Bayesian Investor Blog

I’ve been shifting my investments to more AI-focused bets.

Here’s an overview of my larger positions, in descending order of how eager I am to recommend further purchases now:

• MU – AI is memory hungry, and MU is in a good position for almost NVDA-like growth
• CSIQ – solar, for AI-related power demand
• ASML – semiconductor equipment
• SCIA – semiconductor equipment
• GOOGL – for TPUs, DeepMind, and Waymo
• SMCI – datacenter
• NVDA
• AMPX (and AMPX.WS) – batteries suitable for drones
• TSSI – datacenter
• CLS – close to half of its business involves datacenters
• PDEX – a fairly safe way to diversify my portfolio
• MTG – a fairly safe way to diversify my portfolio
• ASTS – expanding cell phone coverage; I’ll likely sell when I can get long-term gains

Short positions:

• SP500 futures – hedging against risks such as tariffs
• SOFR futures dated 2029 through 2032 – betting that interest rates will rise due to AI
• WMT – it’s got a high PE ratio, and little sign of growth

[This is not at all a complete list, as I have smaller positions in something like 150 other companies.]

AI stocks are likely to form a bubble someday, but I’m guessing the peak of that bubble is more than a year away.

I still have some concerns about tariff-related damage causing some declines sometime this year. I’m optimistic that the courts will strike down the per-country tariffs this fall. It shouldn’t take long for the country to recover from the tariff damage once the tariffs have been removed.

Beware that even in a strong bull market, there will be periodic scares, such as January’s DeepSeek-trigger panic, that cause sharp drops in leading stocks. AI-related stocks had a big rally in June, and are likely to consolidate for a while before the next such rally.