Nasdaq Enters Correction at 24,630, S&P 500 Sinks to 7,447, Dow Sheds 2.4% as NVDA Drops 8%

Wednesday’s relief rally lasted exactly one session. By 11:46 a.m. ET on Thursday, November 5, the S&P 500 had dropped 241.19 points, or 3.14%, to 7,447.43. The Nasdaq Composite had shed 1,105.93 points, or 4.30%, to 24,630.58, and the Dow Jones Industrial Average was down 1,258.07 points, or 2.36%, at 52,098.05. The Russell 2000 slid 2.92% to 2,773.64. Every one of those numbers wipes out more than Wednesday’s gain, and in the Nasdaq’s case it wipes it out three times over.

Put the two sessions side by side and the reversal is brutal. On Wednesday the Dow ripped 1,569.40 points, or 3%, to close at 53,356.12, its best single day since April 2025. The S&P 500 added 119.70 points, or 1.6%, to 7,688.62, and the Nasdaq climbed 364.50 points, or 1.4%, to 25,736.51. Thursday’s S&P loss is double Wednesday’s gain. The Nasdaq loss is three times larger. The Dow has handed back 80% of its record-setting pop in a single morning.

The Nasdaq is the headline. The tech-heavy benchmark is now more than 10% below its October 6 record close, which puts it in correction territory for the second time in 2026. It closed in a correction in March, rallied to records in June, nearly cracked again in late July, then made fresh highs in September and October. Thursday’s slide undoes that last leg. At the intraday low the Composite traded near 24,400, a drop of more than 5.1%. The Nasdaq-100 fared worse, falling as much as 5.7% to 26,826, a loss of 1,611 points.

The S&P 500 is on pace for its worst session since the April 2025 tariff crash. It traded as low as 7,425 and now sits 4.8% below its 52-week high of 7,818.9. Its year-to-date gain, which stood at 11.25% at the open, has been trimmed to under 9%. The Nasdaq’s year-to-date gain has been cut to 6%.

The fear gauge confirms the damage. The CBOE Volatility Index, which fell 20.3% to 16.61 on Wednesday, jumped as high as 23.46, a 41.6% spike. That is the largest one-day volatility jump in months and pushes the VIX back above the 20 line that separates calm tape from stressed tape.

The thesis of the day is simple. This is not a recession trade and it is not a rate scare. It is the market repricing the entire AI capital-spending supply chain for political risk after Tuesday’s Democratic sweep of Congress and several key governorships. Chips, memory, servers, neoclouds and power suppliers got hit. Energy, healthcare and blue-chip defensives held up. The index damage is a function of where the market’s weight sits, and right now that weight sits squarely in the names being sold.

From Green Futures to a Red Tape: How the Session Unraveled

The morning did not start like a crash. Before the open, Dow futures were up 329 points to 53,832 and looked set to build on Wednesday’s 3% jump. S&P 500 futures were up 13 points to 7,733. Only the Nasdaq was flashing a warning, with futures down 132 points, or 0.46%, to 28,434. Semiconductors were the soft spot. The VanEck Semiconductor ETF (SMH) fell 2.25% premarket to $493, pushing further below its 200-day moving average of $521. Wednesday had marked its first break of that average since May 2025.

The opening bell delivered a split tape. The Nasdaq fell 1.99% to 25,225.30 in the first minutes, the S&P 500 lost 0.95% to 7,615.65, and the Dow added 0.11% to 53,412.20. Small caps were still bid, with the Russell 2000 up 1.58% to 2,856.95. For a few minutes the rotation story held: sell chips, buy everything else.

That story died fast. By 9:57 a.m. ET the Dow had flipped to a 351.87-point loss at 53,004.25. The S&P 500 was down 127.74 points, or 1.66%, at 7,560.88, and the Nasdaq had dropped 795.08 points, or 3.09%, to 24,941.44. Decliners outnumbered advancers 1.36-to-1 on the NYSE and 2.17-to-1 on the Nasdaq. The Nasdaq Composite was printing 247 new 52-week lows against 132 new highs.

At 10:10 a.m. the S&P 500 was down 2.03% at 7,532.64 and the Nasdaq down 3.61% at 24,806.33. The Dow was still holding a sub-1% loss at 52,856.67. Twenty-five minutes later the Nasdaq was down 4% and the S&P 500 down 2.3%, and the chip complex had already erased $2 trillion in market value from its Wednesday close. By 10:58 a.m. the Nasdaq was down 4.9%.

