Market structure · 5 November 2026 session
Who won, who lost and why
Investors punished exposure to future AI construction and still rewarded businesses delivering stronger earnings today. The session combined three forces: a broad reduction in risk, a concentrated unwind of AI infrastructure exposure, and large reactions to individual earnings and political news. Treating every red stock as an AI casualty, or every green stock as a defensive, misses the most useful information.
2,111 listed companies with a market value above USD 10 billion · one-day price change · prepared 6 November 2026, Dubai
2,111
listings worldwide: 908 up, 1,180 down, median −0.47%
85%
of the 757 USD-quoted listings fell, median −2.70%
31.3% up
AppLovin, the biggest winner, on its results
19.8% down
Sandisk, the biggest loser, on nothing of its own
01 · Where the selling was concentrated
Baskets by product show a sharper divide than sector labels.
We grouped USD-quoted names by what they sell and who pays for it, and took the median move of each basket. Memory and storage fell 15.6%, optical and high-speed links 13.9%, AI power and construction 11.7%. Those clusters share one exposure: the pace and economics of building computing capacity. Advertising platforms rose. The broad sector medians sit in between and hide the split.
| Basket or sector | Median move | What is inside |
|---|---|---|
| Memory and storage | −15.6% | STX, MU, WDC, SNDK |
| Optical and high-speed links | −13.9% | GLW, COHR, FN, CIEN, CRDO, LITE, ALAB |
| AI power and construction | −11.7% | HUBB, ETN, FIX, PWR, NVT, VRT |
| Electronic technology, all USD-quoted | −8.5% | sector median, 757-name slice |
| Health technology, all USD-quoted | −3.5% | sector median |
| Technology services, all USD-quoted | −2.9% | sector median |
| Utilities, all USD-quoted | −0.7% | sector median |
| Consumer non-durables, all USD-quoted | −0.7% | sector median |
| Advertising platforms | positive | APP, U |
Median of the constituents’ one-day moves, equal weights. The baskets are analyst-selected examples from the USD-quoted slice, not sectors, indexes or portfolios; small baskets can be dominated by one earnings event. The full basket list is in the notes at the end.
02 · Winners: businesses selling outcomes
Advertising, payments and critical software.
Customers kept paying for products tied to revenue, transactions and essential operations, and the companies that reported stronger numbers were bought on a day when almost everything else was sold.
| Product and customer | Examples | Why they held up |
|---|---|---|
| Performance advertisingAdvertisers and app publishers | AppLovin +31.3% Unity +14.8% | Strong earnings; evidence that advertising technology generates revenue now. |
| Business paymentsCompanies, fleets, cross-border users | Corpay +7.8% | Organic growth and a higher outlook outweighed the wider risk reduction. |
| Insurance data, engineering softwareInsurers and engineering teams | Verisk +6.0% Bentley +3.8% | Earnings-day gainers with embedded workflow products and different customer exposures. |
| Safety communicationsPublic agencies and enterprises | Motorola Solutions +4.7% | Higher guidance and a record backlog supported demand visibility. |
AppLovin: earnings and cash were the catalyst
Quarterly revenue of USD 2.18 billion, up 55%; net income USD 1.44 billion, up 72%; free cash flow about USD 1.42 billion; fourth-quarter revenue guidance USD 2.365 to 2.395 billion. The stock responded to a specific improvement in reported performance and outlook, not to advertising being defensive.
Unity: a different exposure to AI from the builders
Revenue grew 32% to USD 622 million and the company reported its first GAAP-profitable quarter since listing. Improving the economics of existing digital activity is a different proposition from funding another data centre. Advertising budgets remain exposed to consumer demand and competition.
Corpay and the workflow businesses: productivity survives a risk-off day
Corpay reported 11% organic revenue growth, a 32% rise in adjusted earnings per share and raised its full-year outlook. Verisk and Bentley also reported around this session. Bentley is the important counterexample: subscription revenue grew 12.4%, management explicitly cited AI infrastructure investment as a benefit to its engineering customers, and the stock rose despite that exposure. The distinction is business economics and earnings expectations, not a blanket rule that AI exposure must lose.
03 · Winners: households, public services and gold
Stable demand helped, but selection mattered.
