Big Tech Beat the Numbers. The Market Sold Them Anyway. Here Is Why.

Wednesday evening was the most consequential 80 seconds in corporate earnings history. Meta, Microsoft, Alphabet, and Amazon reported their first-quarter results within moments of each other after the closing bell. The numbers were good. The market reaction was not. By Thursday morning, Meta had fallen 7%, Microsoft was down 4%, and Amazon had shed 3%. Only Alphabet escaped with a gain, rising as much as 6% in after-hours trading. The pattern across all four companies told the same story: beating the numbers is no longer enough. The bar is now beating the expectations behind the expectations.

The combined quarterly revenue of the four companies exceeded $430 billion. Their combined AI capital expenditure commitments for 2026 alone now exceed $650 billion. In a single year, four American technology companies have pledged to spend more on AI infrastructure than the GDP of most countries. The question the market spent Thursday answering is whether those returns are arriving fast enough to justify the price.

Alphabet: The One That Got Away

Google's parent was the clear winner of the night. Revenue came in at $107 billion, up 19% year-on-year. Google Cloud surpassed $20 billion in quarterly revenue for the first time in its history, growing 63% over the prior year. Net income surged 81% to $62.57 billion. The company raised its full-year capex guidance to between $180 billion and $190 billion. The AI narrative held up: search proved resilient against fears of disruption, cloud growth accelerated sharply, and the capex commitment was interpreted as a company investing from a position of strength rather than desperation. Shares climbed 6% after hours. When the returns are visible and accelerating, investors are willing to fund the infrastructure. When they are not, the same spending becomes a liability.

Microsoft: Azure Beat, Costs Shocked

Microsoft delivered a clean beat on the numbers that mattered most. Azure grew 40% in the quarter, ahead of the 38% to 39% guidance range that had made investors nervous heading into the report. Revenue of $82.89 billion and earnings per share of $4.27 both cleared consensus. The AI monetization story is working: Copilot adoption is accelerating, Azure AI services are growing faster than the core cloud business, and the company's enterprise relationships give it a distribution advantage that is difficult to replicate.

What unsettled investors was the capex update. Microsoft guided to $190 billion in capital expenditure for 2026, up 61% from last year and approximately $35 billion above what analysts had modeled. CFO Amy Hood attributed $25 billion of that increase to higher component prices, a reminder that the AI infrastructure buildout is getting more expensive, not less. The stock fell roughly 4% on Thursday. The business is performing. The cost of building it is rising faster than the market had priced.

Meta: The Best Quarter in Years, the Worst After-Hours Reaction

Meta's underlying business is in exceptional health. Revenue grew 33% to $56.31 billion, its fastest quarterly growth since 2021. Earnings per share of $7.31 beat the $6.79 estimate. Ad impressions grew 19% and price per ad rose 12%. Daily active users across the family of apps reached 3.43 billion.

And yet shares fell more than 7% after hours. The culprit was a single line in the earnings release: Meta raised its full-year capex guidance from a range of $115 billion to $135 billion up to a new range of $125 billion to $145 billion, a $10 billion increase at both ends. The company cited higher component pricing and additional data center costs. Investors who were already uncomfortable with the original range got a larger number than they expected. The business beat. The spending scared.

Amazon: Cloud Acceleration, Capex Reaffirmed

Amazon delivered what may have been the most straightforwardly impressive set of numbers of the four. Revenue of $181.5 billion grew 17% year-on-year. Earnings per share of $2.78 demolished the $1.64 consensus estimate. AWS revenue grew 28% to $37.6 billion, its fastest pace in more than three years, and the company posted a record worldwide operating margin of 13.1%. CEO Andy Jassy reaffirmed the company's $200 billion capex commitment for 2026, supported by AI partnerships with OpenAI, Anthropic, and Meta. The stock fell roughly 3% anyway. AWS is growing at 28% annually on a $150 billion revenue run rate. The question investors are asking is whether $200 billion in annual infrastructure spending is the right price to sustain it.

The $650 Billion Question

Taken together, the four companies have committed to spending more than $650 billion on AI infrastructure in 2026 alone. The returns on that spending are real but unevenly distributed. Alphabet's cloud is growing at 63%. AWS is growing at 28%. Azure is growing at 40%. These are not speculative bets. They are businesses generating tens of billions of dollars in revenue and accelerating. The infrastructure spending is not going into a void.

But the market is asking a harder question: at what point does the spending outpace the returns? The component price inflation that Microsoft and Meta flagged is a structural issue, not a one-quarter anomaly. Nvidia's chips are expensive and in short supply. Data center construction costs are rising. The companies building AI infrastructure are competing with each other for the same constrained inputs, which means the cost of the buildout is likely to stay elevated even as the revenue it generates grows.

The Fed and the Macro Backdrop

The earnings deluge landed on the same day the Federal Reserve held rates at 3.50% to 3.75% in an 8-4 vote, the most divided FOMC decision since October 1992. One governor voted to cut. Three others objected to language implying future cuts remain on the table. Core CPI is running at 3.3% year-on-year and core PCE at 2.7% to 3.0%, both well above the 2% target. The Strait of Hormuz remains effectively closed, with Brent crude up more than 83% year-to-date. The macro environment in which these companies are spending $650 billion is one of elevated inflation, a divided central bank, and an unresolved geopolitical shock.

In a falling rate environment, long-duration spending commitments are easier to justify. In a higher-for-longer environment, the discount rate applied to future returns rises, and the present value of infrastructure spending that will generate returns over five to ten years shrinks. The market's reaction to Wednesday's earnings was not irrational. It was a rational repricing of long-duration AI bets in a world where the cost of capital is higher than it was two years ago. The AI supercycle is real. The returns are arriving. The spending is also real, and it is getting more expensive. Wednesday night was the moment those two facts collided in the same earnings call, and the market spent Thursday deciding which one to weight more heavily.