The 11 a.m. print was the ugliest so far. The S&P 500 was off 3.2%, the Dow was down 1,146 points, or 2.1%, and the Nasdaq Composite was 5% lower. Gains in Chevron and Johnson & Johnson were not enough to keep the 30-stock Dow from following tech down.

The shape of the day tells you who was selling. A slow bleed from the open with no meaningful bounce, accelerating through the first 90 minutes, is institutional distribution, not a retail panic. Funds that bought chips into Wednesday’s rally used the open to get out. Anyone who bought the 9:30 a.m. dip in semiconductors was underwater by 10 a.m. and deeper underwater by 11 a.m. There was no reflex rally at 10:30 a.m., which is the time of day dip buyers usually show up. The bid faded every time it tried to form.

The leveraged ETFs show the intensity. The Direxion Daily Semiconductor Bull 3X (SOXL) traded 325 million shares and was down 26.8% at $48.82. Its inverse twin, SOXS, was up 27.3% at $74.25 on 96.8 million shares.

The Semiconductor Index Drops 11%: Anatomy of a Sector Wipeout

The PHLX Semiconductor Index (SOX) fell 11% intraday, its steepest slide since March 2020. It had already dropped 3% on Wednesday to its lowest level since April, even as the Nasdaq rose. Two days of selling have now stripped double-digit percentage value out of a group that was still up 70% for the year through October.

The breadth inside the chip group was total. AMD fell as much as 15.9%. Micron dropped between 12% and 16.8% as the morning wore on. Nvidia was down 8.3% at 11 a.m., Intel lost between 4.4% and 8%, Broadcom shed 6% to $304.51, and Marvell Technology dropped more than 8%. Qualcomm, which reported Wednesday night, slid 8.9% to $157.82. Arm Holdings fell 7% to $177.80. There was no chip stock with any size that held green.

Three forces collided. The first was political. Democrats won control of both the House and the Senate on Tuesday, and on Wednesday Representative Yassamin Ansari called for a pause on advanced AI development until regulations are in place. That pushed AI legislation from campaign rhetoric into the realm of committee hearings and subpoenas for the next two years.

The second force was state-level. Democrats won governorships in Texas, Ohio and Georgia on platforms that included restricting data-center construction. Gina Hinojosa, who defeated Greg Abbott in Texas, has promised an immediate moratorium on new data centers and the elimination of tax breaks for developers. Texas alone hosts 335 data centers with another 247 in planning, and accounts for a quarter of planned U.S. AI infrastructure. States run by Democrats who support moratoriums will control 45% of America’s planned data-center capacity starting in January.

The third force was earnings. Qualcomm’s handset revenue collapsed 26%, and Arm’s royalty revenue grew just 8%. Both pointed to a smartphone market being squeezed by soaring memory costs, which undercuts the one part of the chip universe that was supposed to be independent of the data-center cycle.

Put those together and investors had no place to hide inside semiconductors. Data-center names were hit by politics. Handset names were hit by earnings. Memory names were hit by both, since high memory prices are crushing phone demand while data-center demand for memory now looks politically capped.

The selloff spread globally before the U.S. even opened. Samsung Electronics dropped 7.5% and SK Hynix fell 9.5% in Seoul. Japan’s Nikkei 225 and the Kospi both declined sharply, and AI shares in Europe retreated. Chinese chipmakers were the lone exception, rising on Beijing’s latest market-support measures.

AMD: A Record Quarter Gets the Worst Reaction in the Group

Advanced Micro Devices is the clearest example of fundamentals being overruled by politics. The stock opened at $452, then broke straight down to $412 by mid-morning, a loss of 15% to 16% on the session. It now trades 37% below its 52-week high of $658.52, and had already shed 22% over the prior month before Thursday’s open.

What makes the move remarkable is what AMD reported two days earlier. Third-quarter adjusted earnings came in at $2.09 per share, beating the $1.95 consensus and rising 74% year over year. Data-center revenue nearly doubled, up 97%. Management guided fourth-quarter revenue to $17 billion. CEO Lisa Su then went further, saying AMD expects revenue to grow more than 80% in 2027, more than twice the long-term growth model the company laid out a year ago. She also said demand outstrips supply.

That report sent AMD higher in after-hours trading on November 3. By Wednesday morning the stock was flat in premarket. By Thursday it was the single worst large-cap performer in the semiconductor space, falling harder than Nvidia, Intel or Broadcom.

