Is the CoreWeave credit crisis more like the 2023 SVB panic than the 2008 subprime crisis?

Original author: Xu Chao

Original source: Wallstreetcn

CoreWeave's unsecured bond yields have surged to around 15%, and its CDS spreads have breached 1,000 basis points, prompting intense market skepticism about systemic risk across the entire AI capital expenditure cycle.

Yet top Wall Street institutions are speaking in unison: this is not a systemic collapse akin to New Century Financial in 2007, but rather more like the episodic panic triggered by Silicon Valley Bank (SVB) in 2023—a very small, structurally fragile instrument is being treated by the market as a barometer for the trillion-dollar AI capex cycle.

Ajay Rajadhyaksha, Global Head of Macro Research at Barclays, notes that CoreWeave's approximately $1.5 billion in unsecured bonds occupy a structurally subordinated position within roughly $35 billion of total debt. Using the price signal from this instrument to negate the entire AI cycle is akin to using a roughly $4 billion regional bank ETF to predict a US recession in 2023. Goldman Sachs, meanwhile, argues that the market has chosen the wrong reference point—the unsecured bond market does not represent CoreWeave's true incremental financing cost.

Historical Precedent: SVB or the 2007 Financial Crisis?

At the heart of this debate is the choice of historical analogy.

In April 2007, New Century Financial, the largest US subprime mortgage lender, filed for bankruptcy, after which the entire credit market gradually collapsed and the S&P 500 fell about 40% over the following two years. In March 2023, Silicon Valley Bank was shut down by regulators, and the market quickly formed a narrative: regional banks are the lifeblood of small and medium enterprise credit, and a credit crunch would trigger a recession. The bond market rapidly priced in 75 to 100 basis points of rate cuts for that year, and the VIX fear index spiked even more than in 2007.

As it turned out, the Fed did not cut rates even once in 2023, instead continuing to hike. US economic growth for the full year far exceeded expectations, and the recession never arrived. The regional bank ETF KRE recovered all its losses, and the S&P 500 rose about 40% over the following two years.

Bloomberg columnist John Authers points out that both crises appeared severe in real time, yet their ultimate trajectories diverged completely. SVB was an over-extrapolated panic, not a harbinger of systemic risk; the bankruptcy of New Century Financial, by contrast, exposed a structural flaw that ran through the entire financial system. CoreWeave's current situation is morphologically closer to SVB.

Rajadhyaksha summarized the core error of the SVB panic: "The mistake was not in recognizing that Silicon Valley Bank had failed. It did fail. The mistake was in extrapolating from a small, volatile, structurally fragile instrument to the entire US economy."

Barclays: A $1.5 Billion "Tail" Wagging the Whole "Dog"

Rajadhyaksha draws a direct analogy between the logical structure of the current market panic and the SVB period.

The "tail" in 2023 was an equal-weighted regional bank ETF with roughly $4 billion in assets under management—KRE—many of whose constituents traded thinly. It was the decline of this small ETF that drove the repricing of trillions of dollars in rates, investment-grade credit, and equity positioning, creating a self-reinforcing panic loop.

The "tail" in 2026 is CoreWeave's approximately $1.5 billion in structurally subordinated unsecured bonds.

Rajadhyaksha lays out CoreWeave's three-tier debt structure: The most significant is the secured delayed draw term loan (DDTL), held in separate special purpose vehicles (SPVs), each backed by specific GPU hardware, customer contracts, and data center leases—when Meta signs a multi-year "pay whether you use it or not" contract and assigns it to an SPV, the loan's substantive collateral is Meta's payment commitment; next is approximately $9 billion in convertible bonds with coupons ranging from 1.75% to 2.875%; at the very bottom are the unsecured notes, behind which hundreds of billions of dollars in claims stand in line at liquidation.

"A 15% yield tells you something about that piece of paper; it cannot tell you that this company is about to cease operations," Rajadhyaksha writes, offering an analogy: "If a person has a $35 million mortgage plus a small credit card, a high credit card rate tells you about the credit card problem, not that every mortgage borrower in the city is in trouble."

The numerical comparison is equally telling: in March 2023, the market priced in 100 basis points of rate cuts on the back of a $4 billion ETF decline; in 2026, the market is questioning a trillion-dollar capex cycle because roughly $1.5 billion in structurally subordinated bonds are trading at 80 cents on the dollar.

Goldman Sachs: The Market Has Chosen the Wrong Financing Cost Benchmark

The core argument of Goldman Sachs analysts Gabriela Borges and Maura Hager likewise hinges on a misreading of the debt structure.

Goldman notes that CoreWeave's primary financing instrument is the DDTL—asset-backed project financing secured by specific customer contracts, GPU infrastructure, and related cash flows. Lenders underwrite the cash flow quality of specific contracts rather than CoreWeave's overall creditworthiness. This structure transfers financing risk from the company's balance sheet to the underlying assets and reduces refinancing risk. The unsecured bond market has been closed to new CoreWeave issuance for months, and using yields from that market to gauge financing costs is equivalent to referencing a price signal that has ceased to function.

Based on annualized Q2 interest expense of $2.56 billion divided by approximately $30 billion in total debt, Goldman estimates CoreWeave's current weighted average cost of debt at roughly 800 to 900 basis points. For incremental financing, Goldman assumes investment-grade customers (such as Meta) correspond to DDTL pricing of approximately SOFR+300 basis points, high-yield customers approximately SOFR+600 basis points, with a blended cost range of 750 to 1,050 basis points, and a conservative baseline of 1,000 basis points—implying incremental financing costs rise only 100 to 200 basis points from current levels.

