2026-09-04 — Chip stocks decouple from broader tech, rallying as September hike odds cross 60%

• Intraday, chip stocks ran well ahead of the broader Nasdaq even as the hot jobs print pushed September rate-hike odds above 60% — a move that “should have hammered high-multiple chip stocks” on rate-sensitivity grounds but did the opposite. As of an 11:12am ET read: Intel +4% to $95.41, AMD +3% to $471.05, Nvidia +2% to $232.37, and the SOXX semiconductor ETF +3% to $517.91, against the Nasdaq-100 ETF QQQ essentially flat at +0.02%. The support cited is earnings, not rate expectations — Nvidia’s most recent quarter posted $96.22bn revenue (+105.8% y/y), AMD’s Q2 2026 revenue was $11.54bn (+50.1% y/y) — and for Intel specifically, CEO Lip-Bu Tan’s foundry turnaround pitch, though external foundry revenue is still just $293 million. This is intraday and not a session close; it extends the digest’s “equities did not sell off” read specifically to the chip cohort outperforming the rest of tech, which is a data point for the breadth question this thread exists to track (AI-specific resilience vs. broad-market resilience) as distinct from the rotation call The Great Rotation tracks. (24/7 Wall St.)

2026-09-02 — A partial morning bounce, even as yields hit a fresh multi-year high

• Stocks actually rose this morning despite the 10-year’s fresh multi-year-high print: Dow +330pts (+0.6%) to 53,097.79, S&P 500 +25.65 (+0.3%) to 7,657.12, Nasdaq +58.64 (+0.2%) to 26,158.42, as of this morning’s session. Oil pulled back slightly on the same tape (WTI’s October contract at $89.12, -1.2%) even as Brent’s own 8am-ET read kept climbing. One analyst’s read, quoted in the coverage: the modest gains suggest “investor optimism that escalation may lead to de-escalation” rather than a reversal of the underlying yield/oil story. ⚠️ This is a single-source, mid-morning read (not a close) — treat the exact index levels as provisional. (Yahoo Finance, “Stock Market Today (Sept. 2, 2026)”)

2026-09-01 — The Nvidia-earnings relief rally gives way to a fresh pullback, driven by Iran and AI-debt-linked yields

• Renewed U.S.-Iran hostilities pushed oil higher and equities lower into Monday’s close, and by Tuesday (09-01) morning, S&P 500 and Nasdaq-100 futures were down roughly 0.3%, with the Nasdaq Composite off 1.1% intraday and Nvidia, AMD and Micron each down around 2%. The stated driver pairs geopolitics with this map’s own debt-risk thesis: “soaring AI corporate debt issuance added to credit supply and lifted yields further,” per market coverage — the 10-year Treasury yield rose to 4.754%, its highest since January 2025, and the 30-year to 5.255%. This is the same AI-debt-issuance-into-yields mechanism AI Debt Gets Rated exists to track, now showing up as a same-day equity-market driver rather than a separate story. (Yahoo Finance/TheStreet market-today coverage, 2026-09-01)

• The S&P 500 closed 7,631.47 (-0.7%), the Nasdaq Composite 26,099.77 (-1.0%, the steeper of the two, “triple the S&P’s loss” per one recap), and the Dow 52,766.88 (-0.8%, -419pts) on 09-01 — the first confirmed closes behind yesterday’s intraday-only “-0.3%/-1.1%” read. Nvidia itself closed $217.44, -1.51% on the day. Coverage keeps the same causal read as yesterday’s digest: rising oil and Treasury yields on the Iran conflict weighing on risk assets, tech hit hardest. (24/7 Wall St, market-close recap, Alain Guillot, stock market recap, stockanalysis.com, NVDA history)

2026-08-31 — The oil/Fed-driven risk-off froze rather than deepened or reversed through the session’s first four hours