The logic of the selling is straightforward. AMD’s 80% growth forecast for 2027 depends on hyperscalers and AI labs building data centers at the current pace or faster. If Texas, Ohio and Georgia slow permits, if Congress launches hearings on power costs, and if local opposition keeps killing projects, the 2027 number becomes a ceiling rather than a floor. In the second quarter of 2026 alone, 45 data-center projects worth $68 billion were disrupted by local opposition. That was before Tuesday’s vote.

Positioning also explains why AMD fell harder than its peers. A stock that runs on a forward growth story carries more downside when the market questions the forward part. Nvidia trades at a lower multiple of forward earnings, and Intel’s story rests more on foundry turnaround than AI share gains. AMD is the purest high-multiple bet on the next leg of AI capex, and the market is cutting that bet first.

There is a valuation question inside the selloff, too. At $412, AMD’s market value has shrunk by tens of billions of dollars in two days despite raising its long-term forecast. Either the 2027 forecast is wrong or the stock is mispriced. The tape on Thursday was not interested in resolving that debate. Sellers wanted out of the highest-beta AI exposure regardless of the earnings line, and AMD gave them the deepest liquidity pool to do it.

For Friday and beyond, the $412 low is the first level that matters. A close below it would mark a fresh post-earnings low and extend a one-month drawdown already exceeding 35% from the highs.

Nvidia and the Weight Problem: Why One Stock Drags the Whole S&P 500

Nvidia was by far the heaviest weight on the S&P 500 on Thursday. The stock fell 8.3% by 11 a.m. ET, putting it at $183.52 against a prior close of $200.13. Earlier in the session it traded at $187.24, down 6.44%, on volume of 140.4 million shares, the second-most-active name on the exchange after the leveraged chip ETF. That follows a 3.2% drop on Wednesday, so the two-day loss now exceeds 11%.

Size is the whole story. Nvidia’s market value is so large that a 1% move in the stock moves the S&P 500 more than a 1% move in dozens of smaller companies combined. When it falls 8%, the index loses meaningful points before the other 499 names get a vote. That is why Thursday’s S&P 500 drop looks catastrophic even though two out of every five stocks in the index were higher at 11 a.m.

Nvidia also had its own problem on top of the political one. Around 600,000 of its H200 chips are stuck in Hong Kong awaiting import approval from China. That shipment backlog raises fresh doubts about how much of China’s AI market Nvidia can actually serve under the current export regime, even though Chinese companies still need its compute. For a stock that already carried data-center political risk into the session, the China headline removed the last argument for buying the dip.

The selling in Nvidia has a technical flavor as well. Below $190 the stock has broken through a band where buyers stepped in several times in October. The next visible floor sits at the $180 round number, which also lines up with where the stock traded before its last leg higher. A clean break of $180 would put the 52-week range in play in a way that has not been tested since the spring.

The irony is that the fundamentals have not changed in a day. Nvidia still sells everything it can make, and the backlog from hyperscalers is not going to disappear because of a midterm election. What changed is the duration of the growth story. If data-center construction slows in states that account for 45% of planned capacity, the order book still gets filled, but it gets filled over a longer timeline. For a stock priced on the next three years of growth, stretching the timeline is the same as cutting the estimate.

Wednesday told a different story about which parts of tech the market trusts. Microsoft surged 6% to a record, Amazon gained more than 6%, and Apple climbed above a $5 trillion market cap. Investors were betting that a slowdown in construction protects the value of existing cloud capacity and forces Big Tech to rein in capex. That is a bullish read for hyperscaler margins and a bearish read for the people selling them chips. Thursday’s tape played out the second half of that trade in full.

Qualcomm and Arm: The Handset Leg of the Selloff

Qualcomm opened the floodgates before the opening bell. The stock reported fiscal fourth-quarter results after Wednesday’s close, beat on the headline, and still opened down 13% at $150.57. It traded as low as $149.79 before bouncing to $157.82 by 10:21 a.m. ET, a loss of 8.9% on the day against a prior close of $173.20. Some prints had it down 11% at $154 in early trading. The decline erases Wednesday’s 5.3% gain, which came on a new collaboration announcement with SpaceX.

The quarter itself was fine. Adjusted revenue came in at $10.44 billion, ahead of the $10.11 billion consensus, though still down 7% from $11.27 billion a year earlier. Adjusted earnings of $2.17 per share edged past the $2.16 estimate. Automotive revenue jumped 89%, the strongest growth line in the company. Qualcomm also renewed long-term licensing deals with Apple, Huawei and Transsion and announced a data-center collaboration with Amazon.

