What next?/Big picture with GenAI/BJ's and their growing coupon clippers

What next for Jay Powell? Well, it's now obviously clear what he'll do in three weeks and the October fed funds contract covering the September meeting has the odds at 90% (after initially getting to 100% immediately after his speech) of a 25 bps cut. After that he'll have 5 meetings left and I want to highlight again that while he's been the boss this year, the Fed dot plot has told us that a 1% REAL fed funds rate is what they consider 'neutral.' And I'll say again, the 3% nominal rate only is appropriate in their eyes IF inflation is sustainably at 2% and which instead is about 100 bps above that. Thus, with core PCE expected for July at 2.9% (with core CPI at 3.1%), one or two rate cuts gets us to that 1% REAL rate.

Thus, while Powell is in the seat, getting more cuts than two under his tenure will take another leg down in inflation and/or a much sharper slowdown in the labor market and coincident rise in the unemployment rate. We of course root for the former because rate cuts in response to the latter should not be hoped for and the two recessions over the past 25 years in the early aughts and the GFC did not help stocks under that circumstance. Either way, the market is pricing in a 100% chance of a total of 100 bps of cuts through next April with Powell as Chair (and 100% chance of two this year).

As we gear up from Nvidia's earnings because not only does the stock make up 8% of the S&P 500, a figure not seen for any one stock going back at least 45 years, but also what it means for the entire massive AI capital spending ecosystem which has been the main driver of economic growth over the past two years, outside of upper income spending and government largess. I want to make a point though as we think about Gen AI globally in the coming years because I don't think there is enough appreciation of the intensity of Chinese competition that American companies are going to experience.

We've already lit a fire under China by incentivizing them to build out their own semiconductor industry where in the coming years they will not need our chips at all. Cutting them off has been an epic fail but great for them as they are far more advanced technologically than they would have been otherwise. And, they are a huge competitor now in the creation of Gen AI models and many more than just DeepSeek (Moonshot Kimi, Qwen from Alibaba, MiniMax to name a few). Imagine being that shoestring start up in Singapore and you need access to a reliable, cheap GenAI model, who are you going to go to? The cheapest one most likely, assuming many models are pretty similar which they will be as the software becomes commoditized, and China will be able to provide that.

As an example of the new reality, in the Weekend FT, I read this headline, "Chinese semiconductor shares surge after DeepSeek boosts domestic chips." In it, "Cambricon Technologies led a rally of Chinese chipmakers yesterday after artificial intelligence start-up DeepSeek unveiled an updated model that would be compatible with domestically made semiconductors."

My point here is that US tech became a global, dominant powerhouse with high market share around the world, outside of China blocking Google, Facebook, Instagram, and Twitter (X). With semi's and with GenAI software, that is not going to be the case in coming years as China will be a tough rival.