Is AI Overhyped? Apollo Economist Sees "Zero Financial Payoff" Beyond Tech Companies

Apollo economist Torsten Slok’s research says revenue per worker and profit margins are rising for the Magnificent Seven but remain flat for the other 493 S&P 500 companies, indicating AI has delivered no financial payoff to corporate America outside tech so far. If AI spending fails to generate broader gains, highly valued tech stocks could face a sell-off; the article notes this outcome is uncertain and cites more optimistic views from Vanguard. The Vanguard S&P 500 ETF returned about 13.8% annualized over the past five years, including 2022, when the S&P 500 fell about 18.2%; it remains exposed to major tech stocks but may decline less than tech-heavy ETFs in an AI bust.

Analysis

The key risk is not that AI has no value, but that returns accrue unevenly: model and infrastructure sellers can monetize spending before customers realize measurable productivity gains. That leaves the AI supply chain exposed to a second-order reversal: if adoption fails to translate into customer ROI, buyers may delay renewals or deployments, weakening future data-center demand and putting pressure on the earnings expectations embedded in AI-linked equities. Alphabet, Microsoft, and Meta also face the possibility that heavy investment earns inadequate returns, even if their existing businesses remain resilient.

Treat Slok’s aggregate productivity indicators as a warning signal, not proof of failure. Benefits may arrive with a lag, and index-level revenue-per-worker and margin measures can obscure differences across industries. A near-term market reaction may therefore be limited; the more important test is whether upcoming company commentary shows sustained customer adoption and returns on AI investment. Over 6–18 months, broad-based margin improvement would weaken the bear case; continued capex growth without evidence of monetization would strengthen it.

The contrarian point: broad-market exposure reduces concentration but does not hedge an AI unwind, because large technology constituents remain embedded in the S&P 500. A relative-value hedge is cleaner than an outright bearish bet while the productivity evidence remains inconclusive.

Key Decisions for Investors

• For the next 1–3 months, consider trimming concentrated AI/mega-cap exposure in favor of VOO rather than treating broad-market exposure as crash protection. This reduces, but does not eliminate, sensitivity to an AI-led selloff.
• Use a relative-value framework: underweight a tech-heavy ETF versus VOO, rather than shorting the AI theme outright. Reassess after earnings commentary on AI customer adoption, monetization, and capex returns; avoid adding to the hedge if evidence of productivity gains broadens.
• Watch whether AI infrastructure spending plans are maintained while customer demand and measurable efficiency gains remain weak. A capex slowdown or disappointing utilization would be a negative catalyst for infrastructure suppliers; broad-based margin or revenue-per-worker improvement would falsify the bearish thesis.
• Do not infer a trade in APO from this research alone: the article offers no evidence that Apollo’s own earnings or assets are materially exposed to the finding.