Is Meta’s AI spending working? The stock’s next move depends on answer
Shares of Meta Platforms are recovering from a bruising stretch earlier this year, when investors shied away from the social media giant’s massive investment in artificial intelligence. Wednesday’s earnings will determine whether that rebound will continue. The stakes are high: Meta is in the midst of one of the most aggressive AI buildouts of all the megacaps. In recent weeks, the Facebook and Instagram parent laid out investments spanning cloud infrastructure and custom chips and inked massive compute commitments — all part of the company’s plan to spend as much as $169 billion this year, most of which will go to artificial intelligence. Investors are increasingly focused on whether all that spending is paying off yet. So far, the market hasn’t known quite how to digest Meta’s ambitious plans. The stock initially jumped after its quarter on Jan. 28, when management forecast 2026’s operating and capital spending would be materially above expectations and would account for nearly all of the 2026 productivity and revenue growth. For a moment, investors cheered and the stock closed at $738. Then the market turned chilly. Shares fell roughly 29% in two months, to $525 in late March. But shares have risen 28% since then, ending Tuesday at $671. While some of that gain was driven by the broader market’s recovery from the March 30 Iran war bottom, investors also seemed optimistic about a series of announced investments aimed at increasing Meta’s compute capacity. The release of a new AI model helped sentiment too. The stock is up almost 2% on the year. On Friday, Meta announced a multibillion-dollar partnership with Amazon Web Services to deploy AWS Graviton processors at scale, making Meta one of the world’s largest customers of Amazon’s in-house chips. The processors will support workloads tied to boosting its core advertising business. Earlier this month, Meta committed $21 billion to AI cloud infrastructure firm CoreWeave, adding to a prior $14.2 billion agreement. In March, the company signed an agreement worth up to $27 billion with Dutch cloud provider Nebius. The company also expanded its partnership to buy Broadcom’s next-generation AI chips, while planning four customer silicon options of its own. META YTD mountain Meta’s stock performance in 2026. In a well-timed note dated March 29, Morgan Stanley analysts argued that the market had become too focused on the cost of AI and not enough on returns, noting that Meta’s business is still growing strongly even as shares have become cheaper. “We believe engagement [time spent] is accelerating [off of large numbers], which gives META even more time and engagement to monetize, while we believe the time spent growth is high quality and monetizable too, given the surge in video-based content,” analysts wrote then. Jim Cramer struck a similar tone during the April Monthly Meeting, telling members, “I don’t like to bet against Mark [Zuckerberg] when it comes to money.” He added, “You’re buying a call on that incredible talent that Zuckerberg got. We used to think that was a negative. Not anymore cause of this Muse Spark, their new flagship model designed for personal intelligence.” Heading into the quarter, investors want to see more proof that the strategy is translating into stronger growth in its advertising business, better products and higher profits. The vast majority of Meta’s revenue comes from ads. Investors, including us at the Club, will want to see more details on the effectiveness of Meta’s AI-powered ad tools, such as Advantage+, AI-generated ads, and automation, which so far have been game-changers in improving ad performance, with Reels being a major beneficiary. Last quarter, Instagram Reels watch time increased 30% year over year in the U.S., while Facebook video watch time grew in the double digits. Muse Spark , Meta’s first project from its newly created Meta Superintelligence Labs, could be the next chapter in the company’s ads growth story and help shares keep their momentum alive. Investors liked what they’ve seen initially. Meta stock closed 6.5% higher after Meta released Muse Spark on April 8, on the enthusiasm that it could improve its core ad model and justify all its AI spending. For Meta, LLMs are a core part of its ad growth strategy, since AI models predict what content people want to see and which products they care about. An LLM is an AI system trained on massive amounts of data so it can understand language, recognize patterns, reason through prompts, and generate responses. Muse Spark is a multimodal reasoning model that handles text, images, and audio, designed for use on Facebook, Instagram, WhatsApp, Threads, and business tools. It’s supposed to make Meta’s apps more engaging and its ads more effective, which should help accelerate the company’s top-line growth. Meta also has other models that serve its recommendation engine, and the company has been working on merging them. The goal is the same. If Meta serves the ad that will most likely lead to user action — like buying a product — advertisers are willing to spend more. Case in point: Threads, the Instagram-linked text-based app launched in July 2023, saw a 20% increase in time spent last quarter, driven by recommendation optimization. Muse Spark was formed after Meta reorganized its AI efforts under new chief AI officer Alexandr Wang, the former CEO of Scale AI . Wang is one of the AI researchers that Meta brought on as part of its massive talent hiring spree last year . Spark also gives Meta a chance to compete with AI leaders like OpenAI and Google. Analysts at Cantor Fitzgerald argue that Meta is still in the early stages of extracting value from LLMs. “Over the next few quarters, we expect META to leverage Muse Spark to deploy LLM’s reasoning capabilities to improve engagement and monetization of the platform across various apps and services,” analysts wrote in an April 11 research report. Morgan Stanley claimed that Meta’s “visibility on forward growth from its core investments remains high,” with “one of the next biggest unlocks” in 2027 likely to be the use of LLMs to analyze Meta’s leading data. As a consumer-facing company with experience deploying LLMs – though still unproven – Muse Spark could further enhance Meta’s ad performance. The real AI monetization opportunity through Muse Spark would be getting the LLM adopted with enterprise customers. MoffettNathanson said Meta’s push into enterprise is “uncertain, and largely fantastical” at this stage, but sees enterprise as one of the clearest paths to monetizing Meta’s massive AI investments through areas like subscriptions, agents, API access, and cloud services. OpenAI and Anthropic already have meaningful market share in this area, but competition hasn’t historically stopped Meta from pursuing a sizeable opportunity. The company will need to prove how it can translate its Muse Spark frontier model into a credible enterprise business. In the meantime, another catalyst for Meta’s stock is how it is managing its rising costs. Meta is trying to help fund its infrastructure buildout with fewer people. Last Thursday, the company said it plans to cut about 8,000 jobs , or about 10% of its workforce, beginning in May. It will also eliminate 6,000 open roles as it reallocates resources toward AI. Meta has been reducing its payroll since late 2022 as it redirects resources toward AI-related investments. “We’re doing this as part of our continued effort to run the company more efficiently and to allow us to offset the other investments we’re making,” Meta’s chief people officer, Janelle Gale, wrote in a memo announcing the news to staff. While layoffs are never welcome news, Morgan Stanley called the workforce reduction a “bullish development” based on the math. A potential 20% workforce reduction could save anywhere from $3 billion to $10 billion annually, or could add $1 to the company’s 2027 earnings per share. (Jim Cramer’s Charitable Trust is long META. See here for a full list of the stocks.) As a subscriber to the CNBC Investing Club with Jim Cramer, you will receive a trade alert before Jim makes a trade. Jim waits 45 minutes after sending a trade alert before buying or selling a stock in his charitable trust’s portfolio. If Jim has talked about a stock on CNBC TV, he waits 72 hours after issuing the trade alert before executing the trade. THE ABOVE INVESTING CLUB INFORMATION IS SUBJECT TO OUR TERMS AND CONDITIONS AND PRIVACY POLICY , TOGETHER WITH OUR DISCLAIMER . NO FIDUCIARY OBLIGATION OR DUTY EXISTS, OR IS CREATED, BY VIRTUE OF YOUR RECEIPT OF ANY INFORMATION PROVIDED IN CONNECTION WITH THE INVESTING CLUB. NO SPECIFIC OUTCOME OR PROFIT IS GUARANTEED.
<
