Apollo’s Slok says US companies now treat AI spending as a cost, not growth

Apollo Global Management chief economist Torsten Slok said AI has delivered “zero financial payoff” to corporate America outside the tech sector. His note compared the Magnificent Seven with the other 493 S&P 500 companies, but the available article text provides no further figures or market reaction.

Analysis

The investable signal is not that AI has no value; it is that monetization may be accruing first to infrastructure and software vendors while customers have yet to show measurable productivity gains. That raises a second-order risk: enterprise AI budgets can remain protected initially, but renewals and expansion become harder to justify if savings, throughput, or incremental revenue do not appear in operating results. The pressure would fall most on vendors whose growth case depends on broad enterprise adoption, while infrastructure suppliers may face a later demand air pocket if customer ROI disappoints.

Treat the cited conclusion as a hypothesis, not a sector-wide measurement: revenue per worker can miss quality improvements, cost avoidance, and benefits with long implementation lags. The excerpt provides no sample methodology, period, or company-level dispersion, so it does not support a blanket short of non-tech equities.

Near term, the note may reinforce valuation dispersion between AI suppliers and adopters. Over 1–3 months, watch earnings commentary for quantified AI-driven revenue, labor productivity, and capex/opex changes. Over 6–18 months, the thesis strengthens if adoption costs persist without margin or revenue gains; it weakens if customer results become auditable and repeatable. A broad AI-adopter short is vulnerable to a sharp reversal if measurable benefits emerge or if labor scarcity makes automation valuable despite flat revenue per worker.

Key Decisions for Investors

• Do not trade this single, truncated analyst note as proof of zero ROI. Verify the underlying sample, measurement window, and whether results include cost avoidance and implementation expense.
• Avoid paying a premium solely for an enterprise-AI adoption narrative; require evidence in guidance or reported productivity, margins, or AI-linked revenue before adding exposure to purported beneficiaries.
• Watch upcoming earnings for quantified customer ROI, renewal/expansion trends, and whether AI-related spending is incremental or displaces other software budgets. These are the near-term catalysts for revising the thesis.
• A relative-value short of AI-adoption-sensitive businesses versus AI infrastructure suppliers is only a watch item, not a recommendation: establish company-level exposure first. Falsify the bearish view if multiple adopters report repeatable, material operating gains; reassess the supplier side if customer spending slows or guidance weakens.