AI Earnings Hold Up; IPO Liquidity Still Has to Clear the Tape

The signal: growth is real, but uneven—and financing conditions matter

The one-year hypothesis is that earnings growth and resilient demand can support DDOG, SNOW, CRWV, AN, RH, WSM, LZB and ETH. The evidence available by November 3 is not a single broad demand story. AI-related cloud and compute indicators are strong; home and furniture results range from positive retail metrics to cautious, execution-dependent outlooks; AutoNation combines buybacks with a mixed operating picture; and ETH has faced meaningful spot-fund outflows. Meanwhile, October U.S. IPO activity was unusually light and aftermarket results were narrow.

That combination matters for market structure. Public companies compete for capital not only with one another but also with new listings, insider resales, buybacks and cash flows through ETFs. A convincing growth story can attract capital, but an IPO calendar’s health is not measured by headline deal size alone: breadth, deal completion, post-listing trading and the amount of freely tradable stock all matter.

Issuance: October’s small cohort, November’s thin opening

Renaissance Capital counted eight U.S. IPOs raising $1.2 billion in October, plus one direct listing—far below its 10-year October average of 22 IPOs and $5.1 billion. Biotechs accounted for seven of the nine listings. Traditional IPOs averaged an 8% return from offer, yet nearly two-thirds ended below issue; the Renaissance IPO Index fell 10% during the month versus a 1% decline for the S&P 500. One large deal, Lycia Therapeutics, rose 67% on its debut, but that standout did not describe the typical deal.^[1]

The near-term calendar looked similarly hesitant: an IPO-market report dated November 2 said no U.S. IPOs were scheduled to price during the week beginning November 2, with volatility and the U.S. midterm election among the cited backdrop. This is a short-window snapshot, not proof that the full-quarter pipeline has vanished.^[2]

The contrast is useful: a much-discussed potential mega-listing can attract attention while the ordinary IPO market remains selective. Issuers need to ask whether real buy-side demand will absorb shares without a weak aftermarket, rather than treating a large addressable market or a famous name as a substitute for price discovery.

AI-linked demand: good reported growth, with duration still to prove

The strongest hard operating evidence in this group comes from cloud software. Datadog’s second-quarter 2026 revenue grew 36% year over year to $1.12 billion, and its count of customers with more than $100,000 of annual recurring revenue reached about 4,720, up from about 3,850 a year earlier. Those metrics support the view that spending is broadening beyond a few experimental projects, although they do not establish that growth will persist at the same pace.^[3]

Snowflake reported fiscal Q2 2027 product revenue of $1.49 billion, up 37% year over year, and a 126% net revenue retention rate. It also reported 828 customers with more than $1 million in trailing-12-month product revenue, a 27% year-over-year increase. Retention and large-customer counts lend more support to usage expansion than a single quarter’s headline sales alone.^[4] Reuters reported that Snowflake raised its annual product-revenue outlook following demand for its cloud data platform and AI offerings.^[4]

For CoreWeave, the supportive evidence is more paired with capital intensity and contract-duration questions. The company said it continued signing compute-capacity contracts at higher prices and that Q3 included customer contracts with terms of roughly three to six months. Higher contracted pricing signals demand, but short durations leave more revenue exposed to renewals, utilization and customer spending cycles.^[5]

Consumer and housing-linked names: resilience depends on the business model

The consumer-facing evidence is mixed rather than uniformly weak. AutoNation’s Q3 2026 release reported adjusted EPS of $5.11, a seventh consecutive year-over-year quarter of adjusted EPS growth, and $620 million of year-to-date share repurchases that reduced shares outstanding by 9%; it also announced an additional $1 billion authorization. The company highlighted record after-sales gross profit and 6% customer-pay growth. Buybacks can support per-share results and absorb stock, but they are not a substitute for durable growth in the underlying vehicle-sales business.^[6]

Williams-Sonoma’s fiscal Q2 2026 comparable brand revenue increased 6.2%, with GAAP operating margin of 22.9% and non-GAAP diluted EPS of $2.10; the company raised its full-year outlook. That is a notably firmer datapoint for discretionary home retail, though one quarter cannot establish a sector-wide housing recovery.^[7]

