Meta's "Selling Compute" Crashes AI Hardware? Wall Street Analysis: Don't Panic, This Isn't Overcapacity or a Sector Inflection Point

Original author: Long Yue

Original source: Wall Street News

A piece of news about Meta selling off its excess computing power has placed several of the most sensitive issues in the AI trading narrative on the table simultaneously: Is computing power truly in short supply? Will Meta revise down its capital expenditure? How much longer can Neocloud remain profitable?

Wall Street News mentioned that Meta is formulating a cloud business plan, potentially offering two types of services: one is managed model/API access, similar to AWS Bedrock; the other is renting out "raw computing power," akin to Neocloud.

Upon the news, CoreWeave, a star next-generation GPU cloud service provider, saw its stock price plummet 13%, while Nebius dropped 15%. The broader AI hardware sector, including chips, subsequently suffered heavy losses. If Meta starts selling computing power, investors will naturally ask three questions:

First, did Meta over-purchase computing power?

Second, is Meta reducing its heavy investment in models and AI products?

Third, is the demand curve for AI hardware and Neocloud about to change?

According to information from the trading desk, on July 1st, Wall Street investment banks including UBS, Morgan Stanley, and Bernstein quickly dissected this event. This might not signal the collapse of AI fundamentals, but rather a pragmatic move by a tech giant seeking balance between computing power constraints and financial returns. This matter cannot be simply equated to "Meta no longer needs computing power." However, the implications differ for various assets.

For Meta, leasing out computing power could be a transitional bridge for revenue and EPS. UBS assesses: "Selling cloud computing power or model access rights could theoretically generate near-term revenue faster than waiting for Meta Business Agents and Meta AI chatbots to scale, and could alleviate concerns about flat or contracting EPS in 2027."

For Neocloud companies like CoreWeave, this represents potential competitive pressure.

For the chip and server supply chain, the market is more concerned about whether the subsequent pace of capital expenditure will change.

"Having Surplus Capacity to Lease" Does Not Equal "Industry-wide Computing Power Glut"

The market's shortest chain of logic is: leasing computing power = excess computing power = downward revision of capital expenditure.

Meta may have surplus computing power available for lease on a temporary basis, but this does not automatically translate into an industry-wide glut. Different institutions use different capacity metrics, which cannot be directly summed.

In Morgan Stanley's model, Meta is expected to add approximately 2GW and 3.5GW of its own operational IT capacity in 2026 and 2027 respectively, against a baseline of around 3GW by the end of 2025. By comparison, hyperscale cloud providers like Amazon and Google could add IT capacity on the order of 5GW and 9GW respectively in 2027. In other words, even if Meta dedicates a portion of its own capacity for external leasing, it is unlikely to single-handedly alter the overall landscape of cloud provider construction over the next three years.

Bernstein uses a broader metric of total data center footprint: Meta's current global capacity is estimated at ~20GW, with another ~14GW expected to come online in the coming years, comprising a mix of owned and leased facilities. This number looks large, but it is not "all leasable AI computing power," nor does it represent the same GPU generation, workload type, or pricing curve.

The market's calculations also include a more aggressive back-of-the-envelope estimate: using contracts and capacity plans like Google with Anthropic, AWS with Anthropic/OpenAI, and Microsoft with OpenAI as anchors, the total AI computing power of several major cloud providers in the future could be around 20GW each or even higher. OpenAI's own Stargate project, and its ~10GW-scale arrangements related to NVIDIA and Broadcom, are also factored into the demand side. The purpose of this metric is not to provide precise predictions, but to illustrate a point: Meta's localized leasing is insufficient to prove that global AI construction has entered a phase of oversupply.

More counter-intuitively, Bernstein also noted that there were weekend reports suggesting Google was limiting Meta's computing usage due to its own capacity constraints. If this claim holds, the situation where Meta is simultaneously seeking external computing power while planning to sell some of its own in the future looks more like a reallocation across "different generations, different use cases, and different time windows," rather than a simple case of "having too much to use."

This Isn't the First Time Meta Has Put "Selling Computing Power" on the Table

On May 27, 2026, a shareholder asked Meta whether it would build a cloud business to compete with AWS, Azure, etc. Zuckerberg replied:

"Sure, it's definitely something we consider... We haven't done it yet because we think we can use this computing power ourselves. But obviously, if at some point we feel like we've overbuilt, then it's an option we have, and it's part of the reason we're confident to keep investing in building."

