A hyper-growth AI cloud with a $104 billion contract backlog, funded almost entirely with debt the bond market now prices at distressed levels. Demand is not the question the Scorecard raises; the balance sheet is. Price sits below all historical support, 56% off its high, and the next year turns on whether CoreWeave can keep borrowing at a cost its contracts can carry.
CoreWeave rents GPU computing capacity, built on NVIDIA hardware in its own and leased data centers, to the companies training and running large AI models. Customers sign multi-year committed contracts, so most revenue is contracted before a data center opens. The company began as a crypto miner, pivoted to AI cloud, and listed on Nasdaq in March 2025.
Revenue is almost entirely capacity rental. It grew 168% in 2025 to $5.13 billion, and roughly doubled again year over year in each of the first two quarters of 2026. Backlog, the value of signed contracts not yet billed, was $104.2 billion at June 30, 2026, up from $66.8 billion at the end of 2025, with more than $25 billion of new commitments signed early in the third quarter. Microsoft was about two-thirds of 2025 revenue; Meta (an order form worth up to about $21 billion through 2032), OpenAI, NVIDIA and Jane Street are the other large commitments.
Two things changed in the last two quarters. Customer mix broadened — the largest customer fell to 36% of Q2 revenue from 71% a year earlier — and the funding bill rose sharply. Capital spending for 2026 is now guided at $35–39 billion against $6.9 billion of cash, net interest expense reached $640 million in Q2, and in mid-October the company's bonds sold off to yields of 15–16%, taking the shares to a new low.
Scored as a Speculative Growth name. The archetype weights growth, runway, technicals and risk highest and down-weights valuation and margins, which carry little meaning before profitability.
1 · Growth 9. Revenue up 168% year over year — hyper-growth territory, and the main reason the name trades at its multiple.
2 · Margins 3. Gross margin of 67% says the unit economics work; net margin of -25% says the company is nowhere near keeping it.
3 · Balance sheet 1. Net debt of 11.5× EBITDA — leveraged-buyout territory; operating profit does not cover interest (-0.1×); a current ratio of 0.46 means near-term obligations exceed near-term assets; at the current burn, cash on hand lasts about 6 months, so the company depends on raising more — and each raise at today's rates costs more than the last. This is the dimension to watch.
4 · Moat 5. Held neutral; competitive position is an analyst judgment the engine does not make.
5 · Technical 3. Price sits below its 20- and 200-day averages; RSI of 25 is oversold.
6 · Valuation 7. At 4.5× sales the market is not paying up for the name, which reflects the leverage above rather than a bargain.
7 · Risk 2. Annualized volatility of 105% is the highest band we score.
What the Score read cannot show is the shape of the debt. The engine's cash runway treats every dollar of capital spending as burn; in practice most of it is financed by project loans secured against specific customer contracts and repaid from those contracts' cash flows. That makes the six-month figure harsher than the real liquidity position, but it does not make the company self-funding — it needs new debt every quarter to keep building. Principal due is $4.4 billion for the rest of 2026 and $6.2 billion in 2027 (June 30 schedule), and rates run from about 2–3% on convertible notes to 9–10% on senior notes and as much as 15% on the oldest delayed-draw facility.
Moat stays neutral at 5.0. In words: the edge is execution — speed to stand up large GPU clusters and priority access to NVIDIA supply — rather than anything structural. Hyperscalers and newer AI clouds can buy the same chips, and the largest customers are also building their own capacity.
Moat is held neutral (5.0) in every Scorecard; the analyst's moat view appears in words above and does not alter the number.
CRWV is below support: there is no tested support zone beneath $62.13. The former support zone at $75.41–$81.63 now sits about 21% overhead and has been broken since mid-October. The last close at or below this level was $60.82 on Jul 29, 2026, the low before August's earnings rally. With an average daily range of $6.21, about 10% of price, a single session can cover nearly half the distance to former support. These are places buyers have appeared before, not levels to act on.
No established support zone in the 52-week window.
Daily closes, last completed session 2026-10-20. Delayed data; not real-time.
No tested support below the current price. See former support and the last time at this level below.
| MA | Level | Vs price |
|---|---|---|
| 20-Day SMA | $80.52 | -22.84% |
| 50-Day SMA | $85.92 | -27.69% |
| 100-Day SMA | $88.84 | -30.07% |
| 200-Day SMA | $92.30 | -32.69% |
RSI 24.8 · ATR $6.21
$75.41–$81.63 (21.4% above price) was a historical support zone that price has since broken below. Former support often acts as a ceiling on the way back up.
CRWV last closed at or below $62.13 on Jul 29, 2026 ($60.82), about 3 months ago. It has closed below its lowest support zone ($77.58) since Oct 14, 2026.
Shown when price is below all historical support. Daily closes, split-adjusted. Historical context only, not a price target.
The options market is pricing a move of about ±20.7% for CRWV into Nov 13, 2026 (the first expiry after the next earnings date, 23 days out): a range of $49.28 – $74.98 around $62.13. The options market prices this stock to swing about 104% a year; over the 23 days to expiry that works out to roughly the ±20.7% above. This is what buyers and sellers of options are paying for, not a forecast of where the stock goes.
Realized-vs-implied history for CRWV will appear once the earnings log carries reported quarters for this name.