The problem was everything after the quarter. Handset chip revenue, Qualcomm’s largest business, fell 26% and missed estimates. Rising memory prices have forced smartphone makers to raise prices, and higher prices are killing unit demand. Qualcomm’s first-quarter outlook also assumes no further Apple product sales, as Apple completes its switch to in-house modems. Fiscal first-quarter adjusted earnings guidance of $2.15 to $2.35 per share brackets a midpoint of $2.25, just below the $2.26 to $2.27 consensus. The GAAP guide was far worse: $1.03 to $1.23 per share against a $1.67 consensus. Revenue guidance of $10.4 billion to $11.2 billion did little to offset it. Management also warned that rising input costs will squeeze chip margins through the April-to-September 2027 period.

Arm Holdings sang the same song in a different key. Fiscal second-quarter revenue rose 27.1% to a record $1.44 billion, beating the $1.38 billion estimate. Adjusted earnings of $0.52 per share topped the $0.47 consensus. But royalty revenue grew only 8% to $670 million, short of forecasts, and GAAP earnings of $0.12 per share missed. CEO Rene Haas said the company could do more if it had more supply, and flagged higher memory prices as a drag on smartphone sales.

Arm had already fallen 5.28% to $191.80 during Wednesday’s regular session before the report. It slipped another 3.05% after hours and was down 7% at $177.80 on Thursday. The stock has now lost more than a third of its value in a month.

The handset leg matters because it removes the diversification argument. A chip investor who wanted to avoid data-center political risk could rotate into mobile and auto chips. Qualcomm and Arm just showed that the mobile side is being crushed by the very memory shortage that the AI buildout created. Skyworks Solutions slid 2.9% to $82.92 in sympathy.

Memory and Storage: SanDisk, Micron, Western Digital and Seagate Get Hit Hardest

The heaviest percentage losses on Thursday came from memory and storage. SanDisk plunged 22.2%. Western Digital dropped 18.1%, Seagate fell 17.1%, and Micron slid as much as 16.8% after losing 12% in the first hour. Astera Labs, which makes connectivity chips for AI servers, lost 19.5%. These stocks had already dropped on Wednesday, when Micron fell 5% and SanDisk fell 7% on the first day of post-election selling.

The reason memory got hit so hard is that it had become the most crowded AI trade of the second half. Memory prices have soared all year as AI servers absorbed high-bandwidth DRAM and enterprise NAND, and storage makers rode that pricing to record margins. The trade rested on one assumption: data-center demand keeps rising faster than supply. Tuesday’s election put that assumption on a political clock for the first time.

There is a second-order problem. High memory prices are the reason smartphone demand is falling, as Qualcomm and Arm both made clear in their reports. That creates a feedback loop. If data-center demand slows, memory prices fall. If memory prices stay high, handset demand keeps sinking. Either way, someone in the chip supply chain loses. Memory makers had been the winners of that tension. Now investors are pricing the possibility that the winning side of the loop is about to weaken.

The scale of the moves speaks to positioning more than fundamentals. A 22% one-day drop in SanDisk does not reflect a 22% change in its earnings outlook. It reflects a crowded long trade being unwound by funds that all bought the same thesis. When every hedge fund owns the memory cycle and the cycle gets questioned on the same morning, the exits get narrow and the price gaps lower.

Volume confirmed it. Micron and Western Digital both traded multiples of their average daily volume in the first two hours, and the declines accelerated as the morning wore on rather than fading. A one-sided tape like that tends to keep falling until forced sellers are finished.

There is still a floor under the group if the data-center slowdown proves smaller than feared. Analysts covering the sector broadly doubt that any legislation can meaningfully stall construction already under way, and existing hyperscaler commitments run into 2027 and 2028. A Democratic Congress makes the buildout more expensive and slower. It does not make it disappear.

But Thursday was not a day for nuance. The memory and storage group got smoked because it had the most to lose if the AI buildout loses momentum, and because it was the easiest place to sell size. Until the political picture clarifies, the group is likely to trade as a high-beta proxy for every headline about data-center permits, moratoriums and congressional hearings.