• From the 11:00 ET open through a ~15:03 ET check, the S&P 500, Nasdaq, and Dow moved by less than a tenth of a percentage point each (S&P -0.45%→-0.46%, Nasdaq -0.43%→-0.42%, Dow -0.62%→-0.63%) — the broad-market leg of this morning’s risk-off has neither reversed nor accelerated, it has simply held. Brent and WTI show the same pattern (Brent $90.23 vs. a morning range of $90.7-93; WTI $85.61 vs. ~$86 this morning) — still well above Friday’s ~$88/$83 close, not extending further. September Fed-hike odds are also flat on the day (CME FedWatch ~60-62%, Kalshi ~47-48%, both matching this morning’s read). The one component that kept moving through the afternoon is Treasury yields, not equities or oil — see the Treasury’s Long-End Defense entry above for the 10-year’s push past 4.75%. (Google Finance, tradingeconomics.com, CNBC 2026-08-31 coverage)

2026-08-26 (after close) — The earnings verdict this thread has been waiting for lands decisively bullish

• Nvidia’s Q2 print — the catalyst this thread has tracked for weeks against Burry’s short call, the 08-24 semiconductor selloff and the seven-day losing streak — beat and then guided far above what anyone was pricing. Revenue hit $96.2B (+18% QoQ, +106% YoY) against ~$92B consensus; management gave its first-ever year-ahead growth forecast, guiding to roughly 70% annual growth versus ~44% analyst expectations, with CEO Jensen Huang saying “demand is much greater than 70%, our supply allows us to confidently deliver 70%.” (Full compute-spend/pricing-passthrough detail already resolved on Compute Spend via upcoming.yaml’s nvidia-q2-fy2026-earnings entry — this thread’s angle is the bear-turn verdict, not the capex mechanics.) (Nvidia investor release, cited in upcoming.yaml, Fortune)
• The stock jumped 7.63% after hours the same evening, with Nasdaq 100 futures extending the move (+1.16%) into Thursday’s premarket — a sharp reversal of the week’s own losing streak and the AI-bubble-skepticism narrative this thread has carried since 08-23. S&P Global’s Melissa Otto: “I think the whole market was like, ‘Whoa, 70%.’” ⚠️ This is one earnings cycle’s verdict, not a resolution of the deeper circular-financing question this thread tracks alongside Nvidia as Lender and Circular Financing — a beat-and-raise on demand does not itself settle who is financing the buildout. (Fortune)

2026-08-25 — Nvidia’s longest losing streak since 2022 runs into its own CEO’s admission that AI hasn’t found mass-market adoption yet

• Nvidia closed Monday at $208.48, down 2.91%, its seventh straight daily decline and the longest losing streak for the stock since September 2022 — before ChatGPT existed. The stock is down roughly 8% over the run, a materially smaller drop than the 24% seven-day slide in 2022, but it is the same shape of signal this thread exists to track: a multi-day bleed concentrated in the AI-linked complex the day before the catalyst (Nvidia’s Wednesday print) that is supposed to settle the argument. (24/7 Wall St.)

• Sam Altman, on a podcast released the same day, said AI has “not had the iPhone moment” of mass-market adoption — comparing the current moment to the Palm Treo era before the iPhone changed how people use technology. Coming from OpenAI’s own CEO rather than a short-seller, this is a demand-side admission distinct from the circular-financing mechanics this map tracks elsewhere: even the industry’s most bullish voice is now dating real adoption to the future rather than the present, on the eve of the earnings print investors are treating as the AI trade’s stress test. (9to5Mac)

• Nvidia snapped a seven-day losing streak into its own print, closing up 2.2% at $213. The slide had run roughly 7.5-8%, its longest since 2022, with the S&P 500 up about 0.3% on the day. The streak is the more interesting fact than the snap: the market spent the week before this print selling the stock the whole AI trade is indexed to. (Kiplinger)

2026-08-24 — The bear turn shows up in prices: a broad semiconductor selloff two days before Nvidia’s print

