12 | +3.45% on the day (from 73.58); −1.26% week; +15.75% month; −9.26% year | Investing.com delayed series, stamped 01:00:00 on 6 Aug — read as the most recent published print, not a 6 Aug close |
| VIX | 15.15 | −4.17%; range 15.11–16.03 (board stamped 16:14:59) | 6 Aug 2026 close |
| MOVE / VIX | 5.02× | vs. 4.91× on Wednesday and 4.88× on Tuesday | Pairs a delayed MOVE with a 6 Aug VIX — stated explicitly because the vintages differ |
| 2y / 10y / 30y swap spreads | — | No reliable data available at this time | Neither Bloomberg’s public rates page, WSJ Market Data nor the reviewed vendor boards published a swap-spread series for 6 August |
Rates vol and equity vol have decoupled again, and in the direction that matters. MOVE rose 3.45% and VIX fell 4.17% on the same session, taking the ratio to 5.02× — its highest reading of the past fortnight. That is the correct relative response: the day’s information was entirely in the rates market — a 7 bp move in the 2-year, an 8 bp move in the 2027 terminal rate, a global 10-year selloff — and essentially none of it was in equities, where the median S&P name fell 0.27%. The uncomfortable part is the absolute level of the VIX: equity vol has now been sold on three consecutive sessions in which the S&P closed red, and 15.15 is the lowest close of the fortnight. The Investing.com MOVE card again publishes an internally inconsistent “Prev. Close” of 95.74 against an intraday range of 73.58–76.12; that field is withheld here rather than quoted.
(d) Issuance, leveraged loans and private credit
• The data-centre and power financing pipeline produced two live prints. Partners Group is taking a majority stake in U.K. backup-power provider AVK with an initial investment of more than $1bn (CNBC), with AVK’s chief executive framing “speed-to-power” as “a defining opportunity for European data centre operators.” Hadrian closed a ~$1.4bn round at a $7.9bn valuation led by Baillie Gifford and J.P. Morgan Strategic Investment Group — more than quadruple its January mark. These are equity and structured-equity cheques, not bonds, and that is the point: the marginal dollar funding the power build-out is migrating out of the IG index and into private vehicles, where its spread is not observable.
• IG primary remains on pace to challenge all-time monthly issuance records, with gross IG supply projected above $2tn for 2026 against $1.7tn in 2025; six of the largest hyperscalers have issued more than $150bn of publicly traded debt in 2026.
• The offshore Japanese market is the week’s other supply story. Bloomberg: “Japan Sees Rush for Covenant Shelter as BBB-Rated Bond Sales Fly” and “Top Japan Insurer Urges Companies to Issue Bonds in Yield Hunt.” A BBB issuance rush in Japan into an insurer bid, in the same week the yen weakened 0.47% against a rising dollar, is a credible source of the marginal demand that has kept U.S. IG credit spreads at 78 bp — and it is the leg a yen reversal would remove.
• Named private-credit and single-name watch items. Bloomberg’s “Why Private Credit Got Entangled With Insurance” is directly relevant to the point above. “Trump Refunds to Cancel Offshore Wind Projects Total $4 Billion” is a live project-finance impairment. UWM Holdings’ dividend suspension and capital raise after a $452m quarterly loss is the first mortgage-finance credit event of this cycle. Morningstar LSTA leveraged loan index: No reliable data available at this time — not published in any reviewed source for 6 August. Bank CDS falls under the CDX gap above; the KBW Nasdaq Bank Index at 189.67, −1.25% (WSJ), is the only bank-risk proxy available tonight and it is an equity proxy, not a credit one.
