Meta plans to lease AI capacity

Meta’s reported plan to lease part of its artificial intelligence computing capacity rattled technology stocks, with CoreWeave shares dropping 13% and Nebius falling 15% as traders reassessed competitive dynamics across the AI infrastructure sector.

market reaction hits ai infrastructure stocks

The sell-off extended beyond cloud providers, dragging semiconductor and hardware names lower as crowded positions unwound. Trading desks linked the move largely to technical repositioning and seasonal deleveraging rather than a clear संकेत that demand for AI hardware is weakening.

CoreWeave faced the sharpest pressure given its exposure to Meta, which accounts for roughly one-third of its 35.2 billion dollar contracted revenue. Traders are now weighing the risk that direct leasing by Meta could erode CoreWeave’s pricing power when contracts are renewed.

leasing seen as monetization, not a pullback

Analysts broadly framed Meta’s move as a financial adjustment aimed at monetizing its vast infrastructure rather than signaling reduced investment. The company is still expected to significantly expand spending, with Morgan Stanley estimating capital expenditure could reach 175 billion dollars in 2027 and 205 billion dollars in 2028.

Bernstein estimates Meta’s computing footprint at around 20 gigawatts, with another 14 gigawatts planned in the coming years. Only a portion of that capacity would likely be available for leasing, and it spans mixed hardware generations rather than a uniform pool of high-end GPUs.

potential earnings boost from compute leasing

Projections suggest leasing could provide a meaningful lift to earnings. Morgan Stanley estimated that renting out 250 megawatts at 40 dollars per watt for a year could increase 2028 earnings per share by about 2.97 dollars, or roughly 8%.

UBS modeling indicates such a strategy could also support 2027 earnings, which are expected to hover near 33 dollars per share, helping offset stagnation in other parts of the business.

supply constraints still evident

Despite the leasing plan, supply pressures in AI infrastructure remain. Bernstein noted reports that Google has limited Meta’s access to certain computing resources due to its own constraints, pointing to ongoing bottlenecks rather than excess capacity.

Meta’s internal expansion plans include about 2 gigawatts in 2026 and 3.5 gigawatts in 2027, adding to roughly 3 gigawatts expected by the end of 2025. These additions are modest compared with other hyperscalers, which are projected to add between 5 and 9 gigawatts each by 2027.

broader implications for cloud and ai markets

Analysts emphasized that leasing spare capacity differs significantly from building a full-scale enterprise cloud platform. Developing competitive model and API services would require deeper software integration and sustained infrastructure investment.

At the same time, speculation has emerged that stronger share performance could revive discussions around potential equity issuance if Meta’s valuation climbs above 20 times expected 2027 earnings, though no plans have been confirmed.

crypto markets reflect broader risk sentiment

Shifts in AI and technology sentiment have echoed in cryptocurrency markets, where risk appetite has remained fragile.

• Bitcoin is trading around 58,000 to 59,000 dollars after hitting a 21-month low, with heavy outflows from spot ETFs in recent weeks
• Ethereum remains слаб near 1,577 dollars, struggling below key resistance levels
• The Crypto Fear and Greed Index recently dropped to 11, signaling extreme fear not seen since 2022

Options market positioning suggests potential volatility ahead, with a gap between bullish derivatives positioning in Ethereum and weak spot price action raising the risk of further liquidations.

outlook hinges on earnings season

Historical patterns show July often brings hedge fund deleveraging, with positioning recently stretched well above normal levels. Traders are now focused on upcoming earnings reports through July and August to determine whether corporate AI spending and cloud demand can sustain the current pace of investment.

Analysts maintain generally positive long-term views on Meta, with UBS setting an 865 dollar price target, Morgan Stanley 775 dollars, and Bernstein 850 dollars, even as risks remain tied to advertising trends, regulation, and execution on large-scale data center expansion.