Data-Center Politics: Neoclouds, Servers and Power Suppliers All Crack

The neoclouds were already bleeding on Wednesday. They hemorrhaged on Thursday. CoreWeave fell between 9.3% and 14.9% through the morning, after losing 10% on Wednesday. Nebius Group dropped between 13.4% and 19.7%. Data-center operators IREN, Hut 8 and Applied Digital fell between 6% and 11%. AI server maker Dell dropped 10.5%. Power suppliers tied to the data-center boom, including GE Vernova and Constellation Energy, also traded lower.

These are the companies with the most direct exposure to the pace of physical construction. CoreWeave rents GPU capacity through data-center partners that include facilities in Texas. IREN owns its infrastructure outright. Applied Digital builds and leases. Every one of them depends on new capacity coming online on schedule, and every one of them now faces a political environment in which that schedule can be delayed by a governor’s order.

The Texas result was the biggest single shock. Markets were already pricing a Democratic House and a possible Democratic Senate. A Democrat winning the Texas governor’s mansion was not in the price. Hinojosa campaigned on extending a moratorium Abbott had already imposed, until lawmakers pass stronger AI regulations. In Ohio, Governor-elect Amy Acton called for a conditional moratorium unless developers cover all utility costs, use union labor and meet environmental standards. In Georgia, Keisha Lance Bottoms also campaigned on restricting development.

Public opinion is not on the industry’s side. Exit polling showed 76% of voters nationwide oppose construction of a new data center in their own community. That is a bipartisan number, and it tells you the moratorium push will not stop at the three states that flipped.

Financing makes the neoclouds especially vulnerable. CoreWeave recently projected losses steep enough that its bond yields surged to 15%. Oracle just priced a combined $20 billion stock and debt offering to fund its own buildout. AI infrastructure stocks have already lost half to 60% of their value since summer highs as investors questioned spending levels, leverage and returns. A slower permitting environment raises the cost of capital at exactly the moment these companies need the most of it.

Elon Musk even mused publicly about moving compute into orbit, which says something about how seriously the industry is taking the ground-level opposition. SpaceX, which is building multiple gigawatts of capacity including projects in Texas, slid 3.81% to $135.92 on Wednesday.

The market is not betting the buildout ends. It is betting the buildout gets more expensive and slower, with more hearings, more lawsuits and more local vetoes. For companies priced on the assumption that capacity doubles every year, a slower path is enough to cut valuations in half.

The Winners: AppLovin Rips 30%, Energy and Defensives Hold the Line

Not everything got sold. AppLovin was the standout of the session, ripping 30.46% to $354.54 against a prior close of $271.77. The mobile app marketing company delivered third-quarter revenue of $2.18 billion, beating the $2.07 billion estimate, and earnings of $4.30 per share, ahead of the $4.04 consensus. The bigger surprise was guidance. AppLovin projected fourth-quarter revenue between $2.37 billion and $2.40 billion, well above the $2.25 billion the market expected. The results eased concerns about the company’s growth trajectory after a volatile stretch for the stock.

AppLovin’s surge shows that the market is still willing to pay for growth when it is tied to software margins rather than physical construction. An ad-tech platform does not need a new data center in Texas to grow, and investors drew that distinction instantly.

Other gainers came from corners with no AI capex exposure. Corpay jumped 8.4%. Tapestry gained 7.9%. Motorola Solutions climbed 7.6%. CF Industries rose 7.1%, riding the bounce in energy and fertilizer pricing. Wendy’s rose 1.5% premarket after rejecting a $1.4 billion takeover proposal from Nelson Peltz’s Trian Partners, calling the offer a substantial undervaluation of its standalone prospects in a statement from its independent committee. The company also unveiled a multiyear turnaround plan called Dave’s Way. Shake Shack rose 3.8% after adjusted earnings of 40 cents per share topped the 35-cent estimate.

Energy was the best-performing sector, which makes sense given Wednesday’s carnage in the group. The Energy Select Sector SPDR (XLE) lost 4.8% on Wednesday as crude fell more than 8%, and Thursday’s oil bounce brought buyers back. Chevron was among the Dow’s best performers. Johnson & Johnson also gained, giving the Dow a pair of defensive anchors that kept it from falling as hard as the Nasdaq.

Healthcare and defensives benefited from the same logic. A Democratic Congress is expected to target healthcare pricing, but the sector’s earnings do not depend on data-center construction. On a day when the market is selling anything tied to AI capex, a pharmaceutical name with steady cash flow looks like shelter.