| Take — the credit-spread-versus-equity-vol divergence, and what would break it. Thursday sharpened the divergence rather than resolving it, because for the first time the rates market and the credit market disagreed openly. The 10-year rose 6 bp, the 2027 terminal rate rose 8 bp, LQD fell 0.35% and the U.S. Aggregate fell 0.30% — and HYG fell eight hundredths of one percent. On the FRED series as of Wednesday, the IG credit spread is 78 bp, 1 bp inside its January level, and the HY credit spread is 275 bp, 6 bp inside December. Credit is absorbing a higher risk-free rate without asking for a wider spread. That is the entire trade of 2026 stated in one sentence, and it is the trade that breaks if the 2027 re-rating in §8 continues. What that means mechanically. A 12 bp increase in the expected terminal rate over one week is a direct increase in the refinancing cost of every leveraged balance sheet that terms out in 2027–2028. IG issuers can absorb it — their coupon is fixed and their cash balances earn it. The CCC cohort cannot, and it is the only part of the market pricing that: 1,023 bp, +138 bp year-to-date, against an IG credit spread 1 bp tighter. What would break it, in order of likelihood. (i) A hot CPI on 8/12 confirming the 2027 re-rating. This is now the modal risk, not the tail — Thursday moved 4.00–4.25% to modal at seven of eight 2027 meetings on the back of benign inflation data, so a genuinely firm print has room to add another 10–15 bp of terminal rate, and that is the level at which HY primary reprices. (ii) Funding. ON RRP at $1.429bn — a second consecutive record low — with the 3-month bill +8 bp on the week and reserves still $150bn off the July peak. Overnight funding eased on Wednesday, which buys time; it does not change the September quarter-end arithmetic, and IG credit spreads at 78 bp price none of it. (iii) The private-credit migration. Partners Group’s $1bn into AVK, Hadrian at a quadrupled valuation and Bloomberg’s insurance/private-credit work describe a market where the fastest-growing block of infrastructure leverage has no observable spread. The IG index cannot widen on risk it no longer holds — which means the IG credit spread is becoming a less informative signal each quarter, and that is a reason to weight the CCC series more heavily, not less. (iv) A yen or Japanese-insurer reversal removing the offshore bid. The signal to watch remains the CCC-minus-HY differential at 748 bp against 604 bp in January — and whether it takes out the 1,034 bp CCC high printed on 31 July. |
| 10 · FX — Levels and Moves |
Quote basis — read this before the table. All pairs are spot; a positive %Chg on a USD/XXX pair means the dollar strengthened, and on EUR/USD, GBP/USD and AUD/USD it means the dollar weakened. The TradingEconomics %Chg column is unusable tonight and is not reproduced: the pull was taken at 00:10 GMT on 7 August, ten minutes after the vendor’s daily boundary rolled, so its own daily change column measures ten minutes of Asian trading rather than the U.S. session. The changes below are computed instead as the 24-hour move between this evening’s TradingEconomics levels and the same vendor’s levels in the prior edition (5 August evening) — one vendor, one hour of day, a clean 24-hour basis. WSJ’s 8:10 p.m. ET marks are shown as the independent cross-check.
| Pair | Level | %Chg (24h) | WSJ 8:10 p.m. | Note |
| DXY | 99.957 | +0.28% | 99.95 | Back within a twentieth of a point of 100 — a sixth consecutive close within half a point of the figure. WSJ Dollar Index 96.24 |
| EUR/USD | 1.15222 | −0.28% | 1.1524 | The single largest G10 contributor to the dollar’s advance |
| USD/JPY | 158.445 | +0.47% | 158.44 | The yen weakened on a day the Nikkei fell 0.93% and Japanese chip names fell 5–11% — a haven that would not bid for its own market’s bad day |
| GBP/USD | 1.34524 | −0.11% | 1.3452 | Sterling gave back a tenth of a percent with the 10-year gilt +5 bp |
| USD/CHF | 0.81267 | +0.71% | 0.8127 | The franc was the weakest G10 currency of the session, and by a wide margin — the mirror image of Wednesday, when it was the strongest |
| AUD/USD | 0.70311 | −0.36% | — | Fell on a day copper made an all-time high — the clearest contrarian cross on the board |
| USD/CAD | 1.40143 | +0.03% | 1.4014 | Essentially unchanged on a day Brent rose 3.83%. Wednesday it strengthened 0.38% on a soft crude tape — the loonie has now ignored the barrel in both directions |
| USD/CNY | 6.74800 | +0.01% | 6.7491 | Unmoved; CSI 300 −0.15%. Bloomberg: “Chinese Banks Expand Direct Settlement to Aid Yuan’s Global Role” |
| USD/KRW | 1,424.71 | +0.17% | — | The won weakened just 0.17% on the Kospi’s −4.58% — a 27-to-1 ratio of equity move to currency move |
| USD/SGD | 1.28366 | +0.20% | — | DBS to a record on a record quarter; the currency did not care |
