Geopolitics Tests AI Demand—and Selective Consumer Resilience

The supplied hypothesis is that earnings growth and resilient demand can support DDOG, SNOW, CRWV, AN, RH, WSM, ETH and LZB over the next year. The latest evidence does not reject that idea, but it changes the burden of proof.

The market is now asking two separate questions. First: is demand for AI infrastructure and cloud software real? Second: can that demand compound through a world of higher energy costs, disrupted shipping, tighter financing and selective household spending? Those questions matter differently for each name.

The near-term market tell is the split between operating demand and the cost of supplying it. On October 30, DDOG fell 11.5%, SNOW 7.3% and CRWV 9.7% at the regular close, while AN gained 6.1%; RH and WSM each fell about 2.7%, and LZB was essentially flat. These were 16:00 ET closes, not a live Saturday market print.^[1] The dispersion is more informative than a blanket “risk-off” label: investors are repricing capital intensity and duration risk in parts of the AI complex, while treating some consumer operators as company-specific stories.

The geopolitical shock is moving through energy and logistics

The World Bank’s October 28 Commodity Markets Outlook says global commodity prices are projected to rise 23% in 2026, with energy prices up 32%. It attributes the pressure to the Middle East conflict, depleted oil inventories and disrupted grain shipments, while warning that further escalation or damage to Russian oil infrastructure could push prices higher.^[2]

That is not a forecast that every company will suffer equally. It is a cost and discount-rate shock that travels through several channels: electricity and data-center operating costs, freight and equipment delivery, fertilizer and food prices, household budgets, and the inflation response of central banks. The World Bank also says metals prices are set to rise 25% in 2026, with data-center investment among the sources of demand.^[2]

The practical implication is that “demand resilience” is not enough. A company must also show that it can pass through costs, secure power and equipment, finance expansion without destabilizing the balance sheet, or earn high returns on already-committed capacity.

AI infrastructure: real demand, higher financing standards

CRWV offers the clearest case study. In an October 27 company announcement, CoreWeave said it had brought most of a planned 22.5 megawatts of IT capacity online at a Copenhagen data center, operated 51 data centers with more than 1.5 gigawatts of active power as of June 30, and had approximately 4.2 gigawatts of contracted power as of its August 11 earnings call. It also described a planned network expansion across the Middle East and Asia Pacific.^[3]

Those facts support the “demand is real” side of the hypothesis. Customers are asking for capacity, geographic reach and network connectivity. But they also reveal the risk: this is a capital-heavy business whose economics depend on power availability, hardware deployment, customer concentration, utilization and the cost of debt. A growth runway can coexist with a fragile equity story if financing costs rise faster than returns on new capacity.

Recent coverage points to that concern. CNBC’s search result described a sharp rise in CoreWeave’s credit risk and investor questions about additional financing, while another October 30 market report linked the selloff in CRWV and SNOW to jitters over hyperscaler spending and who ultimately pays for the AI buildout.^[4] The pages were not fully retrievable in this run, so those reports are best treated as market commentary rather than independently verified operating facts.

For DDOG and SNOW, the exposure is different. They are less directly tied to building physical power capacity, but their valuation and growth still depend on enterprise consumption, workloads moving into production and customer budgets surviving a more expensive macro backdrop. The next earnings updates therefore matter less as a simple beat-or-miss exercise than as evidence about usage breadth, renewal behavior, large-customer concentration and whether AI-related demand is incremental or displacing other spending.

The current schedule puts DDOG’s next report on November 4 before the open, with the date marked estimated; SNOW is scheduled for December 2 after the close, also estimated.^[5] Those dates are catalysts to monitor, not promises about results.

Consumer names: resilience is selective, not absent

AN’s October 30 gain stands out against the declines in RH and WSM, while LZB was nearly unchanged at the close.^[1] One session cannot establish a durable trend, but the cross-sectional reaction is consistent with a market that is differentiating between formats, price points, promotional exposure and execution rather than assuming a uniform consumer collapse.

That distinction matters under a commodity and shipping shock. Higher fuel and freight costs can pressure furniture and home-goods margins, while food and energy inflation can crowd out discretionary purchases. Yet operators with strong brands, better inventory discipline, a value proposition or share-taking potential may still grow even when the aggregate household backdrop is weak.

The evidence threshold for RH, WSM and LZB is therefore operational: traffic and conversion, average ticket, inventory turns, promotions, freight exposure and gross-margin protection. For AN, the market’s positive reaction suggests investors saw a more favorable company-specific signal on the day, but it does not prove that household demand has broadly reaccelerated.

The same caution applies to ETH in the requested scope. The quote feed returned an instrument labeled ETH at 21.29 at the October 30 close, down 4.4%; without a verified instrument description in this research pass, it should not be folded into the company-level conclusions above.^[1]

What would confirm or weaken the hypothesis?

The base case is mixed: AI demand remains durable, but capital-intensive supply is repriced more harshly; consumer demand persists, but leadership is selective. For the original hypothesis to be confirmed, the next evidence should show:

• DDOG and SNOW: stable or improving consumption, broad-based customer adoption and growth that does not rely on a handful of unusually large contracts.
• CRWV: contracted demand converting into utilized capacity, credible financing access, manageable power costs and evidence that returns exceed the cost of expansion.
• AN, RH, WSM and LZB: resilient traffic and margins despite freight, energy and household-budget pressure, with inventory and promotions under control.
• Macro: no further deterioration in shipping lanes, oil inventories or conflict-related supply routes, and no renewed inflation shock that pushes financing costs materially higher.

The hypothesis would weaken if AI customers defer deployments, if hyperscaler spending becomes more internally financed or less profitable, if CRWV’s credit risk rises alongside capacity commitments, or if higher energy and food costs spread from logistics into a broad consumer pullback. It would also weaken if the apparent consumer dispersion proves to be a one-day positioning effect rather than a difference in operating results.

What to watch next

1. The next company disclosures: DDOG on its estimated November 4 before-open date, followed by WSM, LZB, RH and SNOW’s scheduled reports in November and December. The calendar currently marks those dates as estimated, not confirmed.^[5]
2. AI infrastructure financing: debt yields, credit spreads, customer concentration and the pace at which contracted power becomes revenue-producing capacity.
3. Energy and shipping: oil inventories, the security of Gulf and Black Sea routes, freight costs and any new sanctions or waivers that alter supply chains.
4. Household pass-through: whether food and fuel inflation reduces discretionary demand, or whether brand strength and value segmentation allow selected retailers to keep taking share.

The cleanest conclusion is not that geopolitics has invalidated the growth thesis. It has raised its quality threshold. Demand can remain resilient while the equity outcomes diverge sharply according to who owns the energy risk, who must finance the capacity, and who can protect margins when households and supply chains are under pressure.