For five consecutive quarters, Meta Platforms did something that very few companies at its scale manage to pull off. It beat earnings expectations every single time, while simultaneously ramping one of the most aggressive AI spending programs in corporate history.
The market rewarded both things. The stock climbed. The narrative held together.
July 29 was different. Meta reported after the bell and the streak was gone. EPS came in at $6.18. The Street was looking for $7.20. The stock dropped nearly 9% in after-hours trading and landed at $542.
The next morning, Bank of America’s Justin Post released a note. His view was that the market got it wrong.
Why Meta’s Q2 earnings missed and what it actually means
The revenue number was fine. Meta posted $60.8 billion in total revenue for Q2 2026, up 28% year on year and ahead of the $60.2 billion Street estimate. Advertising revenue was $59.4 billion, slightly above the $58.9 billion consensus. Other Family of Apps revenue came in at $1.007 billion, up 73% year on year and well above estimates, according to 24/7 Wall St.
The miss was entirely in costs. Total expenses jumped 55% year over year to $42 billion. Inside that number were a $2.4 billion legal charge related to youth safety proceedings and a $1.2 billion severance charge tied to the 8,000-person headcount reduction Meta carried out in May 2026.
Those are one-time items. They don’t say anything about the trajectory of the advertising business. But they hit the EPS line hard enough to snap the earnings beat streak and send the stock down.
More Meta:
- Meta layoffs take disturbing turn in new lawsuit
- Meta business model in trouble from $1.4 trillion lawsuit
- Meta doubles down on smart glasses amid always-on feature concerns
Free cash flow dropped 91% year over year, from $8.55 billion in Q2 2025 to just $784 million. That’s the number that worries investors who are already nervous about Meta’s capital expenditure trajectory. Full-year 2026 capex guidance was narrowed to $130 billion to $145 billion from a prior range of $125 billion to $145 billion.
Management also flagged that compute capacity would continue to grow in 2027 and potentially into 2028, though the company later clarified that it has the flexibility to adjust 2028 spending based on demand.
What Bank of America said about Meta’s AI advertising gains and capacity assets
Post’s note argues that the after-hours sell-off reflects concern about Meta’s investment direction, rather than any deterioration in the underlying business.
Instagram time spent was up 10% globally year over year. Facebook video time spent grew 9% globally. The high end of Meta’s Q3 guidance implies 26% revenue growth excluding currency effects, a deceleration of just one percentage point on a three-point tougher comparison, and still well above the broader digital advertising sector.
“AI capacity strengthening core ad business and provides strong optionality; compelling core valuation at 13x 2027 EPS,” Post wrote in the note. “Buy.”
His thesis rests on two things the market is currently underpricing.
- Meta’s AI investments are already producing measurable results in the core advertising business. AI-powered ranking and recommendation improvements drove an 8.3% increase in ad clicks and a 15.7% uplift in Facebook conversions in Q2. Advantage+ reached an annualized revenue run rate of more than $75 billion in the quarter. These are not future projections. They’re happening now.
- Meta’s capacity assets, the data centers, GPUs, and custom silicon it is building, are significantly more valuable than the market is reflecting in the stock price. “We’re getting a lot of offers for compute at a significant premium over what we paid for it,” Zuckerberg said on the earnings call. As Meta’s data centers ramp over the next 12 months, Post believes the company has significant optionality to monetize that capacity externally through enterprise partnerships, API licensing, and infrastructure deals, as TheStreet reported.
6 Meta AI catalysts Bank of America says could move META stock in 2026
Post’s note outlines a specific catalyst path that he thinks can shift investor sentiment on Meta’s AI spending. The next major milestone is Meta’s Connect Conference on Sept. 23. Beyond that, the note identifies five additional potential catalysts over the following months, as TheStreet reported.
The first is a frontier AI model launch. According to press reports cited in the note, Meta’s next-generation model, internally code-named Watermelon, has already achieved frontier-level performance on internal benchmarks. A public launch at or after Connect could strengthen investor confidence in Meta’s AI execution and move the stock.
The second catalyst is a proprietary chip launch. Meta has been developing custom AI silicon for years. A public announcement of a production-ready chip would be a significant signal that Meta can reduce its dependence on Nvidia hardware.
Third is a personal AI assistant launch.
Fourth is a business AI revenue platform, including subscription products.
Fifth is enterprise API licensing deals that give external companies access to Meta’s AI models and compute capacity.
Any one of these could change the narrative around whether Meta’s spending will generate returns, according to CNBC.
Benjamin/Getty Images
Why Bank of America cut the Meta price target to $810from $825
The price target cut to $810 from $825 was driven by a 1% reduction in 2027 GAAP EPS estimates, to $34.73 from $35.00. Post raised revenue estimates slightly for both 2026 and 2027, but increased expense forecasts to reflect the one-time charges, higher stock-based compensation, and lower other income from higher interest costs. The valuation multiple stayed at 24 times 2027 earnings.
At the after-hours price of $542, Post’s note puts the stock at 16 times 2027 earnings on a total company basis, or 13 times when stripping out Reality Labs losses. The S&P 500 trades at around 20 times 2027 earnings. Historically, Meta has traded at a three-point premium to the index. Right now it’s trading at a meaningful discount, which is the core of Post’s valuation argument.
The $810 price target from the July 30 closing price of $585.61 implies roughly 38% upside. Post acknowledges that meaningful AI monetization outside advertising may take time to materialize.
But his argument is that even if the AI revenue story takes longer than expected, Meta retains the flexibility to moderate its infrastructure buildout, which would drive a significant uptick in free cash flow. At the current valuation, he sees more upside than downside, regardless of which scenario plays out.
What Meta’s Q2 earnings and the BofA note mean for META stock investors
The core debate around Meta hasn’t changed after this quarter. Bulls see a company building the infrastructure for consumer AI, enterprise software, and cloud computing while simultaneously running one of the world’s largest advertising businesses.
Bears question whether the capital spending will ever generate returns that justify the cost. The Q2 miss gave the bears a data point. The advertising metrics gave the bulls theirs.
Post’s note is a bet on the bulls. He thinks the market is too focused on the size of the AI investments and not focused enough on what those investments are starting to produce. The Watermelon model, the Connect Conference, and the possibility of external capacity deals are the events he’s watching to determine whether that thesis is right.
META stock is down roughly 7% on the day. Bank of America is staying long.