RH’s fiscal Q2 revenue grew 2.6% to $922.2 million, but the company explicitly said its adjusted EBITDA margin included a $55.1 million tariff benefit, equal to 600 basis points. RH’s Q3 revenue-growth outlook was 5%–6%, and its much stronger Q4 outlook depended in part on backlog conversion, RH Estates and new galleries. The company’s own disclosures identify backlog timing, inventory availability, tariffs and housing-market conditions as important uncertainties. That makes the growth case an execution test, not just a consumer-confidence bet.^[8]

La-Z-Boy reported Q1 fiscal 2027 retail written sales growth of 16% and written same-store sales growth of 3%, while consolidated sales fell 3% year over year, partly reflecting the divested wholesale casegoods business. Its retail indicators suggest pockets of resilience; the consolidated comparison shows why segment mix and comparable bases must be checked before calling a broad rebound.^[9]

Crypto flows and secondary-market supply are separate liquidity channels

ETH is not an operating company and should not be assessed through earnings growth. Its relevant signal here is the marginal flow of capital into and out of listed products and the token market. A report published November 3 said U.S. spot Ethereum ETFs saw about $139.8 million of net outflows on November 2, extending a run to six consecutive trading days, with reported net assets of $13.18 billion. A separate report put October outflows at $1.61 billion, described as a record monthly withdrawal since the products launched. These data point to a near-term headwind in one regulated access channel, not a complete measure of ETH demand or on-chain liquidity.^[10]

For newly listed equities, lockup releases and follow-on sales add another supply channel beyond IPO issuance. A third-party calendar listed 14 U.S. IPO lockups ending in November, including Cerebras, Innio and Quantinuum; the dates and share counts should be checked against each issuer’s prospectus and filings because lockup conditions can include exceptions or early-release triggers. An unlock makes shares eligible for sale; it does not mean holders will sell them.^[11]

Exchange rules: one floor changed, but it is not the whole listing test

Nasdaq’s amended Capital Market initial-listing rule establishes a uniform $4 minimum bid price and removes the previous lower-price alternatives. The SEC’s Federal Register notice dated October 13 documents the rule change. This is a specific initial-listing screen for the Nasdaq Capital Market—not a general rule for all Nasdaq tiers, a promise of aftermarket liquidity, or a proxy for operating quality.^[12]

The broader plumbing checklist remains unchanged: assess the offered float relative to shares outstanding, concentrated ownership, market-maker and investor depth, lockup terms, any follow-on supply, and whether a company’s cash needs could lead to additional issuance. A minimum listing price can shape the candidate pool, but it cannot manufacture continuous two-sided demand.

What would confirm—or weaken—the one-year thesis?

The base case is therefore conditional: the cloud and AI companies show the clearest current evidence of growth, but the consumer names do not yet form a uniform demand cohort, and crypto has a different flow-based test. October’s IPO statistics also argue against assuming that growth narratives automatically translate into easy issuance or liquid aftermarket trading. For the optimistic case to hold, earnings and cash generation need to validate the growth stories while secondary supply remains digestible. For the cautious case, even genuine demand could coexist with a poor IPO tape if risk appetite, float dynamics or new supply turn against it.

What to watch next

• The next reported quarters from DDOG and SNOW: look beyond revenue to customer growth, retention, usage and management’s demand commentary.
• CoreWeave’s disclosed contract duration, capacity utilization and financing needs; demand growth and capital intensity should be read together.
• Comparable-sales trends and guidance quality at AN, RH, WSM and LZB, separating recurring operating gains from buybacks, tariff benefits, backlog conversion and business-mix effects.
• IPO pricing, completed deals and post-listing performance—not only the pipeline headline—and each issuer’s actual free float and lockup provisions.
• ETH spot-fund flows alongside other market-liquidity measures; a single ETF-flow streak is informative, but insufficient on its own to establish a lasting trend.
• The application of Nasdaq’s updated Capital Market threshold and any further exchange or SEC changes affecting initial listings or continued trading.

This is research and education, not investment advice.