| USD/INR | 95.3400 | +0.27% | — | Still the worst major Asian currency of 2026 |
| USD/MXN | 17.2072 | −0.18% | 17.2068 | The peso strengthened against a rising dollar — one of only three currencies on the board to do so |
| USD/BRL | 5.11170 | −0.17% | — | Strengthened despite the Brazilian 10-year selling off 14 bp (Bloomberg) |
The take — the two crosses that contradict the tape, and one that confirms it. First, and most important: the Swiss franc was the weakest G10 currency on a day Iran published a plan to bar U.S. and Israeli shipping from the Strait of Hormuz. USD/CHF +0.71% is a big move for the franc, and it is precisely the wrong direction for a geopolitical escalation. Set it beside gold closing −0.15% after a seven-week high and silver −0.94% (§11): the three assets that hedge policy and geopolitical risk all fell together, on the day of an actual escalation. The honest reading is that this is partly a rates trade — a 7 bp rise in the U.S. 2-year is a straightforward negative for a zero-yielding haven, and the Swiss 10-year fell 1 bp while every other European 10-year rose — but it is also a market voting that the Hormuz draft is an opening negotiating position rather than a policy. Second, the Australian dollar fell 0.36% on the day copper printed an all-time high near $6.90/lb. A currency levered to the industrial-metals complex declining on a record copper print says the FX market thinks the copper move is a tariff-and-supply story rather than a demand story — which is exactly what CNBC’s own analysis concluded, citing the DRC export ban and Section 232. Third, and the one that confirms the equity read: USD/KRW moved 0.17% on a Kospi that fell 4.58%. Wednesday’s edition argued that foreign inflow had replaced domestic leverage as the marginal buyer of Korean equity. A 4.58% index decline that produces almost no won weakness is that thesis holding: foreign money did not leave, domestic leverage was unwound. That is a materially more benign read of Korea’s session than the index print implies, and it is consistent with the U.S. semiconductor complex closing green. USD/TWD was again not published on the TradingEconomics currencies board at the time of this pull and is omitted rather than carried forward.
| Contract / Instrument | Level | 1-day | Weekly | Monthly | YTD | Driver |
| WTI crude (Sep, settle) | $77.29 | +2.75% | −7.81% | +6.18% | +35.95% | Iranian state media published a draft Hormuz plan barring U.S. and Israeli vessels. Intraday $77.83 |
| Brent crude (Oct, settle) | $82.49 | +3.83% | −5.21% | +6.83% | +36.97% | The largest one-day gain since the Hormuz closure. Intraday $82.72 |
| Gold (spot) | ~$4,240.7 | −0.15% | +5.12% | +4.29% | −1.61% | Reached a seven-week high near $4,271 in the morning (Reuters/CNBC, “highest since June 18”) and closed red. Bloomberg: “Gold Steady as Traders Assess Impact of Hormuz Flare-Up on Rates” |
| Silver (spot) | ~$61.51 | −0.94% | +6.95% | +5.80% | −13.52% | Gave back a fifth of Wednesday’s +4.35% |
| Copper (Sep) | ~$6.716/lb | −0.17% | +4.15% | +10.71% | +17.97% | Printed an all-time high near $6.90/lb intraday and closed lower (CNBC). Catalyst: the DRC formally banned copper and cobalt concentrate exports |
| Platinum (Oct) | ~$1,735.6 | −0.75% | +9.13% | +6.42% | −15.53% | The precious metal that did not participate, for a second session |
| Natural gas (Henry Hub, Sep) | ~$2.633 | −1.31% | −4.35% | −18.20% | −28.72% | Still the worst major commodity of 2026 |
| Gasoline (RBOB, Sep) | ~$2.9559 | +4.47% | −5.21% | −4.88% | +72.53% | Outran crude — the crack widened back out and the refiners still lagged the producers (§3) |
| Heating oil (Sep) | ~$3.9323 | +4.18% | −4.02% | +7.47% | +85.28% | Distillate remains the tightest part of the barrel |
| German gas | €56.96/MWh | +6.23% | −3.67% | +19.75% | +89.86% | The largest one-day move on the board; EU TTF €58.35 |
| LNG (JKM) | $21.14/MMBtu | +1.08% | −1.10% | +30.70% | +120.09% | The best-performing major commodity of 2026 |
| Uranium | $86.50/lb | +0.17% | −0.12% | +0.93% | +5.94% | Unmoved by the AI-power tape for a fourth session |
| Naphtha | $702.68/t | +5.07% | −8.39% | +12.53% | +43.88% | The petrochemical feedstock read on the Hormuz headline |
Weekly / monthly / YTD columns are the TradingEconomics board’s own (header order verified programmatically as Price | Chg | %Chg | Weekly | Monthly | YTD | YoY | Date). Those three columns are struck off the vendor’s live 00:10 GMT 7 August level, not off Thursday’s settlement, so they run marginally rich for crude and refined products and marginally light for gold and silver; they are directional context, not settlement-basis figures. 1-day changes for crude, metals and refined products are computed against the prior edition’s settlements — each derivation is shown in the companion Data Notes.
