The energy shock is testing the software-growth thesis, not yet breaking it

A sharp geopolitical move is now colliding with a crowded technology narrative. Renewed attacks on commercial shipping around the Strait of Hormuz helped push Brent crude above $110 a barrel on October 16, while U.S. crude closed at $101.41. The same report said the S&P 500 was down more than 1% and the Nasdaq Composite about 2% in afternoon trading.^[1]

That backdrop matters for the software and AI-infrastructure names in this scope because it raises two questions at once: can resilient customer demand keep earnings estimates intact, and can higher energy, financing, and infrastructure costs compress the valuation investors are willing to pay for that growth?

The market tell: cloud names sold off while demand evidence held up

At the October 16 close, DDOG fell 7.19% to $247.11, SNOW fell 8.43% to $306.84, and CRWV fell 11.91% to $61.39. AN was a relative exception, rising 1.21% to $161.73; RH fell 4.14% to $112.99, WSM slipped 0.44% to $250.16, and LZB declined 1.28% to $30.07. These are regular-session closes at 16:00 ET, not a claim that the underlying businesses changed by those percentages in one day.^[2]

The cross-section is more informative than any one ticker. The market punished high-duration cloud and AI-infrastructure exposure most heavily, while a retailer and a home-furnishings name were mixed rather than uniformly lower. That is consistent with a repricing of financing sensitivity and risk appetite, but it is not, by itself, evidence that enterprise software demand has stopped growing.

Evidence for the thesis: Snowflake still shows durable usage signals

Snowflake’s latest disclosed quarter provides a useful counterweight to the market move. Product revenue was $1.16 billion, up 29% year over year; net revenue retention was 125%; and remaining performance obligations were $7.88 billion, up 37%. The company also reported 688 customers with more than $1 million in trailing-12-month product revenue.^[3]

Those figures support the “resilient demand” side of the hypothesis. They show expansion within existing customers, a growing contracted backlog, and continued penetration among large enterprises. But the same release identifies the variables that could make the market’s concern rational: consumption is variable, customers may optimize usage, budgets may be rationalized, contract durations may shorten, and higher inflation, tariffs, interest rates, foreign-exchange volatility, and geopolitical instability are explicit risks.^[3]

The balanced reading is therefore conditional. Growth can remain strong while the market multiple contracts. A company does not need to miss its revenue target for its stock to fall if investors decide that the growth is more capital-intensive, less durable, or more exposed to macro shocks than previously assumed.

Why geopolitics reaches software and AI infrastructure

The immediate channel is energy. The Strait of Hormuz is a critical shipping route, and the October 16 reporting linked the oil spike to attacks on commercial vessels as well as uncertainty around Russian diesel supply. It also reported that damaged refining infrastructure and constrained refinery capacity were tightening the diesel market.^[1]

The second channel is rates and inflation. A persistent energy shock can keep headline inflation elevated and make a rapid easing in financial conditions less likely. That tends to matter more for long-duration growth assets, whose valuations rely heavily on cash flows expected further into the future. This is a market mechanism, not a forecast that every company in the group will see an immediate cost hit.

The third channel is AI-infrastructure economics. Data centers and cloud platforms depend on power, equipment, networking, and financing. If those inputs become more expensive or less available, the question shifts from “how large is AI demand?” to “who captures the economics after power, depreciation, financing, and customer-price concessions?” That is particularly relevant to CRWV, while DDOG and SNOW face a different version of the question: whether customers keep expanding usage when finance and operating teams are asked to protect cash flow.

What the sell-off does and does not say about each group

Cloud and observability — DDOG and SNOW. The price action says expectations are high and macro sensitivity is back in the foreground. Snowflake’s disclosed growth and retention metrics support the operating case, but its consumption model makes customer optimization an important variable to monitor. DDOG’s next report is a near-term test of whether demand, usage, and forward guidance can reassert themselves against the valuation pressure.

AI infrastructure — CRWV. The larger decline suggests investors are especially sensitive to capital intensity, customer concentration, financing, and the pace at which new capacity becomes productive. A rebound in extended trading does not erase the regular-session decline: CRWV’s 16:00 ET close was $61.39, while the after-hours print was $61.6492 at 16:27 ET, 0.42% above that close.^[2] The distinction matters because a modest after-hours recovery is not the same as a reversal of the day’s repricing.

Consumer and housing-sensitive names — AN, RH, WSM, and LZB. These companies provide a read on whether higher fuel costs and rates are broadening into the consumer. Their mixed-to-lower performance on this session is less conclusive than the cloud sell-off. The next evidence should come from traffic, orders, promotions, housing sensitivity, freight, and gross-margin commentary rather than from the geopolitical headline alone.

ETH. The supplied market snapshot identifies ETH at $22.98, down 1.71% on the session, but that symbol does not by itself establish which security or product is intended. It is not used here as a proxy for the broader crypto market or as evidence about the software thesis.^[2]

The next catalysts are earnings, not speculation

The available earnings calendar lists DDOG for November 4 before the open, with the date marked estimated; SNOW for December 2 after the close, also estimated; AN for October 30 before the open, estimated; RH for December 10 after the close, estimated; WSM for November 18 before the open, estimated; and LZB for November 17 after the close, estimated. CRWV is marked unsupported by that calendar, so no date is stated here.^[4]

The key questions are concrete:

1. Usage: Are customers expanding consumption, or merely signing contracts and optimizing existing workloads?
2. AI monetization: Are new AI features producing incremental revenue and durable workloads, or mostly shifting activity within existing budgets?
3. Capital intensity: For infrastructure providers, are power, equipment, depreciation, and financing costs moving faster than revenue per unit of capacity?
4. Pricing power: Can companies pass through higher energy, labor, freight, or infrastructure costs without slowing demand?
5. Macro sensitivity: Do management teams change hiring, spending, guidance, or customer commentary after the oil and shipping shock persists?

What to watch next

The base case is mixed rather than binary: the operating evidence can remain constructive while the market demands a larger margin of safety. Confirmation of the bullish thesis would require sustained net retention, usage expansion, healthy enterprise budgets, and improving economics on new AI capacity. Evidence against it would be weaker consumption, more discounting, delayed deployments, rising financing needs, or guidance that acknowledges a longer period of macro caution.

The geopolitical risk is also conditional. If shipping disruption eases and energy prices retreat, the market may refocus on company-specific execution. If attacks continue, diesel remains constrained, or sanctions and supply routes become less predictable, the energy shock could become a broader inflation-and-rates problem. For this scope, the cleanest signal will be whether upcoming earnings reports show that customers are still expanding software and AI workloads after accounting for the higher cost of keeping the system running.

This article is for research and education, not personalized investment advice. Market prices and after-hours data are time-sensitive; corporate-event dates are estimates where the calendar labels them as such.