Combined average of Moat (AI Resilience), Growth, and Valuation scores.
Moat Score
Pure-play NVIDIA GPU cloud with first-mover scale, deep NVIDIA partnership, and earliest access to next-gen architectures — but heavily concentrated revenue and asset-financed business model limit moat durability.
CoreWeave's edge is operational velocity and NVIDIA preferred-partner status, not a structural moat:
- NVIDIA Preferred Partner: CoreWeave is repeatedly first-to-market with new NVIDIA architectures (H100, H200, GB200, Vera Rubin) — a status reinforced by NVIDIA's equity stake. This deployment-velocity advantage compounds during platform transitions.
- Hyperscaler Customer Concentration: Microsoft accounted for 67% of FY2025 revenue; OpenAI represents ~33% of contracted future revenue under agreements totalling ~$22.4B (the original $11.9B five-year deal plus up-to-$4B and up-to-$6.5B expansions); Meta's relationship totals $35B through 2032 (the $14.2B September 2025 agreement plus a $21B expansion announced April 9 2026). Three customers = bulk of book — both a strength (visibility) and weakness (renewal risk). Competitive intensity rose further: Meta launched 'Meta Compute' on July 1 2026 to sell raw GPU capacity into the same rental market, and SpaceX is now leasing Colossus capacity to Anthropic and Google at multi-year scale — another neocloud landlord for the same frontier-lab demand CoreWeave serves. Microsoft has already shown willingness to walk: it declined a $12B expansion option in March 2026 and signed a larger agreement with a competing neocloud.
- Asset-Heavy Capital Model: CoreWeave funds GPU buildouts with collateralized debt. Drawn debt jumped to $35.1B at June 30 2026 (recourse current $6.2B + non-recourse current $1.3B + recourse non-current $25.2B + non-recourse non-current $2.4B), against ~$6.9B of cash, restricted cash and marketable securities. August 10 closed the oversubscribed $2.6B DDTL 5.5 (SOFR+5.50%, Ba2/BB+, maturing 2031) — lenders forced a 100–125bp spread blowout vs initial talk and added DSCR covenants, even as year-to-date debt-and-equity raises topped $30B. Net interest came in at $640M in Q2 (below the $650–730M guide) after $536M in Q1, but the absolute carrying cost keeps compounding. The equilibrium works only while contracts are signed faster than depreciation and interest run. The cost of that capital now depends on what stands behind each loan. In September CoreWeave proposed a $3B convertible note and an at-the-market programme of up to 35M shares; by October 15 its five-year CDS hit a record ~1,195bp and its 9% and 9.75% unsecured 2031 notes traded near 79 cents (15.6–16.2% yields), against SOFR+5.50% on the contract-backed DDTL 5.5 two months earlier. Lenders price the customer contracts, not CoreWeave itself.
Moat Verdict
CoreWeave's moat is narrow and time-bounded: NVIDIA preferred-partner status plus operational velocity advantage during AI hypergrowth. Most of the equity moats are N/A, destroyed or weakened — the business is fundamentally a leveraged, concentrated GPU rental operation. The investment case rests on hypergrowth durability (90 score), not moat depth. Suitable as a speculative position sized to risk.
35.0 resilient · 35.0 vulnerable · 80/20 = 35.0 · = 35
Open a moat to read its note.
CoreWeave sells raw GPU cloud capacity to AI labs and hyperscalers, not consumer UI experiences.
CoreWeave's stack runs CUDA workloads on NVIDIA hardware; no proprietary business-logic moat.
CoreWeave does not derive moat from public data access.
Datacenter operations and large-scale GPU cluster engineering are scarce skills, but the talent pool is growing rapidly and large hyperscalers compete aggressively for the same engineers.
CoreWeave bundles GPU compute + networking + managed Kubernetes (now with cross-cloud Interconnect, SUNK Anywhere, and LOTA) for AI workloads, but customers can replicate the stack on AWS/Azure/GCP, Meta Compute, or SpaceX Colossus — the bundle is convenience, not lock-in.
