Goldman Ties Half of S&P 500 Earnings Growth to AI Spend

Goldman Sachs says artificial-intelligence spending is now responsible for about half of S&P 500 earnings growth, a concentration that leaves index profits exposed to the capital-budget decisions of a small group of buyers.

Goldman Sachs has attached a number to something equity investors have felt for two years without being able to size: roughly half of the earnings growth in the S&P 500 is coming from artificial-intelligence spending. It is a statistic that can be read as a boast or as a warning, and the honest answer is that it is both.

The bullish reading is straightforward. AI is not a story about future revenue that never arrives; it is producing measurable profit growth today, right now, in the reported numbers of the largest index in the world. The bearish reading uses exactly the same arithmetic. If AI is contributing half of the growth, then the other half — the ordinary economy of banks, retailers, insurers, industrials and healthcare — is carrying a much lighter load than the headline index results suggest.

One Company's Capex Is Another's Revenue

The mechanism matters more than the number. AI spending shows up in S&P 500 earnings mostly as a transfer: capital budgets at a handful of very large technology buyers become revenue and margin at semiconductor makers, networking suppliers, power and cooling vendors, memory producers, and the contract manufacturers who assemble the racks. The same dollar is an expense on one income statement and a sale on another, and the index nets out to growth because the sellers convert it at high margins while the buyers depreciate it over years.

That is a fragile structure in one specific way. Revenue diversified across thousands of customers is durable. Revenue concentrated in the discretionary capital budgets of a short list of buyers is a decision, and decisions get revisited every quarter. As 24/7 Wall St framed it, the dependency runs through a handful of buyers whose spending plans could shift at any point.

Nothing about that requires a recession or a burst bubble. A capital budget can be trimmed at the margin — pushed a quarter to the right, rephased across a longer build schedule, reallocated from new capacity to utilisation of existing capacity — and the effect on the supplier chain is immediate even though the buyer's own earnings barely move.

What Gets Hit First If Budgets Shift

Investors trying to work out their own exposure should think in terms of distance from the capital budget rather than in terms of sector labels.

• Closest in: the component and equipment suppliers whose order books are directly tied to data-centre build-outs. These names re-rate fastest in both directions because a single buyer's order pattern can swing a quarter.
• One step out: power generation, grid equipment, construction and industrial cooling, where the contracted work has a longer tail but the new-order flow is just as sensitive.
• Furthest out: the software and services companies monetising AI through subscriptions. Their revenue is customer-diversified, but their multiples were set in the same enthusiasm and would not be spared a sentiment reset.
• The buyers themselves: paradoxically, the companies doing the spending have the most control. Slowing capex lifts free cash flow even as it removes growth from everyone downstream.

The point is not that AI-linked profit is fake. It is that the index-level earnings growth rate has become less diversified than the index's 500 constituents imply — a concentration risk hidden inside a broad-market number.

The Tape Is Not Pricing Anxiety

Whatever the structural question, markets ended the most recent session in a constructive mood. The SPDR S&P 500 ETF Trust (SPY) closed at $764.29, up 0.85% from the prior close of $757.83, with a day range of $763.60 to $766.38, as of the last trade on Friday, 11 September 2026. The Invesco QQQ Trust, the Nasdaq 100 proxy most directly geared to the AI trade, closed at $714.88, a gain of 0.87% against a previous close of $708.69. The Dow 30 tracker DIA finished at $525.79, up 0.97%.

Goldman Sachs shares participated, closing at 1,029.18, up 0.92% on the day from a prior close of 1,019.77, having traded between 1,018.59 and 1,042.99. The near-identical gains across all three major benchmarks suggest a broad-tape session rather than a narrow AI-led one — a small piece of evidence that the market can advance without the AI complex doing all the lifting, which is precisely the question the Goldman statistic raises.

What Would Confirm or Break the Thesis

There are a limited number of things worth watching, and they are all observable without a forecast.

Capex guidance language. The large buyers publish capital-spending intentions alongside results. The tell is not the headline figure but whether the phrasing shifts from expansion toward optimisation and efficiency. That language change historically precedes the spending change by a quarter or more.

Order backlog versus shipment. When suppliers report growing backlogs but flattening shipments, the demand is real and the constraint is supply. When backlogs flatten while shipments hold up, the pipeline is draining.

The breadth of the other half. If the non-AI half of S&P 500 earnings growth accelerates — banks, consumer, industrials, healthcare all contributing — the dependency resolves itself without anyone needing to cut a data-centre budget. That is the benign path, and it is entirely plausible.

Depreciation. Today's capex becomes tomorrow's depreciation expense at the buyers. As the installed base ages, that charge grows whether or not the revenue arrives to match it. It is the slow-acting counterweight to the current earnings boost and it does not require any change in behaviour to show up.

The useful takeaway from Goldman's figure is not directional. It is that a single input now explains half the growth in the broadest measure of American corporate profits, and a portfolio built on index exposure is therefore less diversified than it looks. Knowing which half of the engine you own is the work.

Key facts

• Goldman Sachs estimate: AI spending accounts for roughly half of S&P 500 earnings growth
• GS last close: 1,029.18, +0.92% (as of 11 Sep 2026, 20:00 GMT)
• SPY last close: $764.29, +0.85% from $757.83 prior close
• Nasdaq 100 proxy (QQQ): $714.88, +0.87% on the day

Frequently asked questions

What exactly did Goldman Sachs say?

Goldman Sachs said that artificial-intelligence spending is driving approximately half of S&P 500 earnings growth. The firm's framing is that AI capital spending is propping up index-level profit growth, which also means that growth rate depends on the budget decisions of a relatively small group of large corporate buyers.

Why is that statistic considered a risk rather than good news?

Because concentration cuts both ways. Earnings growth sourced from a few buyers' discretionary capital budgets can be revised at any quarterly planning cycle. Revenue spread across thousands of customers is durable; revenue dependent on a short list of spending decisions is not. The same figure that shows strength shows dependency.

How does AI spending turn into S&P 500 earnings?

It is largely a transfer between index members. Capital budgets at large technology buyers become revenue at chipmakers, networking and memory suppliers, power and cooling vendors and contract manufacturers. The sellers book high-margin sales immediately while buyers depreciate the assets over years, so the index nets out to growth.

Which companies are most exposed if AI budgets slow?

Exposure runs by distance from the capital budget. Component and equipment suppliers tied directly to data-centre builds move first. Power, grid, construction and cooling firms follow with a longer lag. Software companies monetising AI through subscriptions have diversified revenue but multiples set in the same enthusiasm.

Where did the major indices close most recently?

As of the last trade on Friday, 11 September 2026, the S&P 500 tracker SPY closed at $764.29, up 0.85%. The Nasdaq 100 proxy QQQ closed at $714.88, up 0.87%. The Dow tracker DIA closed at $525.79, up 0.97%. Goldman Sachs shares closed at 1,029.18, up 0.92%.

What should investors watch to see if the dependency is easing?

Watch for capex guidance language shifting from expansion to optimisation, supplier backlogs flattening while shipments hold, and whether non-AI sectors such as banks, consumer, industrials and healthcare accelerate their own earnings growth. Rising depreciation charges at the large buyers are the slow-acting counterweight.

Sources

• Goldman Sachs Says AI Is Driving Half of S&P 500 Earnings Growth. That Number Cuts Both Ways. — 24/7 Wall St

Photo: Andrea Piacquadio · Pexels Licence — source