A handbag brand, a water utility and a gold royalty company won for very different reasons.
| Customer and product | Observed winners | Economic transmission |
|---|---|---|
| Households: daily purchases | Costco +1.8%, Kimberly-Clark +1.5% General Mills +1.4%, Hershey +1.3% | Frequent purchases and established distribution give more predictable demand than capital projects. |
| Households: branded discretionary goods | Tapestry +6.3% | Brand-specific sales and margin strength; consumers remained selective. |
| Residents and municipalities: essential services | American Water +1.4%, Waste Management +1.0% Republic Services +0.9% | Recurring necessity spending cushions earnings expectations. |
| Investors: gold exposure | Royal Gold +5.1%, Newmont +1.0% Evolution Mining +14.2% | Gold prices and demand for protection; miners still carry cost and execution risk. |
Tapestry: a consumer winner, not proof of a spending boom
Sales of USD 1.89 billion, up 11%, with Coach up 16% and Kate Spade down 10%; the annual earnings and revenue outlook was raised. That is a successful brand and product mix inside an uneven consumer environment, not evidence that discretionary spending strengthened.
Stable spending was a relative advantage, not universal protection
PepsiCo rose 0.6% but Procter & Gamble fell 0.5% and Walmart 0.7%. Duke Energy rose 0.5% while Southern fell 0.2% and American Electric Power 2.0%. The consumer non-durables median was −0.66% and utilities −0.72%: far better than electronics, but neither group broadly rose.
Gold and copper told different stories
Gold-related gains contrasted with Freeport-McMoRan −7.3% and Lundin Mining −11.0%. Gold offers financial protection; copper is tied to industrial activity. Local trading hours and costs explain why miners did not move together.
04 · Losers: the computing supply chain
The hardest-hit products sat behind AI capacity.
Different technologies faced a shared risk: that customers build or spend more slowly.
| Product | Examples | Who ultimately pays |
|---|---|---|
| DRAM, HBM and NAND memory | Micron −15.4%, Sandisk −19.8% | Cloud operators, server makers, device manufacturers. |
| Hard drives and storage systems | Western Digital −15.8%, Seagate −15.3% | Data-centre and enterprise storage budgets. |
| Processors and accelerators | AMD −15.7%, Arm −11.8% Broadcom −9.5%, Nvidia −8.9% | Cloud, enterprise and device platforms; exposure differs by product. |
| High-speed electrical and optical links | Astera Labs −17.0%, Lumentum −16.3% Credo −15.6%, Coherent −12.3% | Server, networking and optical-system manufacturers. |
| Chip fabrication and test tools | Lam Research −14.6%, Applied Materials −12.0% KLA −11.4% | Foundries and memory producers funding new capacity. |
Why memory and storage fall harder than the processors
Memory and storage carry both volume risk and pricing risk. If buyers slow orders or inventory rises, selling prices and margins fall together. Factories carry large fixed costs, so a modest change in demand produces a much larger change in expected profit. DRAM, NAND and hard drives are distinct markets; on this day they shared one sell-off.
The equipment suppliers are one spending decision further upstream
Lam, Applied Materials and KLA sell the tools used to make chips. Their customers can defer a factory expansion even while existing AI services keep growing. The reaction reflects concern about the next investment cycle, not evidence that today’s end demand disappeared.
A lower multiple is not enough to identify a bargain
A low trailing P/E can reflect unusually strong profits near the top of a cycle. The relevant question is what earnings look like under slower demand, normalised pricing and lower utilisation. A high-quality supplier can also fall with the theme while its competitive position stays intact. The daily return alone cannot tell these cases apart.
05 · Losers: construction, capacity and software
One customer budget reached many sectors.
The sell-off crossed sector labels because the underlying revenue dependencies overlap.
Power equipment, cooling and construction services
Vertiv −13.9%, nVent −12.3%, Quanta Services −12.1%, Comfort Systems −11.3% and Eaton −7.1%. Their products differ: thermal management, electrical protection, grid construction, mechanical systems, power distribution. Incremental data-centre construction is the common demand source. Permitting delays, power constraints or a higher cost of funding push revenue into later periods, and then fixed costs, working capital and the timing of customer payments matter as much as the size of the order book. Hubbell, down 3.1%, held up best in this group.
