Goldman Sachs believes investors are overly concerned about how CoreWeave funds its operations while severely underestimating the company's potential to charge for computing capacity.
Goldman Sachs analysts Gabriela Borges and Maura Hager upgraded CoreWeave, Inc. (NASDAQ: CRWV) from Neutral to Buy on Monday with a price target of $106. Given that the stock has fallen 43% since its August 11 second-quarter earnings release and dropped 24% over the past five trading sessions alone to close at $61.39, the $106 target still implies roughly 73% upside.
Citing a "Finance 101" framework directly in their report, Goldman Sachs noted that CoreWeave's enterprise value is highly correlated with the spread between its unlevered internal rate of return (uIRR) and its weighted average cost of capital (WACC), and argued that the market is underestimating the upward revision potential for uIRR while overestimating the room for WACC to rise.
The upgrade carries direct implications for market sentiment.
Over the past week, CoreWeave's unsecured bond yields climbed to 13.5% to 15%, credit default swaps (CDS) hit historic highs, and Deutsche Bank estimated a financing gap of approximately $35 billion over the next 12 to 18 months—factors that together weighed on the stock. Goldman Sachs's core counterargument is that the unsecured bond market does not represent CoreWeave's true incremental financing costs, because that market has been closed to new CoreWeave issuance for months, and its primary financing tool is asset-backed delayed draw term loans (DDTL) secured by GPUs and customer contracts.
Financing costs: Goldman Sachs says the market is using the wrong benchmark
The central point of contention in Goldman Sachs's report is whether the metrics the market uses to gauge CoreWeave's financing pressure are actually accurate.
Goldman Sachs pointed out that CoreWeave's primary financing method is not traditional corporate bonds but DDTL—an asset-backed project financing structure secured by specific customer contracts, GPU infrastructure, and related cash flows. Lenders underwrite not CoreWeave's overall credit profile but the cash flow quality of specific contracts and the collateral value of underlying assets. This structure shifts financing risk from CoreWeave's corporate balance sheet to specific contracts and infrastructure assets while reducing refinancing risk.
Based on this, Goldman Sachs estimates CoreWeave's current weighted average debt cost at approximately 800 to 900 basis points (calculated by annualizing second-quarter interest expense of $2.56 billion divided by approximately $30 billion in total debt). For new financing, Goldman Sachs assumes DDTL pricing of approximately SOFR plus 300 basis points for investment-grade customers (such as financing linked to the Meta contract) and approximately SOFR plus 600 basis points for high-yield customers, with a blended cost range of 750 to 1,050 basis points. Goldman Sachs ultimately adopted 1,000 basis points as a conservative baseline, implying that new financing costs would rise by only 100 to 200 basis points from current levels.
Goldman Sachs also cited market data as supporting evidence: DDTL 5.0 and 5.5 are both trading at prices no lower than their original issue discounts, indicating that confidence in CoreWeave within the asset-backed financing market has not collapsed.
Pricing power: Revenue per megawatt is the linchpin of the entire bull case
The other half of Goldman Sachs's bull thesis rests on expectations for rising revenue per megawatt of computing capacity at CoreWeave.
Goldman Sachs estimates that CoreWeave's current blended monetization level is approximately $10 million in annual recurring revenue (ARR) per megawatt but believes this figure has upward bias over the next 24 months.
Supporting factors include: short-term contracts (three to six months) already signing at $40 million per megawatt; CoreWeave raising prices by 25% in July and by an additional 10% in October; and contracts signed in the second quarter carrying contribution margins 5 to 10 percentage points higher than recent contracts. Combining next-generation GPUs with higher-margin software products, Goldman Sachs views $13 million per megawatt as a "reasonable starting assumption" for newly signed long-term contracts.
Plugging $13 million per megawatt into Goldman Sachs's illustrative model (a five-year, 200-megawatt contract, 80% debt-financed, 10% financing cost, zero residual value) yields a 15% unlevered internal rate of return, creating a positive spread of approximately 3 percentage points over the 11.8% WACC.
However, Goldman Sachs's sensitivity analysis also reveals the fragility of this logic: if calculated at the current actual blended level of $10 million per megawatt, uIRR would fall to 6.3%, turning the spread negative by 6 percentage points at the same financing cost. In other words, the entire bull case depends on CoreWeave increasing revenue per megawatt by approximately 30% from current levels.
Goldman Sachs's supplementary argument is residual value. Its base case assumes GPUs are worth zero after five years, which it describes as conservative, and cites A100 contracts being renewed through 2029 and H100 contracts being renewed at a premium as references. If full residual value in year six is factored in, uIRR would jump to approximately 20% to 25%.
Capital expenditure and free cash flow: self-sufficiency only by 2030
Goldman Sachs also argued that CoreWeave's most capital-intensive phase has passed, with new capacity as a proportion of the installed base having peaked, which will limit the financial drag from new capacity not yet in production.
But Goldman Sachs's own model data reveals the length and difficulty of this path: free cash flow is negative $28.2 billion in both 2026 and 2027, and negative $15.8 billion in 2028; long-term debt will climb from $44 billion this year to $91 billion by 2028; expected interest coverage is only 0.3 times in 2026 and 0.7 times in 2027; and the ratio of net debt to equity is projected to peak above 3,000% in 2027. Goldman Sachs expects capital expenditure to level off in 2028 and free cash flow to turn positive in 2030, at which point CoreWeave would achieve self-sustaining cash generation.
Goldman Sachs also raised a counterintuitive argument: if financing costs rise broadly across all computing providers, supply growth would slow, delaying the arrival of overcapacity and thereby making existing capacity "more valuable rather than less valuable in the near term"—effectively framing credit tightening as a potential positive for CoreWeave.
Price target and valuation: above consensus, at a discount to hyperscalers
Goldman Sachs set its price target at $106, based on 18 times Q5 through Q8 EBIT (down from a previous target of $139 corresponding to 22 times), which it said implies approximately 10 times EBIT three years out—roughly a 30% discount to Oracle and more than a 50% discount to hyperscalers, reflecting CoreWeave's "higher absolute debt levels."
On earnings forecasts, Goldman Sachs is notably above consensus: 2028 revenue estimates are 11% above the average analyst estimate, and 2028 EBIT is 24% above the average, primarily driven by higher per-megawatt ARR assumptions. Goldman Sachs also raised its 2028 earnings per share forecast to $3.36 (previously a loss of $1.15) but lowered its 2026 earnings per share forecast to a loss of $3.80 (previously a loss of $3.64).
A footnote that cannot be ignored
Goldman Sachs acknowledged key risks in its report: once supply catches up with demand, pricing could collapse, at which point contract renewals in years four through six might not cover WACC while CoreWeave remains locked into seven-to-sixteen-year data center leases; hyperscalers such as Microsoft (estimated by Goldman Sachs to account for approximately 12%) could bring computing demand back in-house; and spot pricing of $40 million to $50 million per megawatt corresponds to equivalent capital expenditure, implying a one-year payback period, which Goldman Sachs acknowledged far exceeds normal return levels and which the market may be questioning as to whether such ROIC is sustainable.
Notably, in its disclosure appendix, Goldman Sachs stated that it is serving as lead underwriter or co-manager for a pending underwriting engagement for CoreWeave, holds more than 1% of CoreWeave shares, has received investment banking compensation from CoreWeave over the past 12 months, and expects to continue seeking related business over the next three months.
Additionally, according to prior reports, CoreWeave's founder cashed out approximately $3.3 billion before the stock price fell sharply.
Comments