Fundsmith 2026 AGM: AI Mania, Big Tech Exit, and the End of Quality Investing?
Terry Smith and Julian Robbins discuss halving big tech positions, AI as the largest speculative craze in history, Novo Nordisk lessons, and why the portfolio is now cheaper than the S&P 500
Fundsmith 2026 AGM: AI Mania, Big Tech Exit, and the End of Quality Investing?
Fundsmith Annual Shareholders' Meeting February 2026
Transcribed by Thomas Chua | March 22, 2026
Terry Smith: "This has been a terrible year"
Terry Smith opened the 2026 AGM with unusual candor:
"If I had to use one word to describe the last 12 months, I would choose 'terrible.' Because I think if we carry on like this, we're all going to be poor."
The fund has underperformed for five consecutive years. Smith pushed back on the narrative:
"Five years ago, we achieved 22% growth. If you can do 22%, who cares what the market is doing?"
He emphasized that what matters is not the past 12 months, but what happens in the next 12 years. The team remains committed to their quality-focused strategy.
I. The Great Tech Exit: Halving Microsoft, Meta, and Alphabet
Why They Cut Big Tech
Fundsmith roughly halved their holdings in Microsoft, Meta, and slightly less so in Alphabet — their largest and most successful positions over the past decade.
The reason? AI capital expenditure concerns. Smith explained:
"These were companies with very high returns on capital, and they were very capital-light businesses. These [AI investments] are not going to be capital-light businesses. This is going to fundamentally change the characteristics of those companies in a way that we don't like."
The Math Problem
The hyperscalers are spending at extraordinary levels:
| Company | Capex as % of Cash Flow |
|---|---|
| Microsoft | 87% |
| Meta | 91% |
| Alphabet | 97% |
| Amazon | >100% (going into debt) |
| Oracle | 200% |
| CoreWeave | 567% |
Smith calculated the required returns:
"If we're looking for a 30% return on capital, those companies are going to need to generate $180 billion of new cash flow — not from the things they've already got, not by cannibalising what they already have — to justify that in terms of return on capital, per annum."
Who Wins in AI?
Smith was remarkably candid:
"We haven't got a clue" how to make money from AI. And he noted: "In aerospace, it wasn't anyone called Wright."
II. Is AI the Largest Speculative Craze in History?
Smith presented data comparing AI hype to historical bubbles:
| Bubble | Total Value |
|---|---|
| Top 5 AI Hyperscalers + Top 5 Semiconductors | $20 trillion |
| NASDAQ at peak (dotcom) | $12 trillion |
| Dutch East India Company | $8 trillion (adjusting for inflation) |
| Mississippi Scheme | $6.5 trillion |
| South Sea Bubble | $4.3 trillion |
"The short answer to the first part of the question would seem to be yes. It is the biggest."
Larry Ellison's Cloud Mistake
Smith pointed to Ellison's history:
"When the cloud began, he said: 'What is it? It's complete gibberish. It's insane. When is this idiocy going to stop?' As a result, he completely missed out on the cloud. He's now number four in a two-and-a-half horse race."
Smith argued this is driving Oracle's aggressive AI spending — "he's not missing the cloud twice."
III. Novo Nordisk: From Triumph to Tragedy
Smith's assessment of their most controversial holding was stark:
"Novo Nordisk has moved from a triumph to a tragedy. We bought it in 2016. We thought it was different to other drug companies... We proved to be right — they came out with the first weight loss drug. And they managed to snatch defeat from the jaws of victory in the core US market."
The Key Lesson
"We really don't like the uncertainty of the drug discovery process. Unlike branded goods, if somebody comes up with a more efficacious product, you're dead. You can't rely on moats — you can only rely on legal protection, and it doesn't work."
"The thing we've learned is we were right first off: don't own drug companies."
The K-Shaped Economy
Smith explained why Church & Dwight was also a detractor:
"People at the top of the K are doing very well and they don't buy Church & Dwight products. And people at the bottom of the K are doing very badly and they're trading down to own-label, or nothing at all."
IV. Portfolio Now Cheaper Than S&P 500 — First Time Ever
This may be the most significant statement of the meeting:
"I can now tell you we own a bunch of companies which are better than the S&P 500 and significantly cheaper. That's the first time I've been able to say that to you."
