CoreWeave, Inc. (NASDAQ: CRWV), initiation of coverage, Nov 21, 2026
A $133 billion backlog doubles sales, but capital costs leave no upside
CoreWeave rents AI computing power to the largest software companies and AI labs across 52 data centers. Its contracted backlog doubles revenue in 2027, but renewal prices after 2027 and a high cost of capital decide what the shares are worth.
- Our rating
- Hold
- 12-month target
- $47.78
- Price, Nov 18, 2026
- $47.83
- Expected return, with dividend
- +0%
By Penmark Team
What CoreWeave is now
Nearly all sales come from multi-year contracts as capacity scales fast
CoreWeave leases powered data-center shells, fills them with NVIDIA chip clusters and sells the capacity on contracts typically four to five years long. It runs 52 sites with 1.7 gigawatts of live power and 4.6 gigawatts contracted.
Ninety-eight percent of third-quarter sales came from committed contracts. A $133 billion backlog lifts sales from $13.2 billion in 2026 to $26.9 billion in 2027.
Share of third-quarter 2026 revenue from committed contracts. Other sales are the remainder. Source: CoreWeave filings; Penmark Research.
The fiscal year ends on 31 December. Source: CoreWeave filings; Penmark Research estimates.
What you are really buying
The $48 price already sits above our $41 base value
We discount cash flow and treat customer prepayments as a claim on the business. That gives a $41.02 base value, or $42.17 weighting the three cases, against the $47.83 close.
The price implies fleet revenue of about $14.4 million per megawatt by 2030, or a 9.94% discount rate against our 10.18%. A 14 times 2028 operating-profit check gives $53.75, but skips the later chip-refresh bill.
Values per share at the 18 November 2026 price date. DCF is of free cash flow to the firm excluding customer prepayments; the check applies 14.0x EV/EBIT to 2028 adjusted operating income. Source: Penmark Research model.
The bull case
Strong 2027 renewals could more than double the shares
Older Ampere and Hopper clusters re-contracted at 30% to 50% premiums in the third quarter. Over 130,000 paid-off chips come up for renewal in 2027, worth about $4 billion a year with no new debt.
If fleet revenue reaches $15.0 million per megawatt by 2030 and live power 6.5 gigawatts, the shares are worth $98.08 today and $111.11 in 12 months. We give that a 20% weight.
Fleet revenue per megawatt is annual sales divided by average active power. Source: Penmark Research model.
The bear case
A 2027 supply wave could leave the shares worth nothing
New hyperscaler and AI-lab capacity could cap renewal prices near $11 to $12 million per megawatt. Then we would stop refreshing the fleet from 2029, and run-off value of $31.8 billion would sit below $53.6 billion of claims ahead of the shares.
That case carries a 25% weight, anchored by senior notes valued at 87.8% of par. Each $0.5 million per megawatt price cut costs $37 a share.
Claims are debt principal, leases and customer prepayments, less cash, deducted at par. Source: Penmark Research model; CoreWeave filings.
What decides it
A one-point change in the discount rate flips the call either way
Secured loans and about $9 billion of near-term prepayments should fund 2027 spending without unsecured bonds. But the cost of capital still decides the rating.
A 1-point lower discount rate would lift our target to $77, a Buy. A 1-point higher rate would cut it to $25, a Sell, and a 3.71 equity beta to $15. That is why conviction is Low.
WACC is the weighted average cost of capital, 10.18% in our base. Source: Penmark Research model (sensitivities).
The odds
Our 12-month target sits in line with today’s price
We weight a $0 bad case at 25%, a $46.47 middle case at 55% and a $111.11 good case at 20%. Those 12-month values weight to our $47.78 target, 0.1% below the $47.83 close.
On today’s $42.17 value the shares look about 12% expensive, but rolling forward at a 13.3% cost of equity offsets that, leaving a Hold.
