AI Demand Is Resilient; Geopolitics Is Raising the Hurdle Rate
Why resilient AI workloads are not the same as a broad risk-on signal
The market is separating AI demand from geopolitical cost pressure
The supplied thesis is not behaving as a single trade. The latest evidence supports a narrower conclusion: resilient AI and cloud demand is still reaching software and infrastructure companies, but the geopolitical shock is raising the hurdle rate for consumer and capital-intensive businesses through energy prices, long-term yields and weaker confidence.
The market tell: DDOG and SNOW are holding up better than discretionary names
At the latest available regular close on October 21, DDOG rose 6.5% to $273.09 and SNOW gained 1.0% to $322.73. In the same session, AN fell 5.5% to $153.01 and LZB declined 1.7% to $28.68. These are not clean one-day verdicts, but they show the market rewarding visible AI-linked demand while remaining more sensitive to the consumer and housing cycle.
RH was up 6.9% at the close, but its pre-market print was $125.02, 2.3% below that close as of 8:13 ET on October 22. The contrast matters because it tests the growth hypothesis directly. Demand resilience can support the software names if it converts into usage, production workloads and durable budgets. It is less obviously sufficient for discretionary businesses when energy costs, financing conditions and confidence are moving in the wrong direction.
AI demand is real, but the timing still matters
Datadog’s latest available Q2 2026 call described revenue of $1.12 billion, up 36% year over year, with non-AI customers growing in the high 20s and about 4,720 customers above $100,000 of ARR. Management also said AI-native customers were scaling rapidly, while enterprise usage was more moderate because customers were still working through cloud migration and production deployment.
That is a constructive signal with an important qualification: the opportunity is not only whether companies announce AI projects, but whether those projects reach production and create recurring observability, security and data workloads. The same call record says more than 6,500 Datadog customers were sending data from at least one AI integration, representing about 80% of ARR. That is evidence of exposure to the cycle, not proof that every dollar of AI infrastructure spending will become durable software revenue.
Snowflake’s Q2 fiscal 2027 call provides a similar read-through. Product revenue grew 37% year over year for the third consecutive quarter, management cited a meaningful step-up in AI revenue, and net-new customer additions rose 32% year over year. Its AI Data Cloud served more than 41% of the Global 2,000 customer base. The encouraging interpretation is that AI is expanding consumption rather than merely shifting existing budgets. The caution is that consumption businesses can also expose customers to sticker shock as agents and new applications create more workloads.
Geopolitics is working through the discount rate and the income statement
Current market research describes a two-sided backdrop: growth indicators remain resilient, but commodity pressure and tighter financial conditions complicate demand and margins. The same risk complex includes conflict-related energy pressure and shipping uncertainty. A separate market report this month linked the Iran war to persistent energy-price pressure and described AI investment and technology sovereignty as part of a larger geopolitical investment cycle.
The macro data reinforces the transmission mechanism. The latest available FRED snapshot shows 3.72% year-over-year CPI inflation, a 5.15% 10-year Treasury yield, an 18.44 VIX reading and a 3.31% high-yield credit spread. Unemployment was 4.2% and real GDP growth was 2.1% year over year, so this is not a recession call from the data. But consumer sentiment was only 51.7, down 11.2% year over year. That combination—growth still positive, but financing and confidence strained—is a difficult backdrop for furniture, home improvement and auto-sensitive demand.
The shipping lane is important because it can turn a geopolitical event into a broad corporate margin problem. Longer routes, insurance costs, fuel prices and inventory buffers affect retailers and manufacturers directly. Even companies that are not energy producers can feel the shock through freight, logistics and household budgets.
What the scope says—and does not say
CRWV sits closer to the AI infrastructure side of the thesis, but its latest regular close was $61.82, down 0.5% on October 21. That weaker tape is a reminder that strong end-demand does not eliminate financing, capacity, customer-concentration or valuation risks for infrastructure providers.
Among the consumer names, WSM closed at $253.32 and was 1.1% lower in pre-market trading at $250.52 as of 8:22 ET. RH’s positive regular-session move also partially reversed before the open. Ethan Allen (ETH) gained 2.5% at the October 21 close, while LZB was lower. The dispersion argues against treating the group as a single macro expression: company-specific earnings, valuation and positioning still matter.
For crypto, the supplied ticker ETH is ambiguous in an equity quote feed and returned a $23.80 quote, consistent with Ethan Allen rather than the Ethereum token. This article therefore does not use that quote as evidence about cryptocurrency markets.
Base case and risk case
The base case is a barbell: AI infrastructure and data workloads continue to grow, while higher energy and financing costs keep pressure on discretionary demand. That would favor companies able to show production usage, pricing power and disciplined capital deployment, without implying that any individual ticker must rise.
The risk case is that the geopolitical shock broadens. A sustained shipping disruption or additional energy escalation could lift inflation expectations and long-term yields, slowing both enterprise projects and household spending. The opposite risk is that the energy premium fades while AI workloads keep accelerating; in that scenario, the market may continue to reward measurable consumption growth even as slower-growth discretionary names lag.
What to watch next
- AI conversion: whether DDOG and SNOW continue to report production usage, expansion and large-customer adoption—not just pilot activity.
- Infrastructure economics: whether CRWV can translate AI demand into durable margins and cash generation while funding capacity.
- Consumer evidence: traffic, order trends, promotions and guidance from AN, RH, WSM and LZB as confidence remains weak.
- The macro channel: the 10-year yield, credit spreads, inflation expectations and energy prices; these determine how far geopolitical risk travels beyond shipping.
- Escalation versus normalization: whether shipping routes normalize and the energy premium fades, or whether disruption becomes persistent enough to alter inventories and corporate planning.
The evidence supports resilient AI demand, but not a blanket all-clear for the full scope. What would have to be true for the bullish case to win is clear: AI projects must reach production, consumption must remain durable, and geopolitical energy and shipping costs must stop worsening. Until those conditions are visible together, the better description is selective resilience under a higher macro hurdle.