The take — an oil rally that nothing else believed, and a copper print that everything else ignored. Positioning first. Crude entered Thursday having fallen 10.4% over two sessions and then round-tripped: Brent is $3.04 higher than 24 hours ago and still 5.21% lower than a week ago. That shape — a violent collapse on a peace headline followed by a violent recovery on the actual text — is the signature of a market that is short gamma and flat directionally, and it argues the next headline moves the price by a similar amount in whichever direction it points. The crack is the tradable expression: RBOB +4.47% and heating oil +4.18% against Brent +3.83%, so the product complex outran the barrel and the 3-2-1 crack widened — yet Marathon Petroleum closed +0.50%, Valero +0.20% and Phillips 66 +1.47%, all behind Exxon +2.13% and APA +5.40%. The equity market is paying for reserves and not for refining margin, which is the opposite of Tuesday and is the cleanest intra-energy pair on the board. Second, copper. An all-time high near $6.90/lb that closes lower is a distribution day, and the catalyst was supply-side — the DRC banning concentrate exports, on top of Section 232 tariff risk and China’s scrap crackdown — not demand. The corroborating evidence sits in FX: the Australian dollar fell 0.36% (§10) and Freeport-McMoRan fell 1.73% (§3) on the same print. When the currency, the equity and the futures curve all decline on a record spot price, the record is about the physical location of metal, not the price of copper. Third, the metals that should have worked. Gold reached its highest level since 18 June and closed −0.15%; silver fell 0.94%; platinum fell 0.75% — on the day of a Hormuz escalation, with Bloomberg’s own headline conceding traders were assessing “the Impact of Hormuz Flare-Up on Rates.” That is the correct causal chain: the 7 bp move in the 2-year outweighed the geopolitics, because a non-yielding asset is priced off the front end. Basis caveats. Every crude figure is the September WTI and October Brent contract; the TradingEconomics rolling crude series carried an 08:10 stamp on 7 August (the Friday electronic session, already reflecting the post-close Iranian strike) and its level of $78.10 is $0.81 above Thursday’s settlement — that figure is withheld rather than quoted as a close. Gold and silver are spot; copper is the September COMEX contract, and the two vendors disagreed by 2.2 cents, so the level is given to three decimals with a tilde rather than as a settlement. Bloomberg’s board showed gold at $4,308.20 in Friday-evening trade against a ~$4,240.7 spot close: that ~1.6% gap is the December futures carry, not a disagreement.
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. Own the semiconductor reversal against the Dow’s price-weighted tail.
Expression: long SOX-complex exposure (SOXX or a basket of AMD, MACOM, Arm, Astera Labs, Micron) against short a basket of the high-priced Dow drags (Goldman Sachs, Caterpillar, Sherwin-Williams, American Express), beta-neutral. Thesis: the tape’s own arithmetic says the index damage was price-weighting, not information, while the SOX absorbing a −4.58% Kospi and a −10% SK Hynix and closing +0.33% is genuine demand. Catalyst: Applied Materials, 8/13 after the close — the only semiconductor capital-equipment print in the window. Invalidation: a SOX close below 11,700, which takes out Thursday’s low and confirms the reversal was short-covering. Sizing: medium; a two-week trade with an event in the middle, so carry it in cash equity rather than options that decay across the print.