• Chip stocks sold off hard through Monday’s session while the broad market barely moved, which is the signature this thread exists to catch. As of ~13:42 ET the S&P 500 was down 0.23% at 7,656.49 and the Dow was UP 0.28% at 53,423.77 — but the Nasdaq was down 0.59% at 26,026.80, and the damage was concentrated: Intel −5% to $85.98, AMD −4% to $454.36, Taiwan Semiconductor −3% to $406.40, the SOXX semiconductor ETF down 3–4% across the session, with Micron, SanDisk and Western Digital each off more than 5% later in the window. This is de-risking inside one sector, not risk-off across the market. (24/7 Wall St., 10:43 ET, Motley Fool midday wrap, 12:15 ET)
• The stated driver has two legs, and the second one is the interesting one. The first is positioning ahead of Nvidia’s Q2 FY2026 print on Wednesday 08-26 — Nvidia’s own five-quarter average day-of reaction has been a 2% DECLINE even after beats, so trimming into it is mechanical. The second is Nvidia’s 08-22 notice that AI-server prices rise more than 15% on DRAM and HBM costs — already on this map’s 08-23 page — which investors are now reading not as pricing power but as demand-softening risk. ⚠️ The price rise is not new; the reading of it is, and that is precisely what a bear turn looks like: no new fact, the same fact repriced. (24/7 Wall St., 10:43 ET)

2026-08-23 — Burry puts a fresh “lights out… 100% circular” call on the record two days before Nvidia’s print

• Michael Burry restated his Nvidia short thesis two days ahead of Wednesday’s earnings, predicting the print will be “lights out” while calling “100% of announced revenue… circular.” This thread tracks whether the AI trade’s bear case survives contact with the earnings catalyst that’s supposed to test it — a high-profile bear putting a dated, specific prediction on the record immediately before that catalyst is exactly the marker this thread exists to hold against the outcome. Three aggregators (TradingView, Yahoo Finance, Stocktwits) carry a consistent quote with same-morning timestamps (2026-08-24T07:50-08:31Z = 03:50-04:31 ET). (TradingView / Yahoo Finance / Stocktwits)
• ⚠️ Same venue caveat as the parallel entry on Nvidia as Lender: the quote is corroborated across three independent aggregators, but Burry’s original post could not be reached to confirm where he actually said it.

2026-08-19/20 — Fund managers turn more bullish even as bond yields hit 2007 highs; a leveraged AI name falls 8.3% on a day the index rallied

• The S&P 500 snapped a three-day losing streak on 08-19, closing +0.21% (7,707.98) after the Treasury’s buyback announcement (full detail on AI Debt Gets Rated) pulled long yields down — but the rally faded hard from its intraday peak (S&P +0.7% at best) and gave back most of the move by the close. Nasdaq closed +0.16% (26,331.09), Dow +0.22% (53,463.05). CoreWeave, the most leveraged AI-infrastructure name this map tracks, fell 8.3% the same session on the same rising-yield pressure the Treasury intervention was meant to relieve — a single stock moving hard against the index it sits inside, the clean breadth signal (“is it AI-specific or macro”) this thread exists to catch. (CNBC)
• BofA’s August Global Fund Manager Survey, reported 08-20, points the opposite direction from this thread’s bear-turn thesis: net equity allocation rose to 56% overweight, the highest since November 2021, on a 14th straight month of overweight positioning. AI-bubble fear — the top-named tail risk for a second straight month — actually fell to 32% of respondents from 45% in July; a disorderly bond-yield rise placed second at 27%. A record 56% of managers now expect a “no landing” economic outcome and 43% expect a “boom,” the most since February 2022. This is the strongest counter-signal on this thread’s record: the survey most attuned to AI-bubble risk got LESS worried the same week the 30-year Treasury hit a 19-year high. (Bloomberg, Investing.com)

2026-08-01 — July’s real number was the semis index, not the Nasdaq

• The iShares Semiconductor ETF (SOXX) fell 22.1% in July, its worst month since December 2002 (−23.3%), while still holding +67.7% year-to-date. (CNBC)
• The S&P 500 posted its first negative July since 2014 — outlets confirmed the direction but none gave a precise monthly percentage, so the magnitude is recorded as slight and unquantified. (Yahoo Finance)
• This map had no record of the SOXX figure at all, which is the cleanest single number on the rotation thesis it has been tracking.

