AI demand is visible; the cost of proving it is rising

The working hypothesis for this coverage set 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 invalidate that view, but it makes the claim more conditional: demand has to outrun the cost of power, logistics, capital and discretionary caution.

The market’s tell is the divergence inside the basket. At the latest available regular close, CRWV rose 7.3% and RH rose 3.1%, while DDOG fell 0.9% and LZB fell 0.9%. In pre-market trading on October 29, SNOW was $326.82, down 3.85% from its October 28 close, while RH was $147.18, up 1.46%; these are extended prints, not regular-session closes.^[1] The tape is rewarding evidence of operating momentum, but it is not granting every growth story the same multiple.

The geopolitical channel is moving through costs, not just headlines

A current trade-infrastructure assessment describes the March damage and temporary suspension of operations at Jebel Ali, alongside shipping restrictions affecting regional cargo.^[2] That matters for this equity set even where companies have little direct Middle East exposure. Disrupted routes can raise freight and insurance costs, increase working-capital needs and make energy-sensitive customers more cautious.

The macro backdrop compounds the issue. September data show unemployment at 4.2%, real GDP growth at 2.1% year over year and inflation at 3.72%, but the 10-year Treasury yield remains 4.99%; consumer sentiment is 48.1, down 12.7% year over year.^[3] This is not a recessionary base case, yet it is an unforgiving one for companies whose valuation depends on distant cash flows or whose customers can delay purchases.

Snowflake shows both sides of the argument

Snowflake’s latest available fiscal Q2 2027 call provides a useful test case. Management said AI workloads are growing quickly, but also lowered expected non-GAAP product gross margin to 74% because AI workloads currently carry a lower contribution margin. At the same time, it raised full-year non-GAAP operating-margin guidance from 13.5% to 14.5% and reiterated a 23% adjusted free-cash-flow margin target.^[4]

That is the central distinction for DDOG, SNOW and CRWV: demand can be real while unit economics are still in transition. Snowflake also reported adoption of its AI products across thousands of accounts, with CoWork surpassing 9,100 accounts in the cited call.^[4] The bullish interpretation is that usage creates a durable control point in enterprise workflows. The cautious interpretation is that the mix shift toward compute-intensive products can dilute margins before pricing and efficiency catch up.

For CRWV, the market is asking whether infrastructure growth can absorb power, equipment and financing costs. For DDOG and SNOW, the question is whether AI-driven workloads translate into durable consumption without triggering customer optimization. A prior Snowflake call said customers had become more cost-focused and were constantly optimizing, even while management reported no major optimization wave at that point.^[4] That is supportive, but it is not the same as immunity.

Consumer names face a different geopolitical transmission

AN, RH, WSM and LZB are less direct beneficiaries of AI demand and more exposed to the household balance sheet, freight, imported goods and housing-linked confidence. Their current price dispersion is therefore informative but not conclusive: RH’s positive extended move contrasts with LZB’s weaker regular session, while WSM’s pre-market print was modestly above its prior close.^[1]

The relevant risk is not simply that a shipping disruption stops a product from arriving. It is that higher logistics or input costs meet a consumer who is already reporting low confidence. The upside case requires companies to preserve traffic, pricing and gross margin simultaneously. The downside case requires only one of those variables to fail.

What would confirm or break the thesis

The base case remains mixed rather than binary. Earnings growth and resilient demand can support parts of the basket, especially where management can show contracted revenue, expanding operating margins and efficient AI monetization. But the geopolitical backdrop raises the hurdle: companies must demonstrate that demand is incremental and that cost inflation is controllable.

What to watch next:

• Cloud consumption: net retention, usage growth, new workloads and evidence of customer optimization at DDOG and SNOW.
• AI unit economics: product gross margin, inference or compute costs, pricing and the pace at which higher-margin core revenue offsets newer workloads.
• Infrastructure financing: power availability, capital intensity and debt-market conditions for CRWV and adjacent data-center suppliers.
• Trade and energy: shipping interruptions, insurance and freight rates, sanctions implementation and energy prices—especially if disruptions broaden beyond regional routes.
• Consumer proof: traffic, promotions, order backlogs and gross-margin commentary from AN, RH, WSM and LZB.

The cleanest reading of the market is not “risk-on” or “risk-off.” It is selective validation: demand is getting credit where it is measurable, while geopolitics is forcing every company to show who pays for the extra cost of growth.