The gainers list is short, and it is not big enough to offset the losses in chips. Five of the 11 S&P 500 sectors were lower early in the session, with technology and consumer discretionary leading the losses. That means six sectors were flat or higher. But those six sectors combined carry less index weight than technology alone. The rotation was real. It was simply overwhelmed by the math of market-cap weighting.

Earnings Casualties Outside Chips: Snap, Veracyte, Dutch Bros and HubSpot

Chips dominated the headlines, but the earnings tape outside semiconductors had its own wreckage. Veracyte collapsed 54% after a third-quarter beat on revenue and earnings was overshadowed by a cut to 2026 revenue guidance on slower Decipher test volume growth and preliminary 2027 guidance far below consensus. A one-day loss of more than half the company’s value is the kind of move that only happens when the forward story breaks completely.

Snap slid 12% after missing third-quarter revenue estimates. The social media company has repeatedly struggled to convert user growth into advertising dollars, and another top-line miss put the stock back in the penalty box. The contrast with AppLovin is stark: one ad-tech business beat and raised and ripped 30%, the other missed and got smoked.

Dutch Bros fell 15%. Its third-quarter results were broadly in line, including system-wide same-store sales growth of 4.4%, but revised 2026 guidance pointed to the low end of its prior same-store sales range and EBITDA below its prior range. For a growth restaurant chain priced on expansion, landing at the bottom of guidance is a miss.

HubSpot slid between 2% and 9% through the morning after a beat on results was offset by weak guidance. The company effectively guided fourth-quarter constant-currency growth down to 12% from 14%, and said it expects net revenue retention to decline two points year over year in 2026 to 101%. A falling retention rate in a subscription business is a signal that existing customers are spending less, and the market treats it harshly.

Arista Networks, which reported earlier in the week, guided fourth-quarter earnings of $1.26 to $1.30 per share on revenue of $3.9 billion to $4.0 billion. As a networking supplier to the same AI data centers now under political scrutiny, it traded with the broader AI hardware group.

Earlier in the week, Fidelity National Information Services tumbled 24% after slashing its full-year earnings outlook, while Block rallied almost 16% after a beat-and-raise, with adjusted earnings of $1.13 per share on revenue of $6.72 billion.

The pattern across these names is consistent. The market is not punishing misses on the quarter. It is punishing weak forward guidance with extreme prejudice, and rewarding strong guidance just as aggressively. Veracyte and Qualcomm both beat the quarter and got crushed on the outlook. AppLovin beat and raised and got rewarded.

The earnings calendar does not let up. Airbnb, Cloudflare, Lumentum, Monster Beverage, Duke Energy, EOG Resources, Cigna, Simon Property Group, Parker-Hannifin and Southern Company all report Thursday. Lumentum, which supplies optical components to data centers, will be the next test of whether the AI hardware supply chain can find any bid at all.

Rates, the Fed and Friday’s Jobs Report: Bonds Rally as Equities Crack

Treasuries caught a safe-haven bid as tech cracked. The 10-year yield fell more than 6 basis points to 4.831%, extending Wednesday’s 7.4-basis-point drop to 4.888%. The 2-year yield, which tracks Fed expectations most closely, dropped more than 8 basis points to 4.50% after falling 5 basis points to 4.581% on Wednesday. The 30-year yield declined more than 3 basis points to 5.249%. Over two sessions the 2-year has fallen 13 basis points.

That is a rally driven by a mix of flight to safety and a reassessment of the inflation path. Wednesday’s 8% collapse in crude lowered the inflation outlook that had been pushing the Fed toward a hike, and a divided government means less new fiscal spending. Treasury supply has not eased, though: the government just borrowed a record $839 billion in a single week.

The Fed held its benchmark rate at 3.75% to 4.00% on October 28, with two officials voting for a hike. September CPI ran at 3.7% year over year, up from 3.4% in August, and producer prices rose 6.3%. Fed funds futures still price between a 79% and an 82% chance of a 25-basis-point hike in December. The Senate result also changes the Fed’s political backdrop. Democratic control means any new nominee to the Board of Governors needs Democratic approval, which blocks efforts to reshape the central bank and removes one source of uncertainty around Fed independence under Chair Kevin Warsh.

Labor data is now the swing factor. Friday’s October payrolls report, due at 8:30 a.m. ET, is expected to show between 59,000 and 65,000 jobs added, with unemployment holding at 4.2%. September’s report showed just 29,000 jobs added, with July and August revised down by a combined 60,000. Early indicators lean toward a firmer print. ADP counted 86,000 private jobs in October, above the 69,000 to 75,000 expected, with base pay up 3.2% year over year.