Customer workloads are private; CoreWeave does not derive proprietary data moat from the compute it sells.
no regulatory protection; in fact, CoreWeave faces export-control exposure on GPU re-rentals to restricted geographies.
GPU compute is fungible across providers; no network effect between CoreWeave customers.
Multi-year capacity contracts (~$104B backlog at Jun 30, plus >$25B early-Q3 commitments not yet booked) embed CoreWeave operationally for the contract life, but renewal is open competition with hyperscalers, Meta Compute, SpaceX Colossus rentals, and other GPU clouds — and Microsoft declining its $12B expansion option in March 2026 for a competing neocloud shows the embedding does not survive the contract term.
CoreWeave is not the system of record for any customer's AI workload metadata or training history.
A GPU landlord competing against hyperscalers that buy silicon, power and capital at larger scale and lower cost; CoreWeave's procurement edge is allocation timing, not a unit-cost lead.
Sells a fungible product or capacity: buyers pay the market price, not a brand premium.
Moat in 5–10 years
CoreWeave's moat holds at its low level. Embedding, scale, bundling and talent are already weakened for the reasons a further step would use: hyperscalers buy capacity cheaper, Meta Compute and other neoclouds compete for renewals, and Microsoft's declined expansion option is already in the embedding note. The multi-year backlog keeps the contracts in place for their term, which is why embedding is weakened rather than destroyed. Nothing in view widens a pillar, and the AI-capex risk is charged in growth, not here.
No moat trending · 35 holds
Combined average of Moat (AI Resilience), Growth, and Valuation scores.
Moat Score
Pure-play NVIDIA GPU cloud with first-mover scale, deep NVIDIA partnership, and earliest access to next-gen architectures — but heavily concentrated revenue and asset-financed business model limit moat durability.
Growth Score
Q2 2026 (reported August 11) delivered revenue of $2.575B (+112% YoY, +24% QoQ), the same +112% YoY rate as Q1's $2.08B, and management raised FY2026 guidance to $12.4–13.2B of revenue (from $12–13B), $960M–1.15B of adjusted operating income (from $900M–1.1B) and $18.5–19.5B of year-end annualised run-rate revenue. Adjusted operating margin recovered to 5% from 1% in Q1 and adjusted EBITDA margin was 59%. Revenue backlog reached ~$104B at June 30, excluding more than $25B of net new commitments signed in early Q3. Active power scaled to 1.5 GW (~+500 MW QoQ) with ~3.7 GW contracted, and Vera Rubin NVL72 bring-up completed. The offset is the balance sheet: FY2026 capex was raised to $35–39B, drawn debt reached $35.1B, and Q2 net interest was $640M — the financing loop is still the investment debate.
Valuation Score
At $61.39 (October 16, 2026 close; market cap ~$38B) the stock is down ~33% from $91.72 on October 6 and back at the July trough, ~55% below the unchanged $135 base and ~23% above the $50 bear — valuation score 87. Nothing in the operating picture moved: the Q2 print (revenue $2.575B, adj. op. income $128M, backlog ~$104B plus >$25B early-Q3 commitments) still stands. What moved is the price of money. A hotter September CPI lifted yields, CoreWeave's five-year CDS hit a record ~1,195bp on October 15 and its unsecured 2031 notes traded near 79 cents (15.6–16.2% yields), so the market is now pricing the equity as the residual claim behind ~$35B of drawn debt that costs far more to refinance than it did in August. EV ~$66.6B is ~5.3× the FY26 revenue guide midpoint and ~0.64× backlog. The ladder is held: $50 already assumes a Microsoft renewal miss and covenant pressure, and the stock has not broken it. The bear case is now a refinancing story as much as a customer story; the bull case is backlog converting while the contract-backed facilities (SOFR+5.50% on DDTL 5.5) keep funding the build.
