Torsten Slok, chief economist at US firm Apollo, argues that Wall Street is pricing in exaggerated profit growth for non-tech sectors. He notes that AI valuations for the S&P 493 index rely on rising margins, yet regulated industries like healthcare, banking, energy, pharma, and manufacturing face delays. Process overhauls and strict privacy requirements mean productivity gains could take years rather than months. If earnings lag behind market expectations, many AI stocks face painful repricing. Slok also warns that falling token costs might cap revenue for hyperscalers.
The core issue is that productivity in knowledge work remains difficult to measure. Even where individual employees become more efficient, the gains often disappear into daily operations without appearing on the balance sheet. Without clear metrics, management cannot act on the data. This disconnect means real cash flows may trail the projected earnings growth significantly.
* Metrics are missing for knowledge work productivity
* Management cannot act without clear data
* Real cash flows could trail projections
