Apollo sees no broad AI payoff outside Big Tech—but its charts cannot isolate AI's effect
Revenue per worker and profit margins have diverged between the largest technology companies and the rest of the S&P 500. The aggregate comparison is a useful warning against easy productivity claims, not causal evidence about AI.
The short version
- What happened
- Apollo compares revenue per worker and margins for the Magnificent Seven with the other 493 S&P 500 companies and argues that gains remain concentrated in Big Tech.
- Who it affects
- Executives and workers should not treat aggregate market charts as evidence that a particular AI deployment will raise productivity or justify staffing changes.
- What remains unproven
- No firm-level AI-adoption measure
Keep in mind
No firm-level AI-adoption measure
No causal identification strategy
No uncertainty intervals
Commercial market commentary rather than peer-reviewed research
Source analysis · 1 linked source. Primary analysed source date: 24 October 2026.
Read the full report · 3 min The evidence, methods, implications and limitations
The divergence is real in the displayed aggregates
Apollo Global Management's 24 October commentary contrasts the Magnificent Seven technology companies with the remaining 493 companies in the S&P 500. Its Bloomberg and Macrobond charts show a widening gap in revenue per employee and profit margins, while separate San Francisco Fed and US labour data show no obvious economy-wide productivity acceleration. A Ramp spending series is used to illustrate corporate AI expenditure.
That combination supports a limited observation: the most visible operating gains in these aggregates remain concentrated among the largest technology firms. It also challenges the stronger claim that AI investment has already produced a broad, measurable productivity boom across listed US companies.[1]
Aggregate charts cannot tell us what AI caused
The comparison does not measure firm-level AI adoption, exposure or implementation quality. Big Tech differs from the rest of the index in business model, capital intensity, demand and pricing power, so its performance cannot be attributed to AI alone. The page reports no causal method, uncertainty intervals or denominator linking AI spending to output.
For workers and managers, the practical message is restraint: rising software spend is not proof of a productivity return. The assessment would change with pre-registered firm-level studies that connect verified adoption to comparable output, quality, wages and headcount over time. Apollo labels the article opinion, and its position as a global asset manager is a relevant commercial interest.[1]
The distinction that matters
What the charts can and cannot establish
The series describe groups and trends; they do not identify an AI treatment effect.
Observed
Compare Big Tech with S&P 493
Revenue-worker and margin gap
Not observed
Firm-level AI adoption
No adoption denominator
Not established
Causal attribution
Other differences remain