2. Fade the 2027 terminal-rate re-rating (see §8).
Expression: long ZQZ7 against short ZQV6, DV01-matched. Thesis: an 8 bp one-day and 12 bp one-week increase in the June-2027 implied peak, driven by benign inflation data, is the market extrapolating a supply-side improvement into a permanently higher policy rate. With 4.00–4.25% now modal at seven of eight 2027 meetings and the probability of a sub-3.50% funds rate at end-2027 halved to 3.5% in a session, the strip has done most of the hawkish work available to it. Catalyst: payrolls 8/7, CPI 8/12. Invalidation: July core CPI at or above +0.4% m/m. Sizing: small-to-medium; negative carry, three-week horizon.
3. Long the producers, short the refiners inside energy.
Expression: long Exxon, ConocoPhillips, APA and Occidental against short Marathon Petroleum, Valero and Phillips 66, dollar-neutral. Thesis: on a day the crack widened (RBOB +4.47% and heating oil +4.18% against Brent +3.83%), the refiners closed +0.20% to +1.47% and the producers +1.50% to +5.40%. The equity market is paying for barrels in the ground and not for margin — and if the Hormuz draft is implemented in anything like its published form, that preference gets more extreme, because it constrains transit rather than production. Catalyst: Iranian and U.S. statements on the draft; the next OPEC+ communication. Invalidation: a confirmed Hormuz reopening agreement, which collapses Brent and re-widens cracks in the refiners’ favour. Sizing: medium.
4. Sell the front of the equity-vol curve into the payroll, not through it.
Expression: short VIX futures or one-week S&P variance into Friday 8:30, covered before 8/12. Thesis: VIX at 15.15 has now been sold on three consecutive red sessions, and MOVE/VIX at 5.02× says the rates market is doing all the pricing. With the September FOMC card at 45.4/54.6 and both a payroll and a CPI inside seven sessions, the risk is not the payroll — it is CPI. Catalyst: the 8:30 print itself. Invalidation: a VIX close above 18, or a payroll below 25k. Sizing: small, and explicitly a two-day trade — carrying short vol through 8/12 is the mistake this idea is designed to avoid.
5. Buy protection on the CCC cohort, funded in IG.
Expression: long CCC-exposed credit protection (or short a levered-loan/CCC-heavy vehicle) against long IG cash. Thesis: the CCC credit spread at 1,023 bp is 138 bp wider year-to-date while the IG credit spread is 1 bp tighter, and the CCC-minus-HY differential at 748 bp against 604 bp in January is the only series that has tracked the 2027 terminal rate all year (§9). A 12 bp weekly increase in expected terminal funding cost hits the bottom of the capital structure and nowhere else. Catalyst: CPI 8/12; the September quarter-end funding test. Invalidation: the CCC-minus-HY differential compressing back through 700 bp, or a decisive dovish payroll. Sizing: medium; this is the report’s highest-conviction structural view and the one with the longest horizon.
Volatility note. VIX 15.15 (−4.17%), range 15.11–16.03, lowest close of the fortnight; MOVE 76.12 (+3.45%); ratio 5.02×, the highest in two weeks. The configuration is specific: rates vol rising, equity vol falling, and the equity index doing nothing. It is internally coherent — the day’s information genuinely was in the rates market — but it leaves equity vol priced for a market with no macro event risk inside seven sessions, when there are two. The cheapest hedge on the board is August equity vol, and the reason it is cheap is that it has been wrong for three days running.