Specialist cloud capacity: demand plus financing exposure
Nebius −14.4% and CoreWeave −9.3% sell access to computing capacity, not a component. A project has to be financed, built, powered and occupied, so a slowdown hits utilisation and returns on capital as well as construction timing. Reuters highlighted funding pressure on the smaller participants in the build-out.
Cybersecurity and enterprise software: the product may be essential while the stock reprices
CrowdStrike −10.0%, Palo Alto Networks −9.2%, Zscaler −8.9%, Rubrik −11.2% and Cloudflare −10.2%; Palantir −9.9%, Datadog −7.2%, Snowflake −6.8%, MongoDB −6.6%. None of this shows that customers stopped securing their systems. Investors can cut the price they pay for future growth, sell crowded positions or worry about customer expansion budgets without any company-specific earnings cut.
The large platform buyers fell less than their suppliers
Alphabet −1.7%, Amazon −2.6%, Microsoft −3.9%, Meta −4.1%: diversified businesses and the main infrastructure customers. Slower spending could protect their cash flow while hurting suppliers, or weaken their own growth. The prices do not prove that they reduced budgets.
06 · Company-specific losers
Four declines that need their own explanation.
A broad market narrative cannot replace reading the individual business news.
| Business and move | Reported catalyst | What it means |
|---|---|---|
| Coca-Cola HBC −17.9% | Reports of a request to place its Russian business under external management; the company said it had received no government notification. | Jurisdiction and asset-control risk overwhelmed the usual stability of beverage demand. A seizure was not confirmed. |
| Rheinmetall −16.1% | Cut 2026 sales guidance to about EUR 13.2 billion from EUR 13.7 to 14.2 billion; vehicle programme and ammunition delays. | Government demand does not guarantee timely revenue, deliveries or cash collection. |
| Expedia −13.9% | Reported earnings strength overshadowed by a softer holiday-quarter bookings outlook. | Investors valued the next demand signal more than the completed quarter. |
| MercadoLibre −13.7% | Revenue up 45% to USD 10.75 billion, but operating income down 6% and an operating margin of 6.3%; shipping, capacity and marketing investment weighed. | Strong growth can coexist with weaker economics for shareholders in the current period. |
Coca-Cola HBC is not Coca-Cola, Coca-Cola Consolidated or CCEP
Coca-Cola Consolidated rose 4.5%, Coca-Cola Europacific Partners 1.1%, and The Coca-Cola Company was flat. Similar products do not produce the same legal exposure, geographic cash flows or shareholder outcomes.
Motorola Solutions versus Rheinmetall; Tapestry versus MercadoLibre and Expedia
Both Motorola and Rheinmetall sell to governments, but Motorola raised its outlook on a USD 16.6 billion backlog while Rheinmetall cut guidance: execution and the conversion of demand into cash, not a preference for public spending. Tapestry, MercadoLibre and Expedia all depend on households, but Tapestry delivered a stronger outlook and margins, MercadoLibre sacrificed margin to expand, and Expedia faced doubt about upcoming bookings. The useful grouping is consumer segment, geography, profitability and expectations, not “consumer stocks”.
07 · Finance, commodities and medicine
The rest of the economy was mixed.
The risk reduction spread well beyond AI, but each business model had its own exposure.
| Product and customer | Examples | Economic exposure |
|---|---|---|
| Banking and investment bankingBorrowers, depositors, issuers | JPMorgan −4.1%, Goldman Sachs −6.4% Morgan Stanley −5.8% | Credit conditions, deposit costs, asset values and capital-markets activity. |
| Trading and brokerageInvestors and traders | Interactive Brokers −6.2%, Robinhood −10.2% Coinbase −9.0% | Asset prices, client balances, leverage and trading volumes. |
| ExchangesTrading and hedging customers | Cboe +1.5%, CME −0.5%, ICE −0.4% | Transaction volumes, clearing, data and product mix; not a uniform volatility trade. |
| FertiliserAgricultural supply chain | CF Industries +6.0% | Nitrogen selling prices against natural-gas feedstock costs; results reported ahead of the session. |
| Oil and servicesEnergy users and producers | Exxon −1.7%, Chevron −1.3%, Schlumberger −3.1% | Oil-price expectations, producer spending and operating costs. Oil stocks fell despite an intraday oil rebound. |
| Medicines and distributionPatients, providers, insurers | Cardinal Health +2.8%, Eli Lilly −3.2% Johnson & Johnson −1.4% | Necessary demand still meets drug pricing, reimbursement, pipelines and valuation; health technology median −3.45%. |
More volatility does not mean every financial business wins. Cboe rose, but most banks and brokers fell: transaction volume helps trading venues, while lower asset values, risk limits and weaker issuance expectations hurt the others. These are transmission channels to monitor, not verified company events.