The bull case is about the business: signed contracts worth eight times this year's revenue, a broadening customer list, and project-level margins that work. The bear case is about the funding: every dollar of that backlog has to be built first, the build is financed with debt now priced as high yield, and interest already exceeds operating profit. For the bull case to win, the next two quarters need to show adjusted operating margin climbing toward the low teens while new borrowing gets cheaper, not dearer. For the bear case, it only takes one closed financing window or one slowed customer. Which of those you weight more heavily is your call.
| Period | Revenue | YoY | GAAP GM |
|---|---|---|---|
| FY 2025 | $5.13B | +168% | n/d |
| Q1 2026 | $2.08B | +112% | 66% |
| Q2 2026 | $2.58B | +112% | 66% |
| Q3 2026e (guide) | $3.45–3.6B | +154–165% | n/d |
| FY 2026e (guide) | $12.4–13.2B | +142–157% | n/d |
Gross margin is revenue less cost of revenue; it slipped from 73% in Q1 2025 to 66% in 2026 as new capacity is stood up ahead of billing. Annual and guided margins are not disclosed (n/d). Guidance is from Aug 11, 2026; guided-row growth is our arithmetic against Q3 2025 ($1.36B) and FY 2025.
| Q2 2026 net loss (GAAP) | $(626)M |
| Q2 2026 EPS (GAAP) | $(1.14) |
| Q2 2026 net interest expense | $640M |
| Q3 2026 interest guide | $860–940M |
| Cash & securities (Jun 30, 2026) | $6.9B |
| Debt (Jun 30, 2026) | $35.1B |
| Principal due, rest of 2026 | $4.4B |
| Principal due, 2027 | $6.2B |
| Convertible notes outstanding | $10.8B |
| Shares outstanding (Apr 30, 2026) | 545.6M |
Operating cash flow was $679M in Q2, and equipment purchases were $6.4B, so the quarter's $5.7B gap was filled with $13.5B of new borrowing. Including leases, total obligations are about $52B. The pattern continues through 2026: capital spending of $35–39B against $6.9B of cash means CoreWeave is in the debt markets every quarter. Since June it has added $4.2B of 2.875% convertible notes due 2033 and a $2.6B term loan at SOFR + 5.50% due 2031. The number to track is net interest expense against adjusted operating income: $640M against $128M in Q2. Until the second overtakes the first, every new contract adds to the debt before it adds to earnings.
| Company | Focus | Position vs CRWV |
|---|---|---|
| Microsoft Azure (MSFT) | Hyperscale cloud with its own AI data centers | CoreWeave's largest customer in 2025 and its largest competitor; buys overflow capacity while building its own |
| Oracle Cloud Infrastructure (ORCL) | GPU cloud for model training at large scale | Competes for the same AI-lab contracts with an investment-grade balance sheet and lower funding cost |
| Nebius (NBIS) | AI cloud, Europe and US | Smaller and growing faster — Q2 2026 revenue $582M, up 454% — targeting 5 GW contracted by end-2026 |
| Amazon Web Services / Google Cloud (AMZN / GOOGL) | Hyperscale cloud with custom AI chips | Offer GPU and in-house accelerator capacity; their own chips reduce dependence on NVIDIA supply |
| Lambda (private) | GPU cloud for developers and labs | Same model at smaller scale, also funded with GPU-backed debt |
CoreWeave's edge is speed and scale on NVIDIA hardware: 1.5 GW of active power at Jun 30 and 4.2 GW contracted. That edge is real but not structural — the chips are available to everyone with the capital, and the hyperscalers have cheaper capital. The competitor to watch is Oracle, which bids for the same large training contracts with an investment-grade balance sheet and a funding cost well below CoreWeave's.
At about 4.5 times trailing sales, CRWV screens inexpensive on revenue, and the Scorecard's valuation dimension reflects that. Adding debt to the market value changes the picture: enterprise value is roughly $62B (market value $34.3B plus $35.1B of debt less $6.9B of cash, before leases), or about 4.9 times the midpoint of 2026 revenue guidance and about 3.3 times the year-end run-rate target. The bull range assumes that multiple holds while revenue reaches the run-rate; the base range assumes it compresses as funding stays expensive; the bear range assumes equity absorbs a financing gap. Analysts span a wide range — HSBC at $39 and Goldman Sachs at $106 on Oct 19, 2026 — and price sits below the lowest target in the Scorecard's analyst range. Where the funding cost settles decides which of those you find credible.
Score, grade, support zones, moving averages, RSI, ATR, beta, volume and analyst range are taken unchanged from the BeYourCFO Single-Stock Scorecard dated Oct 21, 2026. Market data is delayed or as of the prior close.
Sample report. Published as an example of the Independent Research Report. Figures were accurate as of the date shown and are not updated; the scoring methodology and price frameworks remain the proprietary work product of BeYourCFO.
This Independent Research Report is produced by BeYourCFO for informational and research purposes only. It does not constitute financial advice, an investment recommendation, or a solicitation to buy or sell any security. All figures reflect publicly available information believed accurate as of the date shown and may contain errors or become outdated; verify independently before acting. Scenarios are estimates with stated assumptions, not forecasts. The BeYourCFO score characterizes business quality and risk; it is not a prediction of price or returns and is directionally validated rather than a proven edge. Market data is not real-time. Past performance is not indicative of future results. BeYourCFO is not a registered investment adviser. Investment decisions are yours to make, independently and, where appropriate, with a licensed professional.
BeYourCFO · CRWV Independent Research · Oct 21, 2026