/The%20CoreWeave%20logo%20displayed%20on%20a%20smartphone%20screen_%20Image%20by%20Robert%20Way%20via%20Shutterstock_.jpg)
Independent investment research firm Hedgeye, popular for its high-conviction short ideas, has focused on newly listed CoreWeave (CRWV) as its new short idea. The firm has cited the company’s dependence on chip giant Nvidia (NVDA) and tech titan Microsoft (MSFT) as reasons for its call. High debt levels and a lack of visibility of profitability in the short term have also been raised as concerns for the company.
So, is there merit in Hedgeye’s assertions or is the firm being unnecessarily bearish about the company? Let’s try and find out.
About CoreWeave Stock
Founded in 2017 as Atlantic Crypto, a cryptocurrency mining venture, CoreWeave is a specialized cloud provider focusing on GPU-accelerated workloads for applications such as artificial intelligene (AI) and machine learning. The company, which transitioned in 2019 to cloud computing services, now provides cloud infrastructure optimized for GPU-intensive tasks, catering to industries requiring substantial computational power. Its market capitalization currently stands at $18.9 billion.
In terms of the share price, CRWV stock is trading roughly in line with its $40 IPO price.
Unpacking CoreWeave’s Financials
As per a recent filing, CoreWeave has seen its top line grow at a remarkable pace, with revenue surging from $15.8 million in 2022 to a staggering $1.9 billion by the end of 2024. Despite this extraordinary expansion, the company’s net losses have deepened — from $31.1 million in 2022 to $863.4 million in 2024. However, a closer look reveals a more encouraging trend: CoreWeave turned operationally profitable during this period. While it reported an operating loss of $22.9 million in 2022, it delivered an operating profit of $324.3 million by 2024, reflecting an operating margin of 16%.
One of the more telling indicators of future growth potential is the company’s remaining performance obligations, which stood at $15.1 billion as of Dec. 31, 2024. This figure marks a 53% increase over the previous year and serves as a proxy for committed but not yet recognized revenue — an important signal of strong demand visibility.
Overall, as of the end of 2024, CoreWeave’s GPU count stood at 250,000 with plans to add 350,000 more GPUs in 2025.
Cash flow from operations has also significantly strengthened, rising from $910 million in 2022 to $2.7 billion in 2024, suggesting improved cash-generating efficiency in tandem with revenue growth. Liquidity, too, has improved meaningfully, with CoreWeave closing 2024 with $1.4 billion in cash, a sharp rise from just $217.1 million a year earlier.
That said, the balance sheet reveals a key vulnerability: the company carries $5.5 billion in long-term debt, far outstripping its current cash reserves. This debt load introduces notable solvency concerns and remains a critical factor for investors to monitor going forward.
Some Drivers, Some Headwinds
CoreWeave’s moat is technical, centered on delivering tangible performance improvements that directly impact end users. At the heart of this lies a substantial disconnect between the theoretical peak performance of GPUs and the results actually achieved in AI applications — a discrepancy CoreWeave terms the “AI Efficiency Gap.” Although GPUs are engineered to operate at full capacity, actual utilization in traditional cloud setups tends to hover around 35% to 45%, largely due to mismatched system architecture, inadequate networking infrastructure, and inefficient orchestration of workloads.
CoreWeave addresses this problem by designing a cloud platform purpose-built for AI, integrating finely tuned hardware arrangements, tailored software layers, and ultra-low latency interconnects. The result is a significantly narrower efficiency gap, enabling faster AI model training, cost reductions, and superior use of compute resources. For customers where speed, scale, and performance are critical, this level of optimization provides a clear competitive edge.
Meanwhile, as highlighted in my pre-listing analysis of CoreWeave and also raised as a concern by Hedgeye later, the company’s customer concentration risk is noteworthy with Microsoft (MSFT) contributing 35% of its revenue in 2023 and 62% in 2024. Despite this reliance, its client portfolio includes other notable names such as Meta Platforms (META), International Business Machines (IBM), and Mistral AI, offering a degree of diversification.
Moreover, the AI cloud infrastructure market, in which CoreWeave operates, is witnessing a significant boom. The company’s filing states:
“According to Bloomberg Intelligence, the market for AI inference/fine-tuning, AI workload monitoring, and training infrastructure, including AI servers, AI storage, training compute, cloud workloads, and networking, will increase by over $300 billion from 2023 to 2028, growing at a CAGR of 38% from approximately $79 billion in 2023 to approximately $399 billion by 2028. This market opportunity is expected to include $330 billion related to training infrastructure, which includes AI servers, AI storage, training compute, LLM licensing revenue, cloud workloads, and networking; $49 billion related to inference infrastructure; and $20 billion related to workload monitoring, all of which are supportable by CoreWeave.”
Moreover, experts report that the global cloud AI market will grow from $80.3 billion in 2024 to $327.5 billion by 2029, while others report the market could be worth about $400 billion by 2030.
As AI spending increases, the addressable market for CoreWeave is sizeable.
A recent nearly $12 billion agreement with ChatGPT-maker OpenAI has been an exciting development and so have been the murmurs about a deal with Google (GOOGL) wherein the search giant would be renting Nvidia’s Blackwell chips from the company. However, no further developments have been reported on this rumored deal.
CoreWeave is contending with formidable competitive pressure from Nebius Group (NBIS), a rival that enjoys the benefit of a more diversified business model, significantly lower pricing, and a much leaner balance sheet. In contrast to CoreWeave, Nebius’ financial structure is less burdened by debt, giving it greater operational flexibility and resilience.
Adding to CoreWeave’s challenges is the pressing threat of technological obsolescence. The majority portion of its GPU inventory — around 250,000 units — consists of Hopper architecture chips, which belong to the previous hardware generation. This concentration heightens the company’s exposure to rapid shifts in AI compute standards. Compounding the risk, CoreWeave is currently consuming roughly $5 billion in free cash flow annually, a pace that is clearly unsustainable. With its current liquidity position, the company will soon need to secure fresh capital to continue funding operations and future expansion.