08 · International markets
Asia and Europe had different winners.
Local sessions and product mixes explain why the global list looks healthier than its USD-quoted slice.
| Quote currency | Listings | Share up | Median move |
|---|---|---|---|
| USD | 757 | 14.7% | −2.70% |
| TWD | 57 | 12.3% | −6.42% |
| KRW | 46 | 39.1% | −1.65% |
| GBX | 57 | 50.9% | +0.07% |
| CNY | 246 | 56.9% | +0.42% |
| EUR | 215 | 61.4% | +0.43% |
| AUD | 33 | 66.7% | +1.82% |
| HKD | 71 | 87.3% | +2.63% |
| JPY | 142 | 73.2% | +2.77% |
Japan. Toyota +16.7%, Subaru +9.0%, Suzuki +8.9% after Toyota raised its annual profit outlook; Konami +10.9%, Capcom +8.2%, Bandai Namco +8.4%; Mitsubishi Estate +9.9%. Furukawa Electric −17.1% and Ibiden −16.3%. Rotation into non-chip businesses while the Nikkei fell.
China and Hong Kong. Xiaomi +8.8%; Alibaba +2.5%, JD.com +3.4%, PDD +1.9% in their USD lines; Rockchip +10.0%. Domestic technology demand was treated differently from US AI capital spending.
Australia. Evolution Mining +14.2%, Northern Star +9.9%, Fortescue +9.2%, BHP +6.2%. Gold and iron ore have different pricing drivers; the Australian session closed before the US one opened.
Europe. Verbund +8.6%, Adyen +4.5%, Bayer +3.7%, BASF +3.3%: electricity, payments, healthcare and chemicals. Rheinmetall and Coca-Cola HBC show that local company news still mattered.
Quote currency is a grouping field, not a verified exchange or country. Local-currency returns over different observation windows; this is not an index comparison.
09 · Customer budgets
The customer map is more useful than the sector map.
The key question is whose spending pays for the product, and whether that spending can be postponed.
| Customer and budget | Products and examples | The day’s message and the next test |
|---|---|---|
| Cloud platforms and AI labsCapital expenditure | Processors, memory, networking, cooling, computing capacity | Broadly negative. Check signed orders, funding, power availability, delivery dates and utilisation. |
| Advertisers and app publishersOperating expenditure | AppLovin, Unity: customer acquisition and monetisation | Positive earnings exceptions. Check advertiser returns, retention, platform access and cash generation. |
| Insurers, engineers, finance departmentsOperating expenditure | Verisk, Bentley, Corpay: data, design and payments | Selected winners. Check renewals and organic growth; do not extrapolate from one earnings day. |
| HouseholdsNecessary spending | Food, personal care, everyday retail, water and waste | Relative resilience. Check real volumes and margins rather than price increases. |
| HouseholdsDiscretionary spending and credit | Coach bags, vehicles, travel, e-commerce | Mixed. Product appeal, financing costs, geography and guidance separated winners from losers. |
| Governments and municipalitiesProcurement budgets | Safety communications, defence systems, infrastructure | Mixed. An approved budget is not a funded order, an acceptance milestone or a collected payment. |
| Investors and market participantsTrading and protection | Gold businesses, exchanges, brokers, crypto platforms | Gold held up; most trading-sensitive stocks fell. Volume, asset prices, leverage and product mix matter. |
Diversification can disappear at the customer level
A chip designer, an optical component maker, a cooling supplier and a specialist cloud operator have different factories and business models. They still depend on the same few hyperscalers, the same electricity connections and the same funding environment, and their stocks fall together even when they sit in different formal sectors.
