We’re about halfway into Q3. Nowcasts put equipment and structures investment — particularly forward looking investment — at near flat relative to Q2.
Figure 1: Equipment and Structures investment (blue line, left log scale), GDPNow nowcast of 8/7 (light blue square), both in bn.Ch.2017$ SAAR; EPU (news) (red, right scale). Source: BEA, Atlanta Fed (8/7), policyuncertainty.com, and author’s calculations.
See what consumer durables are doing, here.
Thanks, Drumpf!
If spending on artificial intelligence weren’t so strong, the picture would be worse. Here’s private nonresidential investment, computer and software investment, and nonresidential investment without computers and software:
https://fred.stlouisfed.org/graph/?g=1Lej0
Take out computers and software, which are flying due to the AI craze, and nonresidential fixed investment peaked back in Q2 of last year.
Which brings us to a potential alternative to a policy-induced recession. Roubini is talking up an AI boom:
https://www.businessinsider.com/recession-prediction-dr-doom-us-economy-nouriel-roubini-markets-outlook-2025-6?op=1
Superficially, stock investors seem to like the idea, but outside of returns to the AI sector, its hard to see where stock investors think AI will pay off. The Magnificent 7, meme stocks and crypo are running wild, while the rest of the equity market is mediocre. If AI isn’t going to boost returns in the wider economy, where does Roubini’s 4% growth come from? Roubini’s 4% growth call depends on productivity climbing 3% per year due to the flowering of artificial intelligence; sectors outside of AI ought to be able to ride that kind of productivity gain. I’d like to see Roubini’s math.