#美股史诗级抛售 #Gate Stock Chat Season
The Second Half of the AI Rally: From Compute to Applications, Who Will Rebound First, Microsoft or Amazon?

The AI rally is shifting gears.
Over the past two years, the market rewarded “compute”: whoever bought more GPUs and expanded data centers faster took center stage. The “selling shovels” logic was simple and blunt. But in the second half of 2026, the narrative has changed—mutual funds are starting to abandon pure story stocks and use “annualized returns” to screen investments, with capital moving from upstream compute hardware to downstream applications. The same dynamic is playing out in U.S. stocks: two cloud giants both delivered better-than-expected earnings, yet Microsoft and Amazon met opposite fates, with one rising and the other falling. Behind this is the market’s repricing of “AI commercialization.”

I. Two Earnings Reports, Two Fates

Microsoft: Growth and cash flow both deliver.

In the quarter just ended, Microsoft revenue was $92.3 billion, up 19% year over year; Azure growth accelerated from 42% to 48%, and for the first time since going public, Microsoft separately disclosed Azure revenue in dollars, at $33.17 billion. More importantly, there were signs of traction on the application side: Copilot daily active users grew nearly 50% quarter over quarter, Microsoft Cloud revenue reached $63.9 billion, up 30%, and commercial remaining performance obligations (RPO) surged 84% in the previous quarter to $678 billion—pointing to exceptional visibility into future revenue.

The market’s reward for Microsoft was unmistakable: while Meta and Alphabet faced repeated scrutiny over massive AI investments, Microsoft delivered more than $10 billion in cash flow in a single quarter to “offset” concerns about capital spending. Its stock rose after earnings as the market endorsed the idea that “AI commercialization is moving from story to profit.”

Amazon: Equally stunning growth, but capital spending spooked the market.

AWS revenue grew 39.3% year over year, accelerating for the sixth consecutive quarter; its backlog reached $578 billion; and AI inference revenue accounted for more than half of AWS AI revenue. Fortune 500 companies’ spending on Bedrock grew 7x year over year, while Token consumption rose 17x—demand was nearly flawless.

But the market’s reaction was a sell-off: quarterly capital expenditures reached $75.9 billion, full-year guidance was around $220 billion, and the company said 2027 would see “more.” In a period of heightened sensitivity to “AI spending,” growth surprises were completely overshadowed by fears of capital spending.

II. The Rotation: Why the “Application Narrative” Is Worth More

The shift in capital from compute to applications reflects a change in investment logic: from “who is building infrastructure” to “who is turning AI into revenue.” Under this framework, the market is looking for companies with visible AI revenue and proven business models, rather than spenders expanding for returns far in the future.

2026 is seen by the market as the year AI applications move from technical validation to large-scale commercialization: coding and video generation have already led the way in proving their business models, with AI office tools, AI marketing, and vertical-industry applications close behind. The application side has had its rallies—previously, the AI applications sector rose more than 13% in 10 trading days—but sustaining those gains requires ongoing earnings validation.

III. Who Will Rebound First—Microsoft Has a More Direct Path; Amazon Is Waiting for the Next Wave

If capital definitively shifts from compute to applications, Microsoft is more likely to rebound first, for three reasons:
First, its application narrative is the “purest.” Copilot, M365, and enterprise agents are the world’s largest-scale examples of AI application commercialization. Microsoft is both a “compute provider” and an “application vendor,” with a higher share of business on the application side and faster monetization; Amazon’s strength remains compute infrastructure, while its application layer is relatively thin.
Second, the paths to valuation re-rating differ. The software sector has already seen the script of “an AI narrative reversal followed by a broad software stock rally” play out. As the world’s largest software company, Microsoft is the most direct beneficiary of this re-rating, while Amazon is still classified by the market as a “compute stock burning cash for growth.”
Third, cash flow is the dividing line. Despite both companies exceeding expectations, Microsoft was rewarded for its abundant cash flow, while Amazon was punished for runaway capital spending—tilting the balance for short-term capital.
But Amazon’s opportunity has not disappeared; its timing is simply later. AWS’s 39.3% acceleration and $578 billion in backlog show that demand is not the problem; the capital-spending narrative is all that is holding back its valuation. Once 2027 capital-spending guidance moderates and AWS margins stabilize, the market could shift from “fearing cash burn” to “confirming returns.” Amazon’s rebound could then be stronger than Microsoft’s—with faster growth and lower expectations.

IV. Four Indicators to Watch to Judge the Rotation

1. Microsoft: M365 Copilot paid penetration and successful price increases, the share of Azure revenue from AI services, and agent-product ARR;
2. Amazon: The “scissors gap” between AWS margins and capital-spending growth, 2027 capital-spending guidance, and advertising growth;
3. The software sector: AI ARR disclosures from SaaS companies and changes in capital flows—leading signals of whether the rotation can continue;
4. Macro variables: Interest-rate expectations and liquidity conditions, which determine overall tolerance for high-valuation growth assets.

Conclusion

Can cloud computing and AI services continue to deliver growth? Demand is not the problem, and growth is accelerating; the real disagreement is over the “timing of returns on capital spending.” In the short term, Microsoft is riding the wave of the rotation toward applications and is more likely to rebound first. Amazon, meanwhile, is like an undervalued “option,” waiting for the capital-spending narrative to turn.
For investors, rather than guessing when the rotation will happen, it’s better to watch the indicators above—the data will provide the answer before the market does.$MSFT {currencycard:stock}(MSFT) ‌

$AMZN {currencycard:stock}(AMZN) ‌