The Speed-of-Deployment Moat
CoreWeave's edge is operational velocity and NVIDIA preferred-partner status, not a structural moat:
- NVIDIA Preferred Partner: CoreWeave is repeatedly first-to-market with new NVIDIA architectures (H100, H200, GB200, Vera Rubin) — a status reinforced by NVIDIA's equity stake. This deployment-velocity advantage compounds during platform transitions.
- Hyperscaler Customer Concentration: Microsoft accounted for 67% of FY2025 revenue; OpenAI represents ~33% of contracted future revenue under agreements totalling ~$22.4B (the original $11.9B five-year deal plus up-to-$4B and up-to-$6.5B expansions); Meta's relationship totals $35B through 2032 (the $14.2B September 2025 agreement plus a $21B expansion announced April 9 2026). Three customers = bulk of book — both a strength (visibility) and weakness (renewal risk). Competitive intensity rose further: Meta launched 'Meta Compute' on July 1 2026 to sell raw GPU capacity into the same rental market, and SpaceX is now leasing Colossus capacity to Anthropic and Google at multi-year scale — another neocloud landlord for the same frontier-lab demand CoreWeave serves. Microsoft has already shown willingness to walk: it declined a $12B expansion option in March 2026 and signed a larger agreement with a competing neocloud.
- Asset-Heavy Capital Model: CoreWeave funds GPU buildouts with collateralized debt. Drawn debt jumped to $35.1B at June 30 2026 (recourse current $6.2B + non-recourse current $1.3B + recourse non-current $25.2B + non-recourse non-current $2.4B), against ~$6.9B of cash, restricted cash and marketable securities. August 10 closed the oversubscribed $2.6B DDTL 5.5 (SOFR+5.50%, Ba2/BB+, maturing 2031) — lenders forced a 100–125bp spread blowout vs initial talk and added DSCR covenants, even as year-to-date debt-and-equity raises topped $30B. Net interest came in at $640M in Q2 (below the $650–730M guide) after $536M in Q1, but the absolute carrying cost keeps compounding. The equilibrium works only while contracts are signed faster than depreciation and interest run. The cost of that capital now depends on what stands behind each loan. In September CoreWeave proposed a $3B convertible note and an at-the-market programme of up to 35M shares; by October 15 its five-year CDS hit a record ~1,195bp and its 9% and 9.75% unsecured 2031 notes traded near 79 cents (15.6–16.2% yields), against SOFR+5.50% on the contract-backed DDTL 5.5 two months earlier. Lenders price the customer contracts, not CoreWeave itself.
Moat Verdict
CoreWeave's moat is narrow and time-bounded: NVIDIA preferred-partner status plus operational velocity advantage during AI hypergrowth. Most of the equity moats are N/A, destroyed or weakened — the business is fundamentally a leveraged, concentrated GPU rental operation. The investment case rests on hypergrowth durability (90 score), not moat depth. Suitable as a speculative position sized to risk.
35.0 resilient · 35.0 vulnerable · 80/20 = 35.0 · = 35
Open a moat to read its note.
CoreWeave sells raw GPU cloud capacity to AI labs and hyperscalers, not consumer UI experiences.
CoreWeave's stack runs CUDA workloads on NVIDIA hardware; no proprietary business-logic moat.
CoreWeave does not derive moat from public data access.
Datacenter operations and large-scale GPU cluster engineering are scarce skills, but the talent pool is growing rapidly and large hyperscalers compete aggressively for the same engineers.
CoreWeave bundles GPU compute + networking + managed Kubernetes (now with cross-cloud Interconnect, SUNK Anywhere, and LOTA) for AI workloads, but customers can replicate the stack on AWS/Azure/GCP, Meta Compute, or SpaceX Colossus — the bundle is convenience, not lock-in.
Customer workloads are private; CoreWeave does not derive proprietary data moat from the compute it sells.
no regulatory protection; in fact, CoreWeave faces export-control exposure on GPU re-rentals to restricted geographies.
GPU compute is fungible across providers; no network effect between CoreWeave customers.
