Goldman Sachs: AI Sparks a Capex Supercycle, Earnings Growth to Drive Stock Prices

N.R. Finch
Published 2026-06-17About 7 min read

Goldman Sachs says AI is fueling the first capital-spending super-cycle in decades; higher borrowing costs will cap valuations, making earnings growth — not multiple expansion — the decisive driver of equity performance.

01

What is a "capex super-cycle," and why now?

Goldman notes that private-sector spending on AI compute, infrastructure, and related technology is accelerating capital demand for the first time in decades.
AI is not the only driver — companies are also reinforcing supply-chain resilience and responding to geopolitical pressure, while governments ramp up defense and critical-infrastructure spending.
In plain terms = multiple streams of capital are pouring into "building things" at the same time, creating a cross-sector investment wave.
02

Borrowing costs are rising — what does that mean for stocks?

Goldman argues that higher borrowing costs will compress equity valuations, closing the old path of riding multiple expansion.
This means → whether stocks rise or fall now hinges on whether companies can actually earn more money — earnings growth becomes the critical variable.
Goldman also expects widening profit divergence across firms, making stock selection more rewarding than broad-index exposure.
03

Why are chipmakers the biggest winners?

Goldman sees chipmakers as among the largest beneficiaries of the AI spending boom; strong demand plus constrained supply have pushed semiconductor net margins close to 50%.
The "Magnificent Seven" — Nvidia, Microsoft, Google, Amazon, Meta, Apple, and Broadcom — now generate a combined 44% return on equity, up 9 percentage points from three years ago.
This reflects powerful pricing ability and competitive moats — record-level profitability has become the core pillar supporting elevated U.S. equity valuations.
04

Cloud giants are spending massively on AI — can they earn it back?

Goldman projects that by 2026, major cloud operators will spend roughly $770 billion in capex — about 100% of their operating cash flow, effectively reinvesting every dollar earned.
Positive signals are emerging, however: improving revenue expectations, growing order backlogs, and expanding margins suggest AI investment is starting to generate returns.
This means → whether cloud operators can sustain profit expansion through peak capex will be the key test of the entire super-cycle's durability.

Content is for reference only, not financial advice.