Thursday’s layoff data reinforced that read. US employers announced 43,994 job cuts in October, up slightly from September but down 71% from a year earlier and the lowest monthly total since 2023. Cuts in the first 10 months of the year are the lowest since 2022. The labor market remains in low-hire, low-fire mode. Initial jobless claims were expected near 196,000 after 194,000 the prior week, with continuing claims at 1.701 million.

The services side is cooling at the margin. The ISM services index slipped to 54.5 from 54.9, with new orders down almost four points to 55.9, but prices paid stayed elevated at 73.0.

The growth backdrop remains strong. Real GDP grew at a 3.4% annualized rate in the third quarter, according to the Bureau of Economic Analysis advance estimate. A hot payrolls number would firm December hike bets, lift front-end yields and the dollar, and add rate pressure on top of the political pressure already crushing growth stocks.

Across the Atlantic, the Bank of England voted 8-1 to raise its key rate a quarter point to 4%, its first hike since 2023, citing inflation it sees rising to 5% by mid-2027. Governor Andrew Bailey said the move should not be seen as the start of a long sequence, but warned it is unlikely to be the last increase unless the Middle East conflict is resolved.

Oil, Gold, Bitcoin and the Dollar: Cross-Asset Signals Point to Risk-Off

Crude bounced after Wednesday’s collapse. WTI rose 1.83% to $89.73 a barrel and Brent climbed 2.9% to $97.36. Both benchmarks tumbled more than 8% on Wednesday to settle at their lowest since August 31, and extended the slide into a fourth day overnight, with Brent hitting $93.82 and WTI $86.92 in Asian trading before buyers returned.

The oil story is pure geopolitics. Markets read Tuesday’s election as reducing the risk of further escalation in the US-Israeli war with Iran, since a Democratic Congress can constrain military options and a weakened president has more incentive to seek a deal. Vice President JD Vance added to that read overnight, posting that the war was the reason his party lost the midterms and suggesting a push for resolution. But uncertainty over whether the administration actually pursues a deal capped the selloff and allowed crude to recover part of its losses. A resolution would restore Middle East oil flows. Until it happens, $90 WTI and $97 Brent leave the inflation picture uncomfortably warm.

Gold extended its run. Spot gold traded between $4,120 and $4,134.83 an ounce, up 1.2% to 2% on the day after jumping 3.4% to $4,073 on Wednesday, its best day since August. Bullion is now at its highest level since October 22. The combination of falling yields, a weaker dollar and equity stress is the textbook setup for gold. The dollar hit its lowest level since September 30 on Wednesday as Treasury yields fell. The key test for gold is Friday’s payrolls: a hot print would reinforce December hike bets and pressure the $4,000 level.

Bitcoin followed equities lower. BTC-USD fell 3.97% to $77,024.24, giving back $3,185.91 on the day. Crypto has been trading as a high-beta risk asset all year, and on a day when the Nasdaq loses more than 4%, it was never going to hold. Bitcoin-linked equities were mixed earlier in the week, with Strategy rallying 11.37% to $103.05 and Circle Internet Group jumping 9.63% to $79.13 during Wednesday’s relief session.

The VIX tells the clearest cross-asset story. A jump from 16.61 to 23.46 in a single session is a volatility shock, and it forces systematic funds that target a specific volatility level to cut equity exposure. That mechanical selling tends to arrive in the afternoon and into the close, which means the market may not have finished pricing in Thursday’s damage when the opening hour ends.

Put the cross-asset picture together and the message is unambiguous. Bonds up, gold up, VIX up, bitcoin down, chips down hard. That is a risk-off tape. Oil’s bounce is the one outlier, and it reflects Middle East supply risk rather than any economic optimism.

Breadth and Leadership: A Narrow Selloff That Feels Like a Crash

The most important number of the day may be the one buried under the index losses. At 11 a.m. ET, two out of every five stocks in the S&P 500 were higher. That is not what a broad-based crash looks like. In a true panic, breadth collapses to 10% or 15% advancers. Thursday’s tape is a cap-weighted selloff concentrated in the heaviest stocks, with a meaningful minority of the market still bid.