Earlier, on October 29, 2025, Zuckerberg discussed a similar logic:

"For any computing power we don't need, we're pretty confident we can absorb a very large portion of it... Of course, it's possible we overbuild. If we do... we see a lot of new demand both internally and externally. Almost every week, people outside the company come to us asking if we can set up an API service, or asking if they can get different types of computing power from us. We haven't done it yet. But obviously, if you reach the stage of overbuilding, this could become an option."

This explains why UBS calls it "not new news."

For Meta Shareholders, Selling Computing Power is More Like an "EPS Bridge," Not a New Core Business

For Meta, the most direct benefit of leasing out computing power is converting long-term AI investments into near-term revenue.

In UBS's table, Meta's diluted EPS for 2026 and 2027 is approximately $32.6 and $33.0, with market concerns that 2027 EPS might be roughly flat or even compressed compared to 2026. Leasing out computing power or selling model access rights could at least provide a buffer for revenue and profit before Meta Business Agents and Meta AI chatbots truly scale.

Morgan Stanley's sensitivity analysis is more intuitive: leasing out 250MW of computing power for a one-year term at a price of $40/Watt could potentially add about $2.97 to Meta's 2028 EPS, representing roughly 8% upside. If the capacity expands to 500MW, 750MW, or 1000MW, or the price differs, the EPS elasticity would amplify or diminish accordingly.

This is also why the market didn't interpret it solely as bearish. From a Meta shareholder's perspective, Zuckerberg has essentially gained another recourse: if internal AI products can't consume all the computing power in the short term, they can sell it to external AI labs first, recovering part of the investment.

The market also draws an analogy with xAI leasing computing power to Anthropic: 500MW corresponds to $1.25 billion per month, translating to approximately $300 billion/GW/year. If this pricing holds, the implied returns are very high, suggesting that high-quality computing power is still tight in some scenarios. It's not evidence that "nobody wants computing power," but rather evidence that "idle windows can be snapped up at high prices."

However, this can only be called a bridge, not the main narrative. Morgan Stanley still places the key to Meta's valuation on front-line product innovation: whether Meta AI, business agents, messaging, diffusion offerings, subscriptions, etc., can drive more durable engagement and revenue growth. Selling computing power can supplement EPS but won't automatically lift valuation multiples.

Capex May Not Be Revised Down; Building a Full Cloud Could Actually Be More Expensive

The market's biggest fear is that Meta might revise down its 2027 capital expenditure, causing the entire AI hardware chain to lower expectations accordingly.

However, Morgan Stanley's current model assumes Meta's capital expenditure rises from $145 billion in 2026 to $175 billion in 2027 and $205 billion in 2028. This model is premised on Meta primarily building capacity for its own first-line products, rather than creating a full-blown hyperscale cloud service provider.

If Meta truly expands its external cloud services, especially building a model/API platform rather than temporarily leasing bare-metal compute, capital expenditure could actually face upward pressure. A complete cloud business requires longer-term data center capacity, a more sophisticated software platform, and the ability to deliver to enterprise customers.

Bernstein also views this issue from a post-2027 perspective. Meta is one of the most important "checkbooks" in the AI market; any change in its construction pace will impact the supply chain. But "temporary external leasing" and "permanent expansion of the cloud business" have different implications for capital expenditure and should not be conflated.

The larger demand side still lies in inference and agent applications. A market summary from HY Computer & AI Power cited OpenAI's recent article on Codex/agentic AI as a demand signal: the number of individual non-developer users grew 137 times, organizational users grew 189 times, and internal OpenAI users grew 12 times. This perspective highlights that the expansion of new use cases may continue to push up inference computing demand.

Thus, the key point of this divergence is not "whether Meta will sell computing power," but whether the AI demand curve is still steepening. If overseas ARR accelerates, inference applications grow, and cloud provider capex continues to be revised upward, Meta's leasing of computing power looks more like periodic asset monetization. If subsequent earnings seasons show collective downward revisions in capex, then this event could become a signal of an industry inflection point.

Selling Bare-Metal Compute is Easy; Building a Complete AI Cloud is Hard

Meta's potential business has two paths, with vastly different difficulties.