The crowded consensuses, with the numbers to stress-test them. (1) “Credit is fine.” IG at 78 bp, 1 bp inside January. HY at 275 bp, 6 bp inside December. CCC at 1,023 bp, 138 bp wider. The consensus is right about the index and has been wrong about the tail all year. Stress test: the 2027 terminal rate rose 12 bp this week. Which cohort refinances into that? Not IG. (2) “The Fed’s next move is a coin flip and that is the whole story.” The September card at 45.4/54.6 has barely moved in two sessions. Meanwhile 4.00–4.25% went from modal at one 2027 meeting to modal at seven, in one day. The consensus is watching the wrong maturity. If the market is right about 2027, the equity market’s terminal multiple is wrong; if the equity market is right, the strip is 12 bp too high. (3) “The AI trade is broken.” The SOX closed +0.33% after trading 2.51% lower, on the day Korea fell 4.58%. Micron closed −1.31% after being down 7.4%. AMD closed +1.50% the session after −7.04%. The guide-downs are real (Western Digital −13.03%, SanDisk −6.74%, Datadog −19.05%, AppLovin −19.63%) and the bid is also real; Applied Materials on 8/13 arbitrates. (4) “Hormuz is resolving.” Iranian state media published a draft banning U.S. and Israeli vessels and conditioning transit on compensation. Bloomberg then reported Iran striking targets in the strait after the U.S. close. Brent +3.83%. Stress test: gold closed red, silver closed red, the franc was the weakest G10 currency and VIX fell 4.17% — nothing outside crude is hedging this. (5) “Low VIX means low risk.” VIX 15.15, the lowest close of the fortnight, on the third consecutive red session, with a payroll in twelve hours and a CPI in four sessions.
The two-sided geopolitical tape. Escalatory: the published Iranian draft; the reported strikes in the strait after the U.S. close; the Houthi threat to Red Sea shipping carried over from Wednesday; “Trump Refunds to Cancel Offshore Wind Projects Total $4 Billion” as a domestic-policy shock to project finance. De-escalatory: President Trump’s Oval Office comment that he thinks the war “will end pretty soon”; the Iran–Oman temporary shipping route; continued U.S. official confidence that a deal is near. The market is priced for the second set and hedged for neither, and that asymmetry is the single largest unpriced risk in this report.
Structural watch items. Funding: ON RRP at $1.429bn, a second consecutive record low; reserves at $2.993tn, $150bn off the July peak; the 3-month bill +8 bp on the week while the 10-year is +1 bp. Overnight funding normalised on Wednesday, which buys time and does not change the quarter-end arithmetic. Mortgage and housing finance: 30-year fixed at 6.77% and 15-year at 6.18%, both 52-week highs; UWM suspended its dividend and raised capital after a $452m loss; D.R. Horton −3.59%, Lennar −3.48%, Builders FirstSource −4.42%. This is the first identifiable credit transmission of the 2026 rate path into a household-facing balance sheet. The private-credit migration: Partners Group’s $1bn into AVK; Hadrian at a $7.9bn valuation; Bloomberg’s insurance/private-credit work. Leverage funding the power build-out is moving where its spread cannot be observed. Market structure: SK Hynix suffered “Another 30% Pre-Market Flash Crash on Nextrade” (Bloomberg) — a recurring execution risk in a top-five Asian semiconductor name. The Topix overhaul: Bloomberg reports over 600 companies likely to be culled — a mechanical flow event for anyone running Japan beta. The Dow’s composition: seven names supplied roughly $79 of the index’s price decline on Thursday.
What VIX is and is not pricing. At 15.15, one-month S&P implied volatility is pricing a daily move of roughly 0.95%. Over the last two sessions the S&P’s realised daily moves have been −0.17% and −0.18%, so realised is running well below implied and the option market is, narrowly, correct. What it is not pricing is the pair of events inside the next four sessions: a payroll against a June baseline of 57k, and a July CPI that is the only release before 16 September able to move the FOMC card by more than 10 points. Nor is it pricing the divergence this report has flagged for three days: MOVE/VIX at 5.02× says the rates market expects to do all the work, and the rates market has just moved the 2027 terminal rate 12 bp in a week without the equity market noticing. One of those two markets is going to be wrong, and equity vol at 15 is the cheaper way to find out. |
Full Source Links and the complete Data Notes & Conflicts section — including every vendor reconciliation, the commodity settlement derivations, the FX 24-hour basis and the earnings-membership exclusions — are in the companion text file US_CrossAsset_Daily_2026-08-06_DataNotes.txt.
|