2026-07-31 — July closes with the AI complex carrying the month’s losses almost alone

• July finished with the Nasdaq down 3.2%, the S&P 500 roughly flat at -0.1%, and the Dow up 0.32% — a fourth straight monthly gain for the Dow. The spread between a flat S&P and a -3.2% Nasdaq is the single cleanest number on what the month did: the damage was concentrated in the AI complex rather than distributed across the market, which is this thread’s whole question stated as one figure. Deltas computed from index levels directly and cross-checked against independently stated monthly percentages.
• Friday itself closed higher — S&P 7,489.72 (+0.70%), Nasdaq 25,373.85 (+1.00%), Dow 52,485.03 (+0.53%) — but on single-name earnings (Amazon +15.3%, Apple -7.4%) rather than a chip-led bid; the AI/semi complex was mixed, not uniformly rallying. A strong last session only partly repaired the month. (Yahoo Finance)
• Micron fell 4.3-6.0%, tied to forced-liquidation selling out of the Situational Awareness unwind (see Circular Financing) and Michael Burry adding to a short — a direct, traceable line from one leveraged fund’s failure into a specific name’s tape. ⚠ Friday closes for Nvidia, AMD and TSMC could not be verified and are deliberately omitted rather than approximated.

2026-07-30 (overnight) — The four prints landed: rewarded for returns, punished for the bet

• Microsoft: broad beat, stock +8-9% — FY2026 revenue tops $331B, Azure crosses $100B annualized, cloud growth accelerated to its fastest since 2022, Copilot paid seats hit 30M; Q4 AI-infra capex hit ~$41B, full-year capex guidance held steady. Rewarded for capex with visible cloud-revenue conversion.
• Meta: revenue +28% to $60.8B, but free cash flow collapsed 91% to $784M, stock fell ~10% — Q2 AI capex alone was $31.1B, and Meta raised the LOW end of full-year capex guidance. Punished for capex without a near-term return story.
• Qualcomm: EPS missed by $0.02 despite a revenue beat, guided to a faster Apple-modem revenue decline, announced a Sept-1 price hike tied to memory-cost inflation. Grouped by outlets with Arm as the two “outlooks rattle markets” prints of the night (Arm’s own result: beat-and-raise, stock fell further on the smartphone-royalty guidance cut — full detail in Arm Royalty Regime).
• The verdict, stated plainly: hyperscaler capex with visible cloud-revenue conversion (Microsoft) got bought; capex framed as a bet without near-term return (Meta) and legacy/smartphone-exposed chip revenue (Qualcomm, Arm) got sold. The bifurcation this thread’s “AI-specific vs. macro” watch predicted — not a clean turn, not a clean relief rally.
• Macro crosscurrent stacked on top: Fed Chair Kevin Warsh’s credibility was publicly questioned after Tuesday’s 9-3 split vote; oil back above $90 on the widening Iran conflict — both pulling on the same tape as the earnings verdicts.

2026-07-29 — FOMC holds 9-3, dissents replace unanimity, credit leg eases

• The Fed held at 3.50%-3.75% on a 9-3 vote, with three dissents for a quarter-point hike — Hammack, Kashkari and Logan — six weeks after a unanimous 12-0 decision. (Federal Reserve)
• The statement’s language barely moved from June; the committee did. The descriptive text is essentially unchanged — the split is in the votes, not the prose.
• The Dow fell roughly 800 points into the close, ahead of the decision and four major earnings prints. (TheStreet)
• Nvidia fell about 2.1% intraday on the standing $250B guarantee reports, not any new disclosure — old news repricing again.
• The bear turn’s credit leg eased today rather than extended — the pressure rotated to Oracle instead of Nvidia — and tonight’s four prints are the real test of whether this was a pause or a turn.

