In this article:

The capital intensity of the AI investment we’re seeing is very high. The hyperscalers – Meta, Alphabet et al – are investing at what for them is an unprecedented rate and making significant use of external funding. Meanwhile the market has priced a lot into the value of the semiconductor companies. For their value to keep rising, one would have to believe there’s another upswing in investment to come. I’m not clear how that can occur.

I am not anti-AI in general. Rather I’m looking for the parts of the market that may do better out of the development we’ll see in the quarters ahead. There will clearly be benefits from all of this investment but not all of it will be profitable, and there are questions now about the circularity of what is happening.

I still believe in the hyperscalers themselves. They are very good businesses, but I have to look at things from a valuation perspective. Not all appear good investments to me through that lens.

A lot of software players are in a downcycle and the read for some of these companies has been that they will be disrupted by AI. But I think some of the existing software players, those that are embedding AI into their systems, will be very well placed to profit from enterprise adoption of AI.

There are semiconductor equipment manufacturers which have not benefitted strongly from the AI trade. All of these stand to gain from the likes of TSMC, Samsung, and others adding to capacity. Most of the chip production so far has come from existing capacity; ramping that up will involve investment flowing through more of those equipment businesses.

The opportunity set across technology remains broad and continues to evolve. Chinese tech players are trying to build their own infrastructure while the consumption recovery there supports lots of domestic names in areas like travel and e-commerce. Video gaming and music streaming look like compelling growth areas: engagement from users is high but, so far they’re under-monetised.

Geographically, I’ll keep looking for a spread of holdings in 2026. Tech changes are happening globally; the companies worth investing in are everywhere from China to Japan, from the US to Korea.

Hyun Ho Sohn

Hyun Ho Sohn

Portfolio Manager