Multi-year capacity contracts (~$104B backlog at Jun 30, plus >$25B early-Q3 commitments not yet booked) embed CoreWeave operationally for the contract life, but renewal is open competition with hyperscalers, Meta Compute, SpaceX Colossus rentals, and other GPU clouds — and Microsoft declining its $12B expansion option in March 2026 for a competing neocloud shows the embedding does not survive the contract term.
CoreWeave is not the system of record for any customer's AI workload metadata or training history.
A GPU landlord competing against hyperscalers that buy silicon, power and capital at larger scale and lower cost; CoreWeave's procurement edge is allocation timing, not a unit-cost lead.
Sells a fungible product or capacity: buyers pay the market price, not a brand premium.
Moat in 5–10 years
CoreWeave's moat holds at its low level. Embedding, scale, bundling and talent are already weakened for the reasons a further step would use: hyperscalers buy capacity cheaper, Meta Compute and other neoclouds compete for renewals, and Microsoft's declined expansion option is already in the embedding note. The multi-year backlog keeps the contracts in place for their term, which is why embedding is weakened rather than destroyed. Nothing in view widens a pillar, and the AI-capex risk is charged in growth, not here.
No moat trending · 35 holds
Growth Analysis
Growth Drivers
Key Risk
The financing loop breaks before the backlog converts. Drawn debt reached $35.1B at June 30 2026 (up from $24.9B at March 31) against ~$6.9B of cash/restricted/securities, FY2026 capex was raised to $35–39B, and DDTL 5.5 closed only after lenders forced SOFR+5.50% and maintenance covenants — proof capital is still available, but at a rising cost. By October 15 2026 the unsecured market had repriced the rest: five-year CDS at a record ~1,195bp (a market-implied ~64% five-year default probability on S&P Global Market Intelligence CDS data), the 9%/9.75% 2031 notes near 79 cents yielding 15.6–16.2%, and an S&P issuer rating of B+, after a September $3B convertible and up-to-35M-share ATM. That only works while contracts are signed faster than GPUs depreciate. Two customer risks compound it: Microsoft (67% of FY2025 revenue) letting its contract decay at 2027 renewal after declining a $12B expansion in March 2026; and Meta — the largest customer at $35B — throttling third-party commitments in favour of Meta Compute (launched July 1 2026), while SpaceX now rents Colossus capacity to Anthropic and Google as another competing neocloud landlord. Either would invert the revenue trajectory while the debt stack stays fixed.
Score Derivation
91.9 base + 2.7 trajectory + 4 margin − 15 risk = 84
Base 91.9 (30–45% CAGR, midpoint 37.5%, decaying from +112% YoY and the ~48% run-rate step into 2027) + 2.7 trajectory (next-gen GPU capacity and sovereign/enterprise accelerating; hyperscaler contracts stable at the same +112% as Q1: (2 − 0) / 3 × 4) + 4 expanding margin (adjusted operating margin 1% in Q1 to 5% in Q2 2026; FY adj. operating income guide raised to $960M–1.15B) − 15 severe risk, held with NBIS on the shared neocloud financing risk. Margin moves from compressing to expanding because the rubric reads the last two quarters; the YoY compression (Q2 2025 ~16%) is already in the base. The NVIDIA refresh driver is restated as the GPU capacity revenue it enables. The formula computed 80 on the old inputs = 84
Price Scenarios (12–24 Months)
Price vs Targets
Dashed lines mark the 12–24 month bear, base and bull targets. Daily closes.
Valuation Multiples
| Trailing P/E (GAAP) | N/M |
| Forward P/E (NTM) | N/M |
| EV / Sales (NTM) | ~5.3× |
| Price / Sales (FY26) | ~3.1× |
| EV / Backlog | ~0.64× |
| Net debt / mkt cap | ~73% |
Traditional P/E remains not meaningful — CRWV is GAAP loss-making. EV is built on $35.1B drawn debt at June 30 2026 net of ~$6.9B cash, restricted cash and marketable securities; the September convertible and ATM are not yet in a balance sheet. At $61.39 the equity has given back the post-print rally while the debt stack stayed fixed, so net debt is ~73% of equity value and EV/Backlog is ~0.64×. The October repricing is in the credit market, not the backlog: unsecured notes near 79 cents mean the next refinancing costs well above the 9–9.75% coupons on the existing 2031 notes, which is a cost the equity absorbs first.
