The market is testing demand — and the cost of the AI story
Why resilient demand is no longer enough on its own
The market is testing demand — and the cost of the AI story
Friday’s tape turned a broad hypothesis into a sharper question: can resilient demand and earnings growth keep carrying high-expectation companies when the cost of financing, infrastructure and household purchases is moving against them?
The answer is not uniform. Datadog (DDOG) and Snowflake (SNOW) still show powerful six-month momentum, while CoreWeave (CRWV) is the clearest stress test for AI infrastructure economics. AutoNation (AN), RH, Williams-Sonoma (WSM), Ethan Allen (ETH) and La-Z-Boy (LZB) sit in a different bucket: their next leg depends more directly on the consumer, housing turnover, tariffs and the durability of margins.
The opening snapshot: a risk-off day with a narrower fault line
The regular session ended Friday, October 16, with SPY down 1.11%, QQQ down 2.50% and DIA down 0.38%. The growth-heavy QQQ decline was the important signal: investors were reducing exposure to long-duration technology and the financing-sensitive part of the AI trade. These were 16:00 ET closes; the market is closed Sunday.[1]
DDOG fell 7.19%, SNOW 8.43% and CRWV 11.91% on Friday. AN rose 1.21%, while RH fell 4.14%, WSM 0.44%, ETH 1.71% and LZB 1.28%. After the close, DDOG was $245.92 at 19:57 ET, SNOW $304.53 at 19:57 ET and CRWV $60.95 at 19:59 ET.[1]
That pattern says the market is demanding proof that growth can survive a higher hurdle rate and a more skeptical view of AI financing.
Six months of returns tell two stories
| Company | Six-month return | Annualized volatility | Maximum drawdown |
|---|---|---|---|
| DDOG | +91.13% | 72.79% | -27.39% |
| SNOW | +103.42% | 68.89% | -19.35% |
| CRWV | -46.69% | 95.32% | -55.92% |
| AN | -21.23% | 36.24% | -33.39% |
| RH | -18.48% | 56.18% | -43.42% |
| WSM | +23.67% | 32.46% | -16.63% |
These figures are not a forecast. They show that AI and consumer are not single-factor trades. DDOG and SNOW delivered large gains before Friday’s reset; CRWV lost nearly half its value with the highest volatility in the group. WSM was positive with much lower volatility than software and infrastructure.[2]
The base-rate interpretation is mixed. Strong demand can support software revenue, but the market is distinguishing recurring, asset-light growth from growth requiring heavy capital spending and debt.
CoreWeave is the financing test inside AI
Reuters described CoreWeave as a loss-making, deeply indebted cloud company targeted during the week’s AI scare. CNBC reported that its equity and debt were trying to find their footing after a sharp selloff, with the 2032 bond yield reported at 15.77%. CRWV’s six-month maximum drawdown was 55.92%, and Friday’s regular-session loss was 11.91%.[3][2]
This does not disprove AI demand. It changes the question. Demand for compute can coexist with a difficult equity outcome if the cost of building, financing and renewing capacity rises faster than cash generation. Demand alone may not be enough when the capital structure becomes part of the narrative.
DDOG and SNOW offer a counterpoint, but not a clean all-clear. Their six-month returns were 91.13% and 103.42%. Recent coverage highlighted concern about financing AI spending, while Snowflake’s finance chief resisted putting early CoCo-agent adoption directly into guidance. Product interest and forecastable revenue are different things.[2][3]
Consumer names test another kind of resilience
The consumer and home group faces a more tangible hurdle: can demand hold while rates, weak sentiment and tariff effects pressure purchases and margins?
The latest macro snapshot is not recessionary: unemployment was 4.2%, real GDP grew 2.1% year over year and industrial production 2.25%. But consumer sentiment was 51.7, down 11.17% year over year, while the 10-year Treasury yield was 5.1%. High yields can make big-ticket purchases and housing-linked decisions harder even when aggregate growth remains positive.[4]
WSM is the strongest evidence here. Its investor release reported second-quarter comparable-brand revenue growth of 6.2%, a 22.9% GAAP operating margin and a raised full-year 2026 outlook. Yet WSM was up 23.67% over six months while RH and AN were down 18.48% and 21.23%.[5][2]
RH adds an important qualification: second-quarter results included a $55.1 million tariff benefit, or 600 basis points of adjusted EBITDA margin, while normalized adjusted EBITDA was $123.5 million. The headline margin needs to be separated from underlying operating performance.[5]
For AN, ETH and LZB, the evidence is thinner in this pass. Their case improves if vehicle and home-furnishing demand stays durable and input costs normalize; it weakens if high yields and low confidence delay discretionary purchases.
Macro: supportive growth, expensive money, selective risk
- Growth: real GDP was 2.1% year over year and industrial production 2.25%.[4]
- Inflation: CPI was 3.72% year over year.[4]
- Long rates: the 10-year Treasury was 5.1%, with the 10-year/2-year curve at 0.35 percentage points.[4]
- Stress: VIX was 16.87 and high-yield credit spread 3.23%.[4]
This is not an economy that has stopped. It is a market that can reward earnings only after asking how much capital, leverage and consumer confidence those earnings require.
What would confirm or weaken the hypothesis?
| Evidence | Supports | Weakens |
|---|---|---|
| DDOG and SNOW | Usage, durable budgets and guidance converting adoption into revenue | Consumption optimization, budget cuts or guidance lagging AI enthusiasm |
| CRWV | Contracted demand translating into cash flow and lower financing pressure | Higher borrowing costs, delayed capacity monetization or worse credit terms |
| AN, RH, WSM, ETH and LZB | Stable traffic, comparable sales and normalized margins | Promotions, delayed housing turnover or tariff benefits masking demand |
| Macro | Falling long yields without a growth shock | 5%+ long rates with softer sentiment and wider spreads |
The hypothesis is plausible but not proven. The strongest support is software momentum and WSM’s reported performance. The strongest counterevidence is CRWV’s financing sensitivity and RH’s reported-versus-normalized margin gap.
What to watch next
- Earnings checkpoints. DDOG is scheduled for November 4 before the open and SNOW December 2 after the close; both are estimated. AN is scheduled October 30 before the open, RH December 10 after the close, WSM November 18 before the open and LZB November 17 after the close; all are estimated. CRWV is unsupported by the calendar source, and ETH has no confirmed date.[6]
- Cash economics behind AI demand: bookings and usage, capex funding, debt costs, contract duration and free-cash-flow conversion.
- Normalized consumer margins: especially separating RH tariff benefits from repeatable earnings power.
- Rates and confidence together: a lower 10-year yield caused by deteriorating growth would mean something different from a lower yield alongside stable demand.
Resilient demand can still support parts of this scope over the next year, but only where revenue growth survives the full cost of capital and reported margins survive normalization. Friday’s tape did not settle the debate; it made the test explicit.
Sources
- Quote: SPY
- Compare stock charts
- Analysis-Why CoreWeave is in the crosshairs of the AI scare By Reuters
- FRED: Unemployment
- Q2 Earnings Roundup: Williams-Sonoma (NYSE:WSM) And The Rest Of The Home Furnishing and I…
- Get earnings schedule