Meta Platforms' aggressive buildout of artificial intelligence infrastructure may open a lucrative new revenue channel if the company decides to lease out spare computing capacity, according to Evercore ISI. The social media giant is on track to amass 14 gigawatts of total compute capacity by 2027, a scale that has weighed on its stock amid concerns over capital intensity.
Shares have slipped roughly 4% over the past month and are down 12% year-to-date, even as the S&P 500 has climbed 12%. The market's skepticism stems from the sheer size of Meta's data-center investments, but Evercore analyst Mark Mahaney argues that the spending could eventually pay off in an unexpected way.
Compute shortage creates opportunity
Meta is the only major hyperscaler without a public cloud business, but recent reports suggest it may establish a Meta Compute division to monetize excess capacity. In a market where nearly every scaled operator is sold out, Mahaney sees Meta as a rare potential merchant seller of surplus compute.
He does not expect Meta to become a full-fledged neocloud provider like Nebius or CoreWeave. Instead, he envisions a limited effort to sell a portion of capacity that would otherwise sit idle. According to his estimates, placing just 0.5 to 1 gigawatt on the market could generate between $11 billion and $22 billion in annual external revenue.
Leasing 1 gigawatt would represent only about 7% of Meta's targeted 2027 capacity, yet could add as much as $4.32 to earnings per share, Mahaney calculates. That optionality is a key reason he raised his price target to $860 from $820, implying more than 50% upside from Tuesday's close near $570.
Zuckerberg remains focused on AI returns
Meta has signaled it will not sacrifice long-term AI ambitions for short-term infrastructure revenue. CEO Mark Zuckerberg noted on the latest earnings call that Meta has received offers to buy computing capacity at a premium, but cautioned that selling all compute would be foolish given the higher margins from selling intelligence directly.
That stance suggests Meta's priority is using its infrastructure to power its own AI products, where it expects to capture greater value. Mahaney describes the potential compute business as a "call option" for investors rather than a guaranteed revenue stream. He advises using the optionality as a reason to own the stock, but not underwriting a full-blown hyperscaler annuity.
Meta is also developing its own Iris chips as part of its broader infrastructure strategy, which could further enhance efficiency and margins.
Investors weigh AI spending
Meta trades at roughly 17 times forward earnings estimates, within 10% of its three-year trough valuation multiple, according to Mahaney. He believes the company has proven it can deploy AI effectively to improve user experience, as seen in recent product enhancements.
For investors, the possibility of Meta Compute adds another potential payoff from infrastructure spending, even as Meta emphasizes that selling AI-powered intelligence remains more valuable than selling raw computing capacity. The stock's current valuation may already reflect some of the risk, but the upside scenario hinges on whether Meta can turn its massive compute buildout into a tangible new profit center.
This article is for informational purposes only and does not constitute financial advice.