Goldman also cites market data: DDTL 5.0 and 5.5 are both trading at or above their original issue discount, indicating that confidence in CoreWeave within the asset-backed financing market has not collapsed.

Notably, Goldman discloses in the report's appendix that it is serving as lead bookrunner or joint bookrunner on 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 seek related business over the next three months.

The Fulcrum and Fragility of Goldman's Bull Case: Revenue per Megawatt Needs to Rise 30%

The other half of Goldman's bull thesis rests on expectations of rising computing monetization at CoreWeave.

Goldman estimates CoreWeave's current blended monetization level at approximately $10 million in annual recurring revenue (ARR) per megawatt. Supporting the upside: short-term contracts (3 to 6 months) are signing at $40 million per megawatt; CoreWeave raised prices 25% in July and another 10% in October; contracts signed in Q2 carry contribution margins 5 to 10 percentage points higher than recent contracts. Goldman uses $13 million per megawatt as a "reasonable starting assumption" for new long-term contracts.

Plugging into an illustrative model (5-year, 200 MW contract, 80% debt financing, 10% financing cost, zero residual value) yields an unlevered IRR of 15%, producing a positive spread of approximately 3 percentage points over an 11.8% weighted average cost of capital.

However, Goldman's sensitivity analysis also reveals the fragility of this logic: using the current actual blended level of $10 million/MW, the unlevered IRR drops to 6.3%, turning the spread negative by 6 percentage points at the same financing cost. The entire bull case depends on CoreWeave raising revenue per megawatt by approximately 30% from current levels.

Goldman's own model data also reveals how long this path is: free cash flow is negative $28.2 billion in 2026 and 2027, negative $15.8 billion in 2028, long-term debt climbs from approximately $44 billion this year to approximately $91 billion by 2028, and free cash flow is not expected to turn positive until 2030.

Supply Chain Data: Demand Reality Is the True "Backstop" of the AI Cycle

Rajadhyaksha turns his gaze to the deepest reaches of the supply chain, using hard data to rebut the systemic collapse narrative.

TSMC recently reported revenue exceeding the upper end of its own guidance by $1 billion, marking its sixth consecutive quarter of record profits, with high-performance computing rising to 68% of revenue, and the company raised its Q4 growth guidance. ASML recorded record sales over the same period, with operating profit more than doubling; CEO Christophe Fouquet stated that planned supply for 2027 and 2028 will not fully meet customer demand, and some customers have begun trying to reserve equipment for 2029 and 2030.

The five major hyperscalers (Microsoft, Meta, Alphabet, Amazon, Oracle) will collectively spend approximately $725 billion to $800 billion on AI infrastructure in 2026, with estimates exceeding $1 trillion in 2027. CoreWeave's 2026 capex guidance is approximately $35 billion to $39 billion, representing about 4% to 5% of that total.

"Even if CoreWeave stopped spending entirely, we would only lose 4% to 5% of total AI capex," Rajadhyaksha said. "Hyperscaler spending is funded primarily from free cash flow, not the unsecured bond market. Extrapolating from one emerging cloud provider to one of the most profitable companies in the world—that logical chain is stretched far too long."

Barclays also acknowledges one key difference from 2023: there is no equivalent central bank tool to serve as a safety net. In 2023, the Fed introduced the Bank Term Funding Program (BTFP), allowing banks to pledge underwater bonds at par in exchange for liquidity, which became the key to breaking the panic loop. AI infrastructure financing has no comparable policy tool. But Rajadhyaksha argues that hyperscalers' genuine demand for computing power itself constitutes this cycle's own "backstop logic."

Rajadhyaksha concedes: "The market has a tendency to always find the smallest, most volatile instrument in a complex system and treat it as a barometer for the whole. In 2023 it was KRE; in 2026 it is CoreWeave's unsecured bonds. Both signals were real. The 2023 extrapolation was wrong, and we believe the 2026 extrapolation is equally wrong."

The Real Risk Trigger: Whether Computing Rent Can Hold

Despite top Wall Street institutions collectively trying to "put out the fire," all parties have clearly identified a common risk trigger condition.

Rajadhyaksha points out that whether the AI capex cycle repeats the subprime mortgage playbook hinges on whether computing rents can be sustained at levels sufficient to cover the cost of capital. "If rents fail to be as high as expected, it could trigger a genuine crisis. CoreWeave is an alarming situation, but it cannot provide evidence in this regard."

Goldman's sensitivity analysis likewise points to this core vulnerability: if ARR per megawatt cannot rise from the current $10 million to $13 million, the entire positive spread logic reverses entirely. And once supply catches up with demand and pricing declines, CoreWeave remains locked into data center leases of 7 to 16 years, with high exit costs.

Brian Stutland of Equity Armor Investments believes CoreWeave's problem lies fundamentally in its own business model rather than in AI's overall prospects, but he also concedes it is "two sides of the same coin"—whether this is an isolated company-level issue or reveals broader structural fragility remains "a 50/50 probability."

For investors, the key question is not whether CoreWeave will collapse, but whether computing pricing can support the financing logic of this cycle. The answer to that question will determine the ultimate fate of this panic.