Earnings growth is real—but demand is splitting into winners and funding tests

The opening snapshot is a clean tell: Nasdaq-linked growth is being rewarded, but the basket is not trading as one “resilient demand” theme. On October 26, QQQ rose 1.11% and SPY rose 0.71%, while the Dow gained 0.31% at the 16:00 ET close. Reuters attributed the broader lift to Anthropic’s IPO filing, which renewed enthusiasm for AI-linked stocks at the start of a heavy earnings week.^[1]

That backdrop supports part of the working hypothesis—that earnings growth and demand can carry selected names—but it also raises the bar. Investors are separating mission-critical software growth from capital-intensive AI infrastructure and rate-sensitive consumer demand.

The tape is already doing the sorting

Across the six equity names with comparable six-month data, the dispersion is unusually wide. From April 29 through October 26, SNOW gained 137.18% and DDOG gained 109.17%, while CRWV fell 51.67% and AN fell 22.12%. WSM gained 37.96%; RH was up only 2.84% after a 43.42% maximum drawdown.^[2]

The last row needs care. A one-day move is not an earnings verdict, and the quote feed reports regular-session closes, not a fundamental diagnosis. But the cross-sectional pattern is useful: the market is paying for visible growth where it sees operating leverage, while discounting businesses exposed to funding costs, big-ticket purchases, or uncertain demand conversion.

What the software winners still have to prove

DDOG closed at $280.24, up 3.93% on October 26, and SNOW closed at $334.95, up 2.93%; their after-hours prints were essentially flat relative to those closes as of 19:59 ET and 19:47 ET, respectively. The immediate tape therefore confirms enthusiasm, but not a fresh post-close fundamental re-rating.

Snowflake’s own fiscal third-quarter release gives the strongest operating evidence in this group: 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%.^[3] Those are meaningful signs that enterprise data workloads remain active. They do not, by themselves, settle the harder questions around durable AI consumption, pricing, margins, or how much of the growth is already reflected in the valuation.

Datadog’s October 26 move also came amid a reported analyst target increase ahead of its November earnings report, with coverage pointing to AI-related demand as part of the rationale.^[3] That is market commentary, not company guidance. The next useful test is whether reported usage, customer expansion, and free-cash-flow conversion validate the enthusiasm rather than merely extend it.

AI infrastructure is the counterexample

CRWV closed at $55.19, down 51.67% over the six-month comparison window, despite rising 1.76% in Monday’s regular session and another 2.26% in the available after-hours print as of 19:59 ET.^[2] The combination is a reminder that a positive daily reaction can sit inside a much larger drawdown.

The market’s concern is not necessarily that AI demand has vanished. It is that demand must outrun the cost of building and financing capacity. For infrastructure companies, earnings growth is only one leg of the case; utilization, customer concentration, capital intensity, debt terms, and the timing of cash returns matter just as much.

The Anthropic filing adds fuel to the AI complex, but it can also sharpen comparisons. New public-market supply and stronger disclosure may make it easier to distinguish recurring software economics from infrastructure economics. That is a sorting event, not proof that every AI-linked balance sheet benefits equally.

Consumers are resilient in aggregate, selective in practice

The macro backdrop is mixed rather than recessionary. The latest snapshot shows unemployment at 4.2%, real GDP growth at 2.1% year over year, and industrial production growth at 2.25%; inflation was 3.72%, the 10-year Treasury yield was 4.96%, and consumer sentiment was a weak 48.1.^[4]

That combination can support employment-linked spending while still making large discretionary purchases harder to justify. It also keeps financing conditions important for housing-adjacent and auto businesses. On October 26, AN fell 0.53%, RH fell 1.31%, ETH fell 3.29%, and LZB fell 1.20%; WSM fell 0.99% in regular trading before a modestly higher reported after-hours print.

These are not independent proof points about each company’s quarter. They are evidence that the demand thesis is conditional. For home and auto names, the next confirmation must come from traffic, ticket sizes, order trends, cancellations, promotions, and gross-margin discipline—not from the existence of a strong labor market alone.

Hypothesis scorecard

Evidence for the hypothesis:

• Enterprise software is showing the clearest price leadership in this scope, with DDOG and SNOW far ahead of the comparison basket over six months.^[2]
• Snowflake’s reported product growth, retention, and remaining performance obligations indicate that customers are still committing to data workloads.^[3]
• The macro data do not currently describe a recession: unemployment and GDP remain compatible with ongoing demand.^[4]

Evidence against an unqualified version:

• CRWV’s six-month decline and 94.4% annualized volatility show that AI demand does not eliminate financing, execution, or valuation risk.^[2]
• Consumer sentiment is weak even as headline growth holds up, a combination that can produce selective rather than broad discretionary demand.^[4]
• RH, ETH and LZB were lower on Monday, and AN remains substantially below its April level.^[2]

Current read: the hypothesis is partly supported, with the strongest evidence in mission-critical software. It is not yet supported as a blanket claim across AI infrastructure, auto retail, and home furnishings. The base-rate interpretation is a two-speed market: recurring digital demand can remain durable while the cost of funding capacity and the cost of financing big-ticket purchases set the terms for everyone else.

What to watch next

1. Software earnings quality: revenue growth, usage or consumption trends, retention, margin progression, and free-cash-flow conversion at DDOG and SNOW.
2. AI infrastructure cash economics: CRWV’s utilization, customer concentration, capital expenditures, debt structure, and path from contracted demand to cash generation.
3. Consumer demand composition: traffic versus ticket, promotions, cancellations, and order backlogs at AN, RH, WSM, ETH, and LZB.
4. Rates and confidence: the 4.96% 10-year yield and 48.1 consumer-sentiment reading are important context for any thesis built on discretionary spending.^[4]
5. Market breadth beyond the headline index move: whether strength broadens beyond AI-linked and software leaders, or remains concentrated in the names with the clearest growth narrative.

The useful conclusion is not that demand is either strong or weak. It is that the market is demanding different kinds of proof: recurring revenue and retention for software, funded utilization for infrastructure, and evidence of willingness to spend for discretionary businesses.