Key metrics show the shift:
| Metric | Fundsmith | S&P 500 |
|---|---|---|
| Return on Capital | 31% | 17% |
| Gross Margin | 62% | 43% |
| Interest Cover | 29x | 8-9x |
| Free Cash Flow Growth (2024) | 16% | — |
The valuation gap has reversed: their companies are now cheaper AND higher quality.
V. What They Bought and Why
New Positions
Intuit — They sold a few years ago after the "incredibly bad" Mailchimp acquisition. Smith:
"The shares went up initially because they were going to be an 'AI beneficiary.' They've now cratered and we've started buying them."
Zoetis — The world's largest veterinary pharmaceutical company. Complements their IDEXX holding.
EssilorLuxottica, Wolters Kluwer — Smith sees these as companies that will "survive and prosper better" when the AI bubble bursts.
They Sold
- PepsiCo — Weight loss drugs + lack of free cash flow growth
- Brown-Forman — Generation Z drinking habits, marijuana legalization, weight loss drugs
VI. The Psychology of Underperformance
Smith quoted John Maynard Keynes:
"It's better in career terms to fail conventionally than to succeed unconventionally."
He shared the story of Robert Sanborn, Morningstar Fund Manager of the Decade in the 1990s:
"He was on the front page of the Wall Street Journal in February 2000. He said he was getting hate mail, and his children were being bullied at school because of his performance."
When to Fire a Fund Manager
Smith cited Cambridge Associates data:
"If you look at managers' performance for three years before the change — the ones who were hired outperformed 4.8%. In the last year before the change, 7.2% for the stars, minus 3.7% for the ones you fired."
Research Affiliates data showed: "The time to fire a fund manager is when he's doing very well."
"The time to fire us, were you going to do so, was about 2021."
VII. The "Do Nothing" Strategy Under Pressure
Smith addressed the concern directly:
"The 'doing nothing' part is my concern. I think a lot could have been improved on in the last few years."
But he defended the core philosophy:
"We very much like the Charlie Munger mantra: 'Only invest in a business that can be run by an idiot, because sooner or later they all are run by an idiot.' What the Novo thing has painfully taught us is that there aren't very many businesses that can be run by idiots."
"We've done more in the last year, and I suspect the times are such that we will continue to do a little more — not to the same degree as other fund managers — because there are things causing change. It's not just AI, though it is the topic du jour. There are plenty of other things: weight loss drugs, the impact across the food and drink industry."
VIII. Most Admired CEOs
Julian Robbins chose Dr. Udit Batra of Waters:
"He didn't put out any snappy-named plans. He just went to people and said, 'Why don't we try something a little different?' The results were extraordinary in an incredibly quick period of time."
Smith chose Kevin Lobo of Stryker:
"They made an acquisition of a surgical robotics company called Mako. At the time, there was absolutely no way it worked on the numbers, and he basically said, 'Trust me.' He was right."
"Whoever said, 'If you build a better mousetrap, the world will beat a path to your door' is an idiot. No, they don't. You've got to actually go and sell them the mousetrap."
Key Quotes
"We're just trying to give you an explanation because I think that explanation is provided in order to give you the information to make an informed decision about what to do about this."
"If we had the market cap of Nvidia in that pie chart, it would actually be the second biggest piece of the pie — bigger than India and Japan."
"There must be something in the water in Denmark." — On both Novo Nordisk and Coloplast being detractors
Summary: What Changed in 2026 vs 2025
| Topic | 2025 AGM | 2026 AGM |
|---|---|---|
| Big Tech | Still held (Microsoft, Meta) | Halved positions |
| Novo Nordisk | Held despite 45% drawdown | Learned: don't own drug companies |
| AI | Skeptical but holding | Called it largest speculative craze |
| Valuation | More expensive than market | Now cheaper than S&P 500 |
| Strategy | "Do nothing" | "Doing more" — but not becoming traders |
Source
This article is based on the Fundsmith 2026 Annual Shareholders' Meeting transcript originally transcribed by Thomas Chua (March 22, 2026). Reprinted for educational purposes.
Transcribed by Thomas Chua | March 22, 2026 | Original source: steadycompounding.com