The first path is selling "bare-metal compute" or raw chip capacity, similar to a neocloud. Customers buy GPU/compute resources, and Meta doesn't immediately need to build out a complete enterprise software suite, developer tools, model platform, and sales system.

The second path is offering managed model/API access, similar to AWS Bedrock or Google Vertex AI. This isn't a business that can run just with "data centers and chips." It requires capabilities in models, software stacks, developer experience, enterprise customer sales, and service support.

Morgan Stanley's model is more cautious about the second path. It notes that Meta's Muse model family does not perform prominently on TerminalBench and SWE Bench Verified, which are tests related to coding ability and third-party usage scenarios. If Meta aims to compete with frontier models like Gemini, subsequent models need significant improvement.

This is also why the inference "Meta selling computing power = Meta exiting the model business" is shaky. The potential plans already include model/API access; first-line products like Meta AI, business agents, messengers, diffusion offerings, and subscription revenue remain the core of long-term valuation. The question isn't whether Meta will do models, but whether it can turn its model capabilities into a cloud service compelling enough for external customers to pay for.

In market discussions, some point to Muse Spark, the closed-source strategy, and management adjustments as evidence that Meta is still at the table for models. But these are better suited as items for subsequent tracking. At least from the framework of the three banks, the more certain conclusion for now is: the execution barrier for selling bare-metal compute is low, but the barrier for building a full-stack AI cloud is high.

Is CoreWeave the Biggest "Victim"? A Customer Turns into a Potential Competitor

This shock hit new cloud/GPUaaS companies like CoreWeave most directly.

Bernstein's rating for CoreWeave is Underperform with a price target of $67; for Meta, it's Outperform with a target of $850. Its logic is straightforward: If Meta provides cloud infrastructure externally, it could compete directly with CoreWeave.

More problematic is that Meta itself is a major customer of CoreWeave. According to Bernstein, Meta currently has $35.2 billion in contracts with CoreWeave, accounting for over a third of CoreWeave's order backlog. Combined with approximately $14 billion in contracts from Microsoft, nearly half of CoreWeave's orders come from clients who could become competitors upon future renewal.

Short-term risk isn't that direct. Existing contracts have strong binding clauses, making it hard to exit immediately. Therefore, CoreWeave's short-term revenue and debt pressures may not deteriorate instantly.

The long-term problem is harder to handle. If customers build their own clouds and sell their own computing power, the bargaining power of new cloud companies will decline. Especially during renewals, CoreWeave will no longer just face a demand side, but a potential supply side with money, technology, and data center experience.

JPMorgan's trading desk noted that the market's reaction to CRWV falling 13% and NBIS falling 15% is relatively easy to understand: Meta overnight transformed from a customer into a potential competitor. For chip hardware, the impact is more indirect; for GPUaaS, the impact feels more like a business model stress test.

Why Hardware Fell First: Beyond Fundamentals, There's Crowded Positioning

From a short-term trading perspective, the market isn't just trading fundamentals.

JPMorgan's trading desk split the debate into two sides: one side is whether the Meta news represents a shift in the narrative for CSP capital expenditure and AI compute demand; the other is that overly crowded positioning, deleveraging, and profit-taking amplified the decline. Its inclination is that the latter carries more weight, and determining whether there is truly a fundamental shift requires observing the statements in the upcoming earnings season.

The positioning backdrop is not light. Just past the major index rebalancing, total flow and leverage starting points are high. Over the past four weeks, both long and short positioning increased by +2 standard deviations. July has historically seen deleveraging by hedge funds over the past five years, with changes typically ranging from -1 to -3 standard deviations. Semiconductor and memory stock positioning is near the 100th percentile.

This explains why a single piece of Meta news could hammer the entire AI hardware chain. A crowded trade encountering the narrative "computing power might not be scarce" easily triggers a 'sell first, ask questions later' reaction. The fact that software, crowded shorts, and Chinese ADRs rose by over 1.4 standard deviations on the same day is also consistent with short covering during a deleveraging process.

For subsequent reversal signals, the market is mainly watching a few things: Whether Meta clarifies its position; whether overseas AI application ARR accelerates; whether cloud provider capex continues to be revised upward; and whether Q2 results beat expectations. The key timeframe is concentrated in July and August. The current situation feels more like an observation period rather than a period of consensus