There has been plenty of talk about elevated levels of market uncertainty this year, but until very recently equities have been contradicting this notion. When markets are exhibiting pronounced levels of dispersion driven by a singular theme – in this case AI – it is a sign of conviction rather than doubt. Yet the extreme moves we have witnessed seem like an exaggeration of the level of confidence anyone can have about how the future unfolds.

There are many excellent examples of the intense equity market dispersion that has occurred. It might be the crushing outperformance of momentum versus quality stocks, TSMC holding a substantially higher weight in emerging market indices than India, or more than 50% of the Korean market being made up by two companies. It seems fair to say that the AI trade has been running extremely hot.

The problem with the ferocity of the move in names within the AI eco-system is that it seems misaligned with the vast number of unanswered questions about the longer term impact of AI on markets and the economy.

Here are a few that spring to mind that I don’t have confident answers to, and I doubt anyone does:

– What level of ROI are companies seeing from AI implementation?

– What levels of ROI are companies hoping to deliver in the future from an AI rollout?

– How much are companies willing to pay for tokens when subsidies are reduced?

– What is the moat on an LLM? How do they ‘sustain’ high margins if their only edge is time to market?

– If AI is to be genuinely transformative, why were the corporate users of AI being left behind in the rally?

There are many more questions without clear resolution, and the dramatic level of market dispersion seems at odds with the profound uncertainties ahead.

This is not to say that the most bullish base case is wrong, rather that the probability of it occurring is not 100%.

Of course, I am being a little naïve here. The market conditions we witnessed in recent months had little to do with fundamentals or probabilistic nuance, and much to do with thematic performance chasing driven by an incredibly powerful narrative. Although there has been some respite in recent days, if that environment resumes the danger is that performance pressure eventually drags everyone in.

In times such as these the market becomes starkly binary. The questions are: is the rally justified or not? Is AI in a bubble? Are software company models irrevocably broken?

This is unhelpful and dangerous. The future is messy, complicated and uncertain, that is always the case – even if markets seem to be telling us otherwise. The most important question to answer remains this: am I appropriately diversified given all that I can’t possibly know?

—

My first book has been published. The Intelligent Fund Investor explores the beliefs and behaviours that lead investors astray, and shows how we can make better decisions. You can get a copy here (UK) or here (US).

All opinions are my own, not that of my employer or anybody else. I am often wrong, and my future self will disagree with my present self at some point. Not investment advice.