Investors Love AI, as Long as You’re a Cloud Host

Investors strongly support the huge capital spending of cloud leaders like Amazon, Microsoft, and Google. Soaring cloud income directly validates their investments in AI infrastructure. This stands in clear contrast to the market’s doubt about other tech firms making large bets without a visible income link. This situation shows how cloud providers are now viewed as the most secure and lucrative part of the AI economy.

Key takeaways

• Cloud leaders gain investor approval for massive AI spending because of strong, directly-connected income growth.
• Amazon’s AWS income jumped 37% year-over-year, validating its increased $220B 2026 spending forecast.
• This investor trust applies to Microsoft and Google but not to companies like Meta, where large spending lacks a clear income source.
• Cloud income is another company’s AI cost, making it fragile if end-user need for AI apps doesn’t appear.
• Amazon’s plan involves long-term commitments to its own AI chips, built to boost cloud profits beyond just data center construction.

Cloud Leaders’ Spending Spree Validated by Rising Income

Amazon reported a 20% increase in net sales. Its AWS cloud income rose 37% year-over-year to $42 billion for the quarter, pushing its stock up almost 10%. Despite common belief, investors aren’t penalizing the company for enormous capital expenditure. This hit $173 billion for the fiscal year and has a higher 2026 forecast of $220 billion.

Income Validates the Spending

Normally, such swelling costs would worry investors. Yet Amazon has an income engine that helps validate the outlay. The jump in AWS income shows demand grows alongside the increased supply of data center capacity. This is reassuring given the multi-year delay between building a data center and selling its capacity.

A Trend Among Cloud Providers

This situation isn’t unique to Amazon. We saw similar trends at Microsoft and Google, whose shares also climbed after reporting solid cloud income. This differs greatly from investor doubt about firms like Meta. They have major spending but no clear income source, leading to stock drops.

The Strategic Infrastructure and Chip Plans Behind the Cloud

Amazon’s capital spending for the fiscal year ended June 30 reached $173 billion on property and equipment. This category covers GPUs, natural gas turbines, and land for data centers, up from $107.65 billion the prior year. The company raised its 2026 spending forecast to $220 billion. It’s using cash reserves to fund this expansion, ending its last quarter with $7.6 billion less cash and marking a time of negative free cash flow.

More Than Physical Data Centers

Importantly, Amazon’s AI plan goes beyond building data centers. The company is making serious long-term commitments to its own chips like the Trainium TPU and the Arm-based Graviton processor. While these projects don’t show up in spending numbers, they are built to substantially improve profits for the cloud business.

CEO Andy Jassy stated during the Q2 earnings call, “We see the AI business following very much the same margin trajectory we saw in the core business before.” He argued that AWS and Amazon Bedrock can have a successful business without its own top-tier model, noting, “there’s not going to be a single model to rule them all.” This spending gets support from AWS income, which rose 37% year over year to $42 billion for the quarter.

The AI Stack’s Fragile Economics and the Cloud’s Protection

Investors currently treat cloud providers as the most dependable part of the AI stack. They remain doubtful about the core economics for AI labs and startups. This is clear in the market’s reaction to earnings. Companies like Amazon, Microsoft, and Google saw their shares rise on strong cloud income. Meanwhile, Meta’s stock fell as investors focused on its large capital spending without a clear income source.

The Direct Connection to AI Spending

However, cloud hosting income is directly linked to the spending of AI companies. Amazon’s AWS income, which rose 37% year over year to $42 billion for the quarter, is another firm’s AI cost. For Anthropic, it is literally the same money. If that spending from major AI labs and their clients proves unsustainable, the income stream for cloud providers like Amazon won’t be stable.

The whole AI expansion depends on what the source calls “David Cahn’s $3 trillion question”: whether enough end-user demand exists to validate the huge investment. Cloud-hosting services sit a few steps away from that final demand problem, but they aren’t completely protected from a potential crash. If broad need for AI applications doesn’t appear, it will hurt every level of the stack.