Author: Bao Yilong
Source: Hard AI
On January 16, JP Morgan's Mark R Murphy research team published a report stating that CoreWeave is at the forefront of explosive demand growth, but investors need to be wary of its high-leverage business model and execution risks.
The report summarizes key signals conveyed by the company's management during a recent investor meeting: AI computing power demand is strong, to the point of being &#34;unsolvable,&#34; customers are locking in longer-term contracts (5-6 years), and are willing to sign contracts 12 months or longer in advance.
CoreWeave has approximately $56 billion in backlog orders, with its 2026 capacity nearly fully allocated. However, this high growth comes with massive capital expenditure requirements and a complex financing structure, with the company needing to invest $27.296 billion in capital expenditures by 2026.
JP Morgan maintains a neutral rating on CoreWeave, setting a December target price of $110, based on an adjusted EBIT valuation of about 33 times for 2026. This multiple is at a premium compared to the industry average of 31 times.
The report acknowledges that the premium reflects the company's high growth prospects and differentiated market positioning, with its capabilities endorsed by top clients such as OpenAI, Microsoft, and NVIDIA. However, the premium is offset by its high debt and the high execution risks necessary to support massive capital expenditures.
The report emphasizes that for investors with a lower risk appetite, this will be a &#34;bumpy journey.&#34;
<h2>AI Computing Power Shortage: Demand Continues to Outstrip Supply</h2>
CoreWeave's management used an extremely strong term to describe the current market situation during the investor meeting: the demand for computing power in the AI field is &#34;unsolvable.&#34; This statement reflects the severe limitations on AI-related computing capabilities relative to customer purchasing willingness.
The report points out two major shifts in customer behavior that confirm this supply-demand imbalance:
<blockquote>
Contract Terms Significantly Extended: The duration of contracts signed by customers is extending from about 3 years to 5-6 years, further supporting the rationale for a 6-year depreciation cycle for GPUs.
Advance Capacity Lock-In: The time for contract signing has extended from 3 months before infrastructure goes live to 12 months or longer. Customers are willing to commit to future capacity well in advance.
</blockquote>
It is noteworthy that although the time for early contract signing has lengthened, this does not mean CoreWeave will deploy capital early.
Once active power is obtained from data centers, the company still expects to convert facilities into revenue-generating GPU computing capabilities within 4-6 weeks, which is an industry-leading turnaround time.
<h2>Capacity Expansion: 2025 Goals in Sight, 2027 Becomes Key</h2>
JP Morgan analysts believe that CoreWeave is likely to achieve its goal of approximately 850 MW of active power by the end of 2025.
More importantly, the available capacity for 2026 has already been largely allocated, either reflected in the approximately $56 billion backlog or in later pipeline discussions with customers
The company's current unallocated capacity is mostly concentrated in 2027, and potential orders for 2027 capacity have also been clarified.
Given that customers are willing to sign contracts with longer delivery cycles, the order conversion for 2027 capacity may begin as early as the first half of 2026.
The report points out that these dynamics reinforce the strong and long-term nature of demand, which is seen as a positive factor for improving CoreWeave's business visibility.
<h2>Financing Strategy: Asset Financing Dominates</h2>
CoreWeave's capital-intensive business model requires a complex financing strategy to support it.
The company stated that 90% of its capital expenditures are supported by asset financing rather than debt financing at the parent company level. This means that changes in the parent company's credit status have little impact on most of its financing capabilities.
The report notes that the company is able to raise debt at more favorable rates and expects this trend to continue. In the future, some contracts may even be financed at the asset level based on revenue and cash flow, with the loan-to-cost ratio potentially increasing from an implied 90% to about 120%.
Additionally, the delays in data center construction caused by third-party developers mentioned in the company's third-quarter financial report are currently progressing in line with the updated guidance.
The report further explains that CoreWeave's management actively manages construction progress, including reserving buffer time in third-party contract completion targets to address potential discrepancies or delays.
The report believes that the market may underestimate the complexity of rapidly building AI data centers in a very short time, with the current main constraint being supply chain challenges such as labor, which is a common issue in the industry.
<h2>Risks That Cannot Be Ignored</h2>
The research report analyzes the potential risks for CoreWeave.
First is the high concentration of customers: most of the revenue comes from a limited number of clients, and nearly all growth in recent years has come from the expansion of existing customers.
Second is supplier dependence: all GPUs in 2024 will come from NVIDIA, with three major suppliers accounting for a significant portion of procurement.
Third is the ongoing massive capital expenditures and debt requirements; if financing capabilities deteriorate, it will directly impact growth and cash flow.
Finally, there are macroeconomic fluctuations. The report warns that any sustained macroeconomic headwinds could prompt buyers to delay purchases.
As a cloud service provider requiring significant customer investment, CoreWeave may be disproportionately impacted. In the event of economic fluctuations, its stock price could suffer severely due to risk-averse sentiment.
JP Morgan emphasizes that CoreWeave is situated at the most certain infrastructure segment of the AI wave, with demand signals strong enough to shift customer behavior paradigms, and its execution capabilities and financing innovations make it a leading player.
However, this high growth is supported by extremely high financial leverage, customer concentration, and macro sensitivity. This is a story that only investors with specific risk appetites should participate in

CRWV

JP Morgan believes that CoreWeave is at the center of the explosion in AI computing power demand, and the company's execution capabilities and financing innovations make it a market leader. However, its growth is supported by a very high customer concentration, massive capital expenditures, and macro sensitivity. The report gives a "Neutral" rating, suggesting that it may offer a "bumpy and highly volatile" journey, suitable only for investors with specific risk preferences

- Morgan Stanley's research highlights CoreWeave's explosive demand growth, indicating a robust AI computing market, while cautioning investors about its high leverage business model and execution risks.  
- The company has a backlog of approximately $56 billion and expects to invest $27.296 billion in capital expenditures in 2026, necessitating complex financing strategies predominantly through asset-backed financing.  
- CoreWeave faces significant risks including concentrated customers, supplier dependency, and economic fluctuations, which could disproportionately impact its stock price and growth if macroeconomic conditions worsen.

CoreWeave Investor Meeting: AI Demand "Unresolvable," Customers Willing to Sign Long-Term Contracts, and Sign 12 Months in Advance