2026-07-28 (overnight) — Korea rallies on the print, then breaks again

• The second circuit breaker in two sessions — this time after good news. The 07-29 KST session opened +1.1%, ran to +2.76% (6,190.04) on SK Hynix earnings relief, then reversed the whole way to an intraday −12.63% (5,262.77), tripping a Level-1 breaker at 12:32pm KST, and closed 5,663.24, ~−6%. SK Hynix itself: +3.74% → −9.61% (1,401,000 won); Samsung +4.55% → −5.23%. (Korea Herald)
• That is the sharper version of this thread’s thesis. A market that cannot hold a bid for three hours on the best operating quarter in its lead company’s history is not repricing earnings — it is repricing the durability of the earnings. The bear turn’s Asia leg no longer needs bad news to fire.
• The divergence says it is a chip event, not regional risk-off: Nikkei −1.49% (61,434.19) but Hang Seng +1.8% (25,762.80). US tape the same day: AMD ~−10%, Micron −8.85% ($820.53), SOX −6% in a fourth straight losing session, Nasdaq-100 9.7% off its record.
• ⚠ NVDA date corrected: the −5% was 07-27, not 07-28 (07-28 closed +0.25%, $197.01, after a ~2% intraday dip). The US leg of this thread is a credit/equity divergence, not a synchronized break — see Nvidia as Lender. (stockanalysis.com)
• Counterpoint kept on the record: Q3 DRAM contracts reportedly settling 20-30% higher this month, and analyst framing that “the panic appears to be indiscriminate.” The tape and the contracts are pointing opposite ways.

2026-07-28 (close) — the loud-no faded by the bell

• NVDA closed +0.3% ($197.01) — the −5% morning move on the record CDS widening fully reversed intraday. The day’s bear signal was a credit-market event (CDS record hold); the equity leg didn’t ratify it. Divergence worth watching into tomorrow’s FOMC + four prints. Elsewhere at the close: INTC −5.9% ($86.30) · MU −8.9% (CXMT read-through day 2) · CRWV −4.9% · SPCX +2.6% ($116.41). (Yahoo chart data, 16:00 ET)

2026-07-28 — FOMC eve: mixed tape, rate odds still elevated

• Going into tomorrow’s 2pm ET decision, the tape is mixed rather than uniformly bearish: CoreWeave −30% in a month, SpaceX touching a fresh post-IPO low (~$110.85, Morgan Stanley pricing its AI unit at zero), and Nvidia’s own CDS still wide — but Intel just posted its best quarter in 15 years (+25% revenue) and oil is now falling sharply (Brent −4.8% to ~$84 on the US-Iran pause), which cuts against one of the week’s own drivers of Fed hike odds. Breadth question from 07-27 stays open: still AI-specific + macro-crosswind, not one clean story.
• Fed hike odds sit ~25-35% across trackers (CME FedWatch ~34.7% as of 07-27; a same-day FinanceFeeds read has it at 25%) — up from ~24% on 07-21, but consensus still points to a HOLD tomorrow with odds shifting to a September hike instead (77% per one prediction-market read). See the FOMC brief for detail.

2026-07-28 — Circuit breakers

• Korea’s session closed −10–11% with circuit breakers triggered — SK Hynix below $140 into its own earnings; NVDA −5% + record CDS move the same day. Both legs of the turn fired at once, the day before FOMC.

2026-07-28 — The Seoul leg

• The turn went global: KOSPI −10.84% (“Black Tuesday”), SK Hynix −14.65% — triggered by CXMT’s debut rather than US capex anxiety, but the same repricing impulse. The bear-turn now has US (index bear territory, NVDA CDS) and Asia (Seoul) legs in the same week.

2026-07-27 — Opened

• The AI-heavy index slid into bear-market territory (IBD); ASML plunged on China news; NVDA −4% on its own guarantee. Running tally into the turn: Mag7 −$797B (07-23), SpaceX ~50% off high, soft AI-debt demand (07-24), hike odds ~40% into Wednesday’s FOMC. Ben: distinct from circular-financing — “it’s its own thing.”