Compare it with Wednesday. Advancers outnumbered decliners 3-to-1 on the S&P 500 during the relief rally. Nine of 11 sectors closed green. The Consumer Discretionary Select Sector SPDR (XLY) jumped 4.3%, Financials (XLF) gained 3.5% and Real Estate (XLRE) rose 3.2%. Only energy declined. Wednesday was broad participation. Thursday is narrow leadership in reverse: a handful of mega-cap chip and AI names dragging the indices lower while the rest of the market splits.

The Dow’s relative strength through the first hour confirms it. With a price-weighted structure and lower technology concentration, the blue-chip index was up 0.11% at the open and down just 0.66% at 9:57 a.m., while the Nasdaq was already down more than 3%. Chevron and Johnson & Johnson held green. The Dow only caught down to a 2.36% loss once the chip selling spilled into its tech components and the broader tape lost its nerve.

Small caps tell a similar story. The Russell 2000 opened up 1.58% at 2,856.95 as investors rotated toward domestic, rate-sensitive companies that benefit from lower yields and less tariff risk. By late morning it was down 2.92% at 2,773.64. That reversal shows how a mega-cap selloff eventually pulls down the whole market through risk reduction and volatility-targeting, even when the underlying rotation thesis still makes sense.

New lows deserve attention. Early in the session the S&P 500 posted 69 new 52-week highs against 3 new lows, which is a strikingly healthy reading for a day the index fell 1.66%. The Nasdaq Composite, by contrast, posted 132 new highs and 247 new lows. That divergence is the whole market in one data point: the broad S&P is holding together while the growth-heavy Nasdaq is breaking down underneath.

The concentration risk cuts both ways. If chips stabilize, the indices can bounce hard because so much of the damage sits in a handful of names. If chips keep falling, the rest of the market cannot carry the indices, no matter how many individual stocks are green. Thursday’s breadth says this is a rotation that got overwhelmed, not a bear market that just began.

Session Verdict: Bearish for Growth, a Rotation Without Capitulation

Thursday was a bearish session for the indices and a brutal one for AI hardware, but it was not a capitulation. The S&P 500 at 7,447.43 is down 3.14% and trading near its 7,425 intraday low. The Nasdaq at 24,630.58 is down 4.30% and officially in correction. The Dow at 52,098.05 has handed back 80% of Wednesday’s record gain. The VIX at 23.46 says the market is stressed, but two in five S&P 500 stocks were still rising late in the morning, and that is not the breadth of a market in freefall.

The driver is clear and singular. Tuesday’s election moved data-center construction from an economic question to a political one. Texas, Ohio and Georgia flipped to governors who campaigned on moratoriums. Congress now has the votes to hold hearings and the will to push AI regulation. States controlling 45% of planned U.S. data-center capacity will be run by moratorium supporters by January. Qualcomm and Arm added a handset slump on top. The market responded by pulling valuation out of every company whose growth depends on physical AI capacity coming online on schedule: AMD down 16%, Nvidia down 8.3%, Micron down as much as 16.8%, SanDisk down 22.2%, CoreWeave down as much as 14.9%, and the SOX down 11% in its worst day since March 2020.

The other side of the ledger matters just as much. AppLovin ripped 30% on a software-driven beat and raise. Energy led the sectors. Treasuries rallied, with the 10-year at 4.831% and the 2-year at 4.50%. Gold pushed above $4,120. Investors are not leaving equities. They are leaving AI construction risk and moving into cash flows that do not need a new permit in Texas.

Three levels define what happens next. On the S&P 500, the 7,425 intraday low is the line. A close below it opens the path toward 7,400 and the early-October trading range. On the Nasdaq, 24,400 is the intraday floor; losing it on a closing basis would extend the correction past 11%. On the SOX, a second consecutive double-digit loss would signal forced liquidation rather than repricing.

The catalysts arrive quickly. Lumentum, Cloudflare and Airbnb report after the close. Friday brings October payrolls, with consensus between 59,000 and 65,000 and a December hike priced between 79% and 82%. A hot number would add rate pressure to political pressure and likely extend the chip selloff. A soft number would ease hike bets, extend the bond rally and give the rotation into rate-sensitive sectors room to work. Next Tuesday’s October CPI, expected at 3.7%, is the next test after that.

The verdict: bearish session, narrow damage, unfinished business. The AI trade lost its political tailwind in 48 hours, and until Washington and the statehouses clarify how hard they plan to lean on data-center construction, the market will keep charging a risk premium on every chip, memory, server and neocloud name in the supply chain.

That's TradingNEWS