Approximate figures as of October 16, 2026.
Where We Are vs Targets
Loading live price…
Microsoft accelerates in-house Azure AI infrastructure and lets CoreWeave contract decay at renewal; AI capex digestion delays new bookings; debt covenants tighten as GPU depreciation runs faster than backlog recognition.
- Microsoft contracts (67% of FY2025 revenue) are not renewed at full scale post-2027 as Azure builds out internal AI capacity with Maia ASICs and AMD MI400 GPUs, removing $8B+ of run-rate revenue
- Meta scales 'Meta Compute' (announced July 1 2026) and SpaceX Colossus rentals to Anthropic/Google redirect AI infrastructure spend away from CoreWeave, stalling renewals on the $35B Meta book and overlapping Anthropic demand
- AI capex digestion in H2 2026–H1 2027 stops new hyperscaler bookings; backlog stalls near $100B despite the early-Q3 $25B pulse and the multiple compresses toward 3× forward sales as growth decelerates
- Net interest — $640M in Q2 on a $35.1B drawn stack — keeps compounding while GPU depreciation runs ahead of revenue recognition, triggering covenant pressure on newer facilities (DDTL 5.5 DSCR) and equity dilution at a depressed share price
FY2026 revenue lands in the guided $12–13B range (~140% growth); customer concentration eases as OpenAI and Meta ramp through 2026; NVIDIA Vera Rubin transition reinforces preferred-partner status; first GAAP profitability glimpsed late 2027.
- FY2026 revenue lands in the guided $12–13B range with Microsoft share declining toward 50% as OpenAI (~$22.4B contracted) and Meta ($35B through 2032) capacity ramps, supported by the >$25B early-Q3 commitments and broadening enterprise logos
- Backlog grows to $125–140B by year-end as the early-Q3 $25B+ converts into backlog and sovereign/enterprise plus Vera Rubin bookings continue, providing 8+ years of forward revenue visibility
- Adjusted operating margin continues the Q2 inflection (5% adj. op. income margin, $128M) toward low-double-digits by Q4 2026 as guided previously, though GAAP profitability remains 12–18 months away
Sovereign AI and enterprise deployments accelerate; CoreWeave wins disproportionate share of Vera Rubin capacity; customer concentration de-risks materially; multiple re-rates on path to GAAP profitability.
- Sovereign AI and enterprise deployments accelerate ($150B+ globally) and CoreWeave captures a disproportionate share of Western non-hyperscaler deployments, converting the early-Q3 $25B+ pulse into a sustained booking run-rate and adding $15B+ to FY2027 backlog
- Vera Rubin capacity at CoreWeave reaches 2× the pace of any rival (including hyperscaler in-house programs and SpaceX Colossus), entrenching the NVIDIA preferred-partner economic moat through 2028
- GAAP profitability is achieved by Q2 2027, the multiple re-rates to ~9–10× EV/Sales on de-risked unit economics, and equity dilution risk fades as FCF inflects positive
CRWV: common questions
Is CoreWeave stock a buy?
We rate CoreWeave Speculative Buy, with a composite score of 63/100: moat 35, growth 84 and valuation 87. The valuation score is struck against our bear, base and bull targets and moves with the price.
What is CoreWeave's moat?
CoreWeave scores 35/100 on moat. Pure-play NVIDIA GPU cloud with first-mover scale, deep NVIDIA partnership, and earliest access to next-gen architectures — but heavily concentrated revenue and asset-financed business model limit moat durability.
How fast is CoreWeave expected to grow?
We estimate 30–45% a year over the next three to five years, for a growth score of 84/100.
What is CoreWeave's price target?
Our base case is $135, with $50 in the bear case and $240 in the bull case.