The stock chart tells the story of the year so far: Meta Platforms shares have fallen almost 10 percent since January, a slide that has left the company trailing most of its big technology peers. The next chapter arrives on July 29, when Meta reports second-quarter results, and investors will be looking for one thing — evidence that the company’s massive bet on artificial intelligence is starting to pay for itself.
The decline has unfolded despite a core business that remains healthy. Advertising revenue has grown steadily, the family of apps keeps adding users, and Meta’s earnings have beaten expectations in recent quarters. The problem is not the ad machine. It is the bill for the AI build-out, which has grown faster than the revenue it is meant to produce.
Moody’s raised the issue in stark terms this year, warning that artificial intelligence capital spending is eroding the credit quality of Meta, Amazon and other technology giants. The rating agency pointed to a simple dynamic: cash flows that used to accumulate now get reinvested at a furious pace into data centers, chips and energy, and the returns on that spending remain years away.
The numbers are easy to trace. Meta’s capital expenditures have climbed to levels that would have been unthinkable three years ago, and the company has said it plans to keep spending at elevated levels through this year and next as it builds out computing capacity for AI training and inference. Free cash flow, the metric investors watch most closely, has narrowed as a result.
Meta has a specific story to tell about where the money goes. Its Llama family of models powers AI features across WhatsApp, Instagram and Facebook, and the company has been integrating AI assistants into its products at a pace that rivals anything in the industry. The question Wall Street keeps asking is whether those features translate into higher prices for advertisers or new products that generate revenue.
The company’s position in the AI race is unusual. Unlike Microsoft and Google, Meta does not sell cloud computing, so it cannot monetize its infrastructure by renting it out to other companies. Its AI spending must be justified through its own products — advertising tools, recommendation engines, consumer assistants. That makes the return-on-investment math harder to demonstrate.
Investors’ patience has limits, and the market has shown signs of losing it. The stock’s decline this year reflects a rotation out of companies whose AI spending outstrips visible returns, and toward the chip suppliers and infrastructure providers that are profiting from the build-out regardless of its eventual outcome.
The July 29 report will be the first test of the year’s narrative. Analysts expect revenue to keep growing at a double-digit pace, but the focus will be on the guidance: how much Meta plans to spend in the second half, whether it will raise its full-year capital expenditure forecast again, and how it describes the path from AI investment to AI income.
Meta has been here before. The company’s stock cratered in 2022 when a previous spending cycle — the Reality Labs metaverse bet — collided with a weak advertising market. Zuckerberg’s response was cost discipline and efficiency, which restored margins and rewarded shareholders. The AI cycle is bigger, and the company has signaled it will not cut its way out of it.
Executives have argued that the AI opportunity is too large to under-invest in. Meta’s advertising business runs on recommendation systems that AI makes better, and the company has shown that better targeting and creative tools command higher ad prices. The bull case rests on that connection: AI spending that improves the core ad business is not a gamble, it is an upgrade.
The bear case is equally simple. If AI improvements do not show up in advertising growth or new revenue streams, Meta will be spending tens of billions of dollars a year on infrastructure with no incremental return, and its credit profile and stock price will both feel the pressure.
Moody’s warning gave the debate a formal voice. Credit rating agencies rarely move fast, and a warning about credit quality is the kind of signal that investors in both bonds and equity notice. For Meta, the practical effect is a higher cost of borrowing at a moment when it has turned to the bond market to fund its build-out.
The second quarter results will also be judged against the company’s own history. Meta has beaten earnings estimates for years, and the market has come to expect it. A miss on the AI investment thesis — guidance that disappoints, or evidence that spending is running ahead of plan — could accelerate the stock’s decline. A beat, with a credible path to returns, could start the recovery.
The stock has drifted near its levels of a year ago, giving back the gains of late 2025. The next few weeks will determine whether the decline is a pause or a repricing. The company’s answer to that question will be delivered in numbers on July 29 — capex guidance, ad revenue growth, and whatever Zuckerberg chooses to say about the future of the AI build-out.
For now, the market’s message to Meta is the same message it has sent to every big technology spender this year: show us the returns, or slow the spending. July 29 is the next chance to do either.