Economic exposure, not a simple capex-versus-opex rule
Operating budgets can also be cut, and some capital projects are contractually committed. The differences that matter are urgency, cancellation rights, the customer’s own return, balance-sheet strength and the point at which the supplier gets paid.
10 · Macro forces
What changed in the market’s assumptions.
Several forces worked at once. No source can assign an exact share of each stock’s move to each cause.
| Factor | Observation | Transmission to businesses |
|---|---|---|
| AI policy and permitting | Reuters linked the selling to concern that election results could slow infrastructure expansion. | Longer delivery schedules and lower near-term returns for construction, equipment and capacity providers. A forecast of policy impact, not enacted rules. |
| Risk appetite and positioning | The VIX reached about 28 intraday; crowded-trade unwinds and volatility-driven selling were cited. | Heavily owned growth themes fall together without fresh deterioration at every company. |
| Rates and financing | Treasury yields eased during the sell-off while funding conditions stayed a concern. | Lower yields help valuations, but wider risk premiums or weaker growth expectations can dominate. |
| Oil, war and inflation | An intraday oil rebound amid uncertainty around the Iran conflict. | Fuel costs hurt consumers, transport and margins. Oil stocks in the list still fell. |
| Gold and protection | Gold strengthened during the session. | Supports gold revenue, but operating costs and company risk still separate the miners. |
| Earnings, margins and guidance | Stronger outlooks at the winners; weaker conversion or guidance at several losers. | A direct change in expected cash generation outweighs the sector or macro narrative. |
What we take from it
Three tests before the next order.
The market rewarded evidence of profitable demand and punished uncertainty about the timing, financing and returns of future spending. Individual earnings and legal risks still overrode that pattern.
One
Test the customer
Can the customer fund the next order and earn an acceptable return on it? For AI capacity that means signed orders, power, delivery dates and utilisation, not announced plans.
Two
Test the supplier
Can it survive delays without issuing equity or losing pricing power? Fixed costs, working capital and the point at which it gets paid decide that, not the headline order book.
Three
Test the valuation
Use earnings under a slower-spending scenario, not an unchanged analyst forecast. A low trailing multiple near the top of a cycle is not a margin of safety.
Nvidia, TSMC and Arista remain on our research shortlist. Strong businesses can be exposed to the same customer cycle; the size of a one-day fall is not, on its own, a buy signal.
How we built it
Universe: every listed company with a market value above about USD 10 billion, 2,111 listings, one-day price change as exported from our screener for the 5 November 2026 session; snapshot returns, not independently verified closes, and counts refer to listings, not unique businesses. Medians use unrounded figures with equal listing weights; no currency conversion. The USD-quoted slice (757 listings) is analysed separately; one MXN line with five shares of volume was excluded. Baskets: advertising APP, U · payments and workflow CPAY, VRSK, BSY, BR, CTSH, TYL, ACN, FICO · everyday goods COKE, COST, KMB, GIS, HSY, CCEP, PEP, KO, PG, WMT · utilities WTRG, AWK, EVRG, ES, DUK, SO, NEE, AEP · oil and gas COP, OXY, CVX, XOM, HAL, SLB, LNG · banks BAC, WFC, JPM, C, MS, GS · enterprise software NOW, MDB, SNOW, ORCL, DDOG, PLTR · AI power and construction HUBB, ETN, FIX, PWR, NVT, VRT · chip designers INTC, NVDA, AVGO, ARM, MRVL, AMD · cybersecurity FTNT, ZS, PANW, CRWD, NET, RBRK · cloud capacity CRWV, NBIS · chip tools KLAC, AMAT, TER, LRCX · optical GLW, COHR, FN, CIEN, CRDO, LITE, ALAB · memory and storage STX, MU, WDC, SNDK. Catalysts come from company releases and Reuters, WSJ, Investor’s Business Daily and Alliance News coverage of 5 November 2026. Customer-budget explanations are our interpretation, not statistical estimates of cause.
Important information
This page is for general information and discussion only and is not investment, legal or tax advice, nor an offer or recommendation to buy, hold or sell any security. Observations and analytical judgements are not personalised trade instructions, and nothing about future returns can be inferred from one trading session. Positive daily returns do not establish low valuations or low risk. All investments involve risk, including loss of capital. Past performance is not a reliable indicator of future results. Analysis as of 6 November 2026.