Wall Street analysts lowered their outlooks for Meta Platforms as investors send the stock lower on fears its capital expenditures on artificial intelligence won't yield as strong of a return on investment. Meta's stock dropped 9% in premarket trading Thursday following its second quarter earnings report , which revealed an earnings miss for the hyperscaler. The company reported earnings of $6.18 per share compared to analysts polled by LSEG's estimate for $7.22. While Meta beat on revenue, its guidance on the metric for the current quarter also came in below estimates. The number of active users across Meta's suite of apps also missed, coming in at 3.6 billion. The company's forecast for its capital expenditures for the year narrowed to between $130 billion to $145 billion. Meta previously expected $125 billion on the low end of the range. That spending has led the Mark Zuckerberg-led company's free cash flow to fall to $784 million, down from $8.55 billion a year ago. "Meta shares are down," wrote TD Cowen analyst John Blackledge in a Thursday note, "likely reflecting the slight 3Q rev miss (at midpoint) alongside questions about timing / scaling of AI monetization amid massive capex cycle." JPMorgan analyst Doug Anmuth said monetizing the company's spending on artificial intelligence beyond advertising appears limited. While investors were gaining confidence in the last few weeks about the company's AI strategy — through its first proprietary foundation AI model Muse Spark — Anmuth wrote in a Thursday note that there wasn't much clarity provided. "We did not come away learning much incremental on the developer API, consumer/business agents, or potential plans to directly monetize compute," Anmuth said. Firms across the street broadly lowered price targets, but none downgraded the stock on worries. That's because, according to Deutsche Bank Analyst Benjamin Black, the fundamentals of the company are still strong. "Meta's 2Q26 results were optically messy, but the underlying business and outlook were considerably stronger than the initial stock reaction suggests," Black wrote in a Thursday note. "The larger takeaway is that Meta's core business has the momentum and earnings power to support its expanding AI ambitions." JPMorgan: Neutral, $640 The bank's price target, down from $725, indicates a 9% gain from Wednesday's close. "We expect Meta to continue to invest heavily N-T & we now project 2027 capex of $243B (+70% Y/Y), while our top-line estimates remain largely unchanged, thereby putting greater pressure on FCF. Overall, we remain encouraged by Meta's continued headroom in core ranking & recommendation improvements, model progression, & potential new forms of AI monetization, but we still need greater clarity on the product pipeline & spending outlook." Wells Fargo: Overweight, $640 The bank's price target is down from $835. "Amid broad uncertainty on the hyperscaler investment cycle, Meta incremental ROI case remains the most opaque. Mgmt commentary suggests seeking flexibility in build beyond '27 & less appetite for NT capacity sales. Cutting ests & target, remain OW." Cantor Fitzgerald: Overweight, $680 The firm's price target, down from $770, represents a 16% gain from Wednesday's close. "Overall, there was a lot to like in META's 2Q print. That said, debates on the company's ability to achieve compelling returns in the M-LT from ongoing capex investments are likely to remain open until we see revenue growth potentially re-accelerate." DA Davidson: Buy, $700 The firm's price target, down from $850, indicates a 19% gain from Wednesday's close. "META plans to maximize compute capacity in 2026 and 2027 with near term capacity viewed as more valuable than long-term capacity. Management underscored the past value of its incremental capacity and maintains confidence in its ability to utilize capacity to serve AI products longer term. Looking out to 2028 and beyond META believes it capacity strategy will enable continued growth in compute while noting the extended duration of data center assets provides inherent flexibility to adjust the pace of investment that aligns with the pace of AI adoption." UBS: Buy, $715 The bank's price target, down from $766, implies a 22% gain from Wednesday's close. "We have left our 2026-2028E revenue largely unchanged post print so this would be upside to our estimates as well. Given Meta formally stated that the contingency plan remains to sell excess compute / model access to external vendors - it would be hard to see a scenario where revenue does not benefit from investments at all." TD Cowen: Buy, $750 The bank's price target, down from $800, represents a 28% gain from Wednesday's close. "Mgmt. raised the lower end of its capex guide by $5BN to $130BN-$145BN, citing focus on maximizing '26 and '27 compute capacity. The outlook implies 97% y/y spend growth at midpoint, vs our & cons. pre-print ests of $135BN (+92% y/y). Mgmt. called out ability to monetize their infrastructure via i) business agent & other agentic tools, ii) API access to Meta's models, and iii) direct monetization of infrastructure, for which mgmt. cited 3p offers to monetize their compute at a meaningful premium to their building costs, noting that 'there is nowhere near enough compute for all the demand.'" Deutsche Bank: Buy, $750 The bank's price target is down from $800. "AI continues to drive measurable gains across engagement, advertising performance and creative, while the company is opening new revenue streams across Business Agents, paid messaging, subscriptions, APIs and enterprise products. Meta still has considerable runway to improve ranking and recommendations, supporting both ad inventory and pricing." Morgan Stanley: Overweight, $775 "META's core leading social platform continues to become stronger, with 3.6bn daily active users across Family of Apps and 2bn DAUs on Instagram. Engagement trends and META's ability to serve more relevant organic and paid content are only improving." Barclays: Overweight, $780 The bank's price target, down from $830, indicates a 33% gain from Wednesday's close. "META's ads business continues to defy gravity, growing 26% ex-fx in 2Q (10 points above Google) and gaining more dollar share on the field. The 3Q guidance suggests much of the ad momentum continues into 2H despite the tough comps… the call sounded very broad while the reality is likely more narrowly focused to what exists in the four mega apps at META, but the strategy and business model expansion seemed almost peanut-butter-manifesto-esque. There are already well-resourced and established competitors in many of these new enterprise AI areas." Citi: Buy, $800 The bank's price target, down from $800, implies a 36% gain from Wednesday's close. "We believe Meta's LLMs are beginning to deliver durable engagement gains on continued R & R improvements and we note global time spent on IG rose DD in 2Q. And with 1M+ businesses using Business Agent, 9M+ SMBs using Meta's AI creative tools, as newer apps launch and Meta's subscription offering ramps, we believe there is increasing clarity on Meta's product roadmap, and we await details of Meta's Enterprise offering." Bank of America: Buy, $810 The bank's price target, down from $835, represents a 38% gain from Wednesday's close. "Recent capacity deals suggest AI capacity is very valuable, Meta should have significant capacity optionality as its data centers ramp over the next 12 months, and we think current valuation reflects little to negative value for Meta's growing capacity assets. At AH price $542, Meta is valued at 16x our '27E EPS (and just 13x for Family of Apps), and we see favorable risk/reward with potential for upcoming product launches and capacity deals to raise 2027 rev,/EPS expectations." Evercore ISI: Outperform, $820 The bank's price target, down from $930, indicates a 40% gain from Wednesday's close. "Despite very solid core P & L results, META traded off due to: 1) a lack of forward clarity on '27 CapEx, where mgmt. declined to provide color beyond the CFO's comments about gearing plans to "maximize" 2026-27 capacity – sounds open-ended; 2) no material update on frontier-model progress, with mgmt. offering only qualitative color that "larger and more advanced models" are still being scaled; 3) commentary about prioritizing selling intelligence over compute (despite fielding offers at a "significant premium" to cost), deferring a nearer-term, more visible revenue / FCF offset to the CapEx build in favor of longer-dated, less-certain monetization; and 4) lack of other, non-advertising monetization proof points." Canaccord Genuity: Buy, $930 "Investors remain focused on the spending trajectory even as the core business continues to see strong momentum. With AI execution intact across the platform and new revenue streams beginning to take shape, we see the market's focus on spend and a softer near-term guide as an entry point rather than a change in our thesis."