Investors love AI, as long as you’re a cloud host
Amazon’s latest second-quarter earnings report has sent a clear signal to Wall Street: when it comes to the artificial intelligence gold rush, investors are playing favorites. Despite a massive, ongoing surge in infrastructure spending that would typically spook shareholders, Amazon’s stock climbed nearly 10% in after-hours trading following a strong performance in its cloud division.
The Capex Conundrum
Conventional market wisdom suggests that ballooning capital expenditures (capex) should be a red flag. Yet, Amazon is doubling down. The company reported that it funneled $173 billion into property and equipment—encompassing everything from high-end GPUs and natural gas turbines to physical real estate—over the fiscal year ending June 30. This is a significant jump from the $107.65 billion spent in the previous year.
Even more striking is the company’s forward-looking guidance. Amazon has raised its 2026 capex forecast to $220 billion, up from a previous estimate of $200 billion. This aggressive strategy has even begun to impact the company’s liquidity; Amazon concluded the quarter with $7.6 billion less cash than it held a year ago, marking its first period of negative free cash flow this year.
Why Investors Are Buying In
If the spending is so high, why are investors cheering? The answer lies in the immediate, tangible returns from AWS (Amazon Web Services).
- Cloud Growth: AWS revenue surged 37% year-over-year, hitting $42 billion for the quarter.
- Demand vs. Supply: While the raw arithmetic of capex versus revenue doesn't perfectly align, the growth in cloud demand provides a necessary buffer, reassuring investors that the massive infrastructure buildout is being put to productive use.
- Strategic Hardware: Beyond standard data centers, Amazon is investing heavily in proprietary silicon, specifically its Trainium and Graviton processors. These chips are designed to optimize margins, allowing Amazon to capture more value within its ecosystem without relying solely on third-party hardware.
“We see the AI business following very much the same margin trajectory we saw in the core business before,” noted CEO Andy Jassy during the earnings call. “AWS and Amazon Bedrock can have a wildly successful business without its own frontier model, and the reason is that there’s not going to be a single model to rule them all.”
The Great Divide in the AI Economy
This favorable market reception is not universal. We are witnessing a distinct bifurcation in how the market values AI-focused companies. Cloud giants like Microsoft, Google, and Amazon are being rewarded for their infrastructure dominance. Conversely, companies like Meta—which are pouring billions into AI without a clear, immediate revenue stream—are facing intense skepticism. Meta’s stock recently dipped 8% as investors scrutinized its cash flow and spending habits.
Key Takeaways for the AI Market
- Cloud as the "Safe" Bet: Investors currently view cloud providers as the most reliable layer of the AI stack.
- The Sustainability Question: Amazon’s hosting revenue is essentially the "AI bill" for other companies, such as Anthropic. If the underlying economics for AI labs and startups prove unsustainable, the revenue stream for cloud hosts could eventually face pressure.
- The $3 Trillion Question: Ultimately, the industry is grappling with whether there is enough genuine demand to justify the massive capital buildout currently underway.
While cloud-hosting services like AWS are currently shielded by their position as the "picks and shovels" providers of the AI era, they are not entirely insulated from the broader market. If the demand for AI applications fails to materialize at scale, the entire ecosystem—from the labs building the models to the hosts providing the compute—will face a difficult reckoning. For now, however, Amazon has successfully convinced the market that its aggressive spending is a calculated path to long-term dominance.