Apollo's Slok Finds No AI Payoff Outside Big Tech
Apollo Global Management chief economist Torsten Slok has published a note stating that AI has delivered no financial payoff to corporate America outside the technology sector. The analysis appeared on October 24, 2026, under the title "No Signs of AI Payoff Outside the Tech Sector", as part of Apollo's Daily Spark series, in which Slok outlines key macro dynamics. Coverage of the note followed on October 25.
The note's core comparison uses what Slok calls the S&P 493, defined as the S&P 500 excluding Alphabet, Amazon, Apple, Meta, Microsoft, Nvidia and Tesla. Within that group, profit margins and revenue per worker appear flat, according to Slok, and he found no evidence that AI is improving productivity there. The seven excluded companies carry most of the AI capital spending narrative, so the split isolates everyone else.
In coverage of the note, Slok argued that the AI story will be decided in the S&P 493 rather than in the so-called Mag 7, the group of seven mega-cap tech names. Most public discussion of AI returns has centred on the handful of hyperscalers, the large cloud operators, and chip suppliers whose earnings have carried the index. Slok's argument is that those firms are not the test. The test is whether the other 493 companies, which employ most of the American workforce, start to show the gains.
The note fits a sequence of Daily Spark pieces on the same question. An August 9 analysis was titled "The Buyers of AI Are Still Waiting for the Payoff." An October 5 analysis was titled "No Signs of AI in the Productivity Data." An October 12 document, "The macro implications of the AI investment boom," asked "Where is the productivity payoff?" and found that AI is taking a growing share of corporate borrowing in the US. The pattern is consistent: capital is flowing into AI, financing is increasingly tied to it, and the productivity return has not appeared in the aggregates Apollo tracks.
The labour data point in a related direction. In remarks reported by Bloomberg on August 22, 2026, Slok said early labour-market data suggest AI's first impact is showing up more in wages than in employment. Bloomberg's headline described AI as weighing on pay without cutting jobs yet. That is a narrower claim than mass displacement.
In my view, the distinction between the two measures matters more than the headline. Flat margins and flat revenue per worker across 493 firms is a blunt reading. It does not separate companies that have deployed AI effectively from those that have spent heavily without results, and it cannot tell a reader whether the gains are simply lagging. Aggregate productivity statistics have historically trailed the adoption of general-purpose technologies by a wide margin, so a flat quarter-by-quarter profile is a weak basis for a verdict on the whole cycle. The more useful question is which of the 493 are moving, and whether the wage effect Slok describes becomes a margin effect.
The broader context here is a shift in where the debate is focused. For much of the past two years, the evidence cited for AI value came from the firms building the models and the infrastructure beneath them. Slok's note moves the measurement to the rest of the economy, the buyers rather than the sellers. If the payoff arrives in the S&P 493, it will show up in operating margins, headcount productivity, and the cost of capital for ordinary enterprises. If it does not, the spending will have been concentrated in a small group of suppliers. Either way, the note sets a scoreboard, and that is useful for anyone tracking the cycle.