- Good, 20%Renewals lift fleet pricing to $15.0m per megawatt by 2030, live power reaches 6.5 gigawatts and build costs are $35m per megawatt.
- Middle, 55%Fleet pricing reaches $14.3m per megawatt by 2030 then fades to $13.5m, with 1.08 gigawatts added in 2027 at $36m per megawatt.
- Bad, 25%A supply wave caps pricing near $11 to $12m per megawatt, the fleet runs off from 2029 and claims exceed the run-off value.
Circle size shows how likely we think each case is. Case values are 12-month values; the target weights all three, rolled forward at a 13.29% cost of equity with no dividend. Source: Penmark Research.
What would change our mind
December’s loan and February’s renewals are the first rating tests
A $4 billion or larger investment-grade loan at 7.0% to 7.5% would ease funding stress. Late-February results bring the first 2027 sales guide and renewal prices on more than 130,000 paid-off chips.
Renewals holding premiums and a 1-point lower discount rate would point to a Buy. Renewal prices sliding toward $11.9 million per megawatt, or a 1-point higher rate, would point to a Sell.
We expect at least $4bn at 7.0% to 7.5%, adding about $6 a share.
We model $26.9bn of 2027 sales against $26.5bn on the Street, and capex of $39bn against $47bn.
A data-center permitting pause is a $3 a share risk if it delays deliveries.
Peer financing gaps that steer demand to scaled operators could add $20 a share.
Inside the full report
What each page covers after the summary on pages 1 and 2. Each opens the full report at that page.
See all 24 pages
- 3Adjusted EBIT margins double from FY26E to FY28E; our FY28E EBIT sits 9% above consensus on lower capex
- 4At 8.5x run-rate EBITDA, CoreWeave trades at a 45% discount to cloud peers, reflecting debt and asset lives
- 5Why we rate CoreWeave Hold: surging backlog meets a $37-per-share pricing tug-of-war and rate risk
- 6Four named debates decide whether CoreWeave compounds to $98 or erodes toward zero
- 7From a 2017 crypto miner to a $133bn AI backlog, funded by $48bn of debt
- 8High EBITDA margins mask depreciation equal to 51% of revenue and capital intensity
- 9Revenue equals active power times pricing per MW; six-year GPU refresh caps cash flow
- 10Power scarcity supports near-term pricing, while a 2027–2028 supply wave looms
- 11CoreWeave leads neoclouds on scale and software, but trails hyperscalers on funding cost
- 12Insiders hold 72% of voting power via 10:1 Class B stock; most capital funds buildout
- 13CoreWeave trades on funding access and pricing power, not near-term GAAP earnings
- 1473% of weekly moves are company-specific; funding and pricing dominate value
- 15Contracted backlog doubles revenue, but post-2027 renewal pricing decides equity value
- 16Secured DDTLs and prepayments cover 2027 capex, leaving WACC as the rating swing
- 17Committed capex lifts FY28E EBIT 9% above Street, but six-year GPU refresh caps value
- 18Reported operating cash flow flatters cash generation via about $12bn of customer prepayments
- 19Revenue reaches $43bn by FY28E as active power hits about 3.9GW and EBIT margins double
- 20Net debt peaks at about $71bn in FY28E before dropping to 1.3x EBITDA by FY30E
- 21DCF values CoreWeave at $41 vs $54 at 14x FY28E EBIT; 6-year refresh explains the gap
- 22At $48, the market prices slightly more than our base case; Hold is balanced
- 23December DDTL and late-February FY27 guidance are the decisive rating triggers
- 24Short interest climbed to 16% of Class A while 42 Street analysts average a $123 target
- 25Annual financial statements (FY24A–FY30E) and quarterly progression (3Q25A–4Q26E)
- 26Methodology, ratio definitions, glossary and source log (as of 18 November 2026)
Penmark Research. This summary draws on our initiation report on CoreWeave, Inc., published Nov 21, 2026. It is not investment advice. Analyst certification, rating definitions and important disclosures are in the full report.