Thursday was the Nasdaq’s worst day since April 2025: down 5.5% and into correction territory, with the S&P 500 off 4.4% and the Dow down 1,833 points. Nvidia lost 9%, Broadcom 10%, the Philadelphia Semiconductor Index 12%. By Friday morning the Nikkei had opened down 7% and Korea had halted trading. Fox Business put it plainly: analysts think “sentiment is changing, which could portend further selling.” One portfolio manager called it “the capitulation phase for the data center capex beneficiary stocks.”
The trigger was political (the midterm result read as more regulation and more scrutiny of who pays for the data-centre buildout), but the kindling has been there for months. A very large amount of capital has been committed to AI infrastructure on the assumption that the returns will follow. This week the market started asking for the receipts.
We have written before about why we like companies that use AI without being a bet on AI. This week is the practical version of that argument, so it is worth saying what our portfolio actually looks like.
We hold 23 companies. They include a surgical robot for MRI-guided neurosurgery, a 3D-printed living seawall, a secondary market for private real-estate stakes, a spatial-biology platform for oncology drug development, an aortic-arch surgery device, a neuromodulation therapy for heart rhythm, a cybersecurity hardware company whose product works by being physically disconnected, a histology lab, a passive health sensor, and a basketball coaching tool. Several of them use machine learning somewhere in the product. Not one of them sells GPUs, builds data centres, trains foundation models, or depends on hyperscaler capex for its revenue. Their customers are hospitals, surgeons, manufacturers, coastal cities, biopharma R&D teams and property investors.
That is not an accident. Early-stage investing already carries plenty of risk; we see no reason to add a correlated macro bet on top of it. The question we ask of every company is whether it would still be a good business if the AI narrative went quiet for three years. If the answer depends on the narrative, we pass.
None of this makes our companies immune to a weaker funding environment, and a rout in public AI names can tighten private markets for everyone. But there is a difference between a portfolio that is exposed to the weather and one that is built out of the thing the storm is hitting. Ours is the former. We did not build it to ride the AI trade, so we are not spending this week working out how to get off it.
Article: https://www.foxbusiness.com/markets/stocks-nov-5-2026
This post is for information only and is not an offer or solicitation of any investment, nor a prediction of the performance of any company or fund.