The money's already spent — now Wall Street is betting it actually worked.
The Summary
- JPMorgan raised its S&P 500 target twice in two months, citing AI capex finally translating to corporate earnings
- The "AI capex logjam" refers to the gap between massive infrastructure spending and actual revenue generation — a gap that appears to be closing
- Tech giants bought the picks and shovels; now they're mining gold, and investors are adjusting expectations upward
The Signal
For two years, hyperscalers poured billions into GPUs, data centers, and network infrastructure while analysts nervously asked when the revenue would show up. JPMorgan's revised S&P 500 forecast signals that question is getting answered. The strategists didn't just nudge the number — they raised it twice in eight weeks, a confidence move that says the deployment gap is narrowing.
The logjam worked like this: companies bought infrastructure in 2023-2024, built models and products in 2025, and are now booking revenue in 2026. The lag was predictable but nerve-wracking for anyone holding equity. You don't typically see ROI the same quarter you rack servers.
"The money's already spent — now it has to work."
What changed? Corporate earnings are reflecting AI products that actually shipped and customers who actually paid. Not demos. Not pilot programs. Revenue. The capex deployment gap meant there was always going to be a period where balance sheets looked scary before income statements looked good. We're exiting that period.
This matters because it validates the Web4 thesis: agents aren't vaporware, and the infrastructure spend wasn't a bubble. The companies that built agent platforms, fine-tuning APIs, and inference optimization tools are now the ones driving earnings beats.
Key points on the timing:
- Infrastructure purchases: 2023-2024 (the "capex" phase)
- Product builds and internal deployment: 2025 (the "logjam")
- Revenue recognition and market repricing: 2026 (the payoff)
The market is repricing risk. If you thought AI spend was speculative six months ago, JPMorgan's double-raise says it's now fundamental. That shift — from "maybe this works" to "this is working" — is what moves trillions in asset allocation. Pension funds and index trackers don't chase hype. They chase proven earnings growth.
The Implication
If you're building in the agent economy, this is your permission slip to think bigger. The infrastructure is paid for, the models are improving weekly, and the market just confirmed it's willing to reward companies that turn AI into actual products. The picks-and-shovels phase is over. Now it's about what you build with them.
Watch which sectors JPMorgan highlights in their next earnings call breakdown. The S&P 500 is an index, but the raises are driven by specific companies proving AI revenue at scale. Those are your case studies for what works and what's still vaporware.