Pieter Slegers — Portfolio Update 2026 (Part I)

Constellation buys small software companies whose products run one specific industry — a marina's booking system, a bus operator's scheduling — and never sells them, using the cash they throw off to buy more.

It is currently in the worst share-price fall of its entire life as a public company, because investors fear artificial intelligence will make this kind of software cheap to rebuild. It is also the smallest position in the portfolio, down about 20%, and rated the strongest thing on the list.

Two pieces of evidence support that. The numbers: the shares cost 17.7 times next year's profits against an average of 31.5 over the past five years, roughly half the usual price for the same company. And the behaviour: the chief operating officer, Mark Miller, has just put $3.6 million of his own money into the stock — the sort of purchase that is hard to explain if the people running it believed the disruption story.

Topicus is Constellation's European offshoot, doing the same thing across Europe — buying niche software businesses and reinvesting everything into buying more.

This entry is where the AI question finally gets a proper answer rather than a denial. Slegers concedes the premise: yes, AI makes writing software cheaper. His point is that writing the software was never the hard part. These products are welded into how a hospital or a council actually operates, so nobody switches; there is no realistic alternative, so prices can rise; and the supplier's real asset is knowing the industry and the people in it well enough to be trusted with something that cannot fail. A machine that writes code faster does not give you any of that.

Meanwhile the company is doing exactly what it should while its shares fall: buying Danish claims software, Polish hospital software, its first Romanian business, and building a 25% stake in Poland's largest IT company. Free cash flow keeps rising. The gap between what the business is doing and what the price says is the whole position.

Brown & Brown arranges insurance for businesses and individuals and takes a commission, never carrying the risk itself — so it needs almost no capital, the revenue repeats, and the cash conversion is excellent.

The useful disclosure here is not about the company but about how the position was built. Slegers wanted it for years and only bought a partial stake because it was expensive. When the shares fell by a third from their April peak, he completed the position in October. That is why a name he rates as strongly as anything he owns is showing a loss: the average was built on the way down, deliberately.

Working backwards from today's price, the company needs to grow cash flow just 5.9% a year to deliver 10% annually to an owner, against roughly 10% expected. That is the widest margin of safety anywhere in this half of the portfolio.

Inter Parfums licenses fashion names — Montblanc, Jimmy Choo, Coach — and makes, markets and distributes the fragrances that carry them. It owns almost no factories' worth of brand risk itself; it rents the names.

Everything hurting it is outside the company: shoppers cutting discretionary spending in Europe, retailers ordering cautiously, and tariffs of 15% on European exports to America and about 55% on Chinese components. The distinction he draws matters — sell-ins to retailers are weak while sell-outs to actual customers hold up, which means the problem is in the channel's inventory rather than in demand.

Meanwhile the pipeline of new licences keeps filling: Off-White from 2027, Annick Goutal from 2026, and Longchamp expected to reach $100 million of sales within three to five years. And the price now assumes none of it: working backwards, the shares are priced for cash flow to shrink slightly forever. Anything better than nothing is upside.

Visa owns the network that moves a card payment between your bank and the shop's bank and takes a sliver of each transaction. It lends nobody money and issues no cards.

The return here is assembled from four disclosed pieces rather than asserted: a 0.8% dividend, about 2.3% of the company bought back each year, and roughly 12% earnings growth — which together should double the share price in about five years. More than half of every dollar of revenue arrives as pure cash, which is what funds the first two.

The one live threat is political: a proposed American cap of 10% on credit-card interest rates, which knocked both Visa and Mastercard. Slegers dismisses it on the view that it would not survive a court challenge. Worth noting that this is a legal prediction rather than a business judgment — and the least evidenced sentence in an otherwise very numerate write-up.

HgCapital Trust is a London-listed way to own a private-equity portfolio without being an institution. Hg buys European software companies that are not listed anywhere, grows them and eventually sells them; the trust's shareholders get the gains plus a dividend.

It is the only holding valued on what it owns rather than what it earns. Hg publishes the portfolio's worth — 550.4p per share at the end of September, up from 539.5p in June, on £2.5 billion of net assets — and the shares change hands below that figure. Slegers' own note is that a quarter's move of eleven pence means very little, and that the case is a twenty-year one.

The standing caution is the same as for any discount-to-assets purchase: it is only real if the stated value of the unlisted holdings is honest. What supports it here is what those holdings are — software a customer's operations depend on, with revenue that repeats.

Judges Scientific buys small British makers of specialised laboratory instruments and runs them as a group. Its customers are universities and publicly funded labs.

That customer base is the problem right now. American federal research budgets have been cut, Chinese competitors are pricing aggressively, and the whole life-sciences equipment sector is weak — which is why this is the worst performer in this half of the portfolio, down 26%.

The bigger event is a handover. David Cicurel, who founded and ran the company for more than twenty years, becomes non-executive chairman in February; Tim Prestidge takes over. The case for him is that he has spent 22 years at Halma and Renishaw, two British companies that grow the same way — by buying and keeping small specialists — and has already been inside Judges for three years. Founder transitions are where roll-ups usually break, so this is the thing to watch rather than the funding cycle.

One honest flag: working backwards from the price, the shares need cash flow to grow 9.2% a year, against a long-term forecast of 6.6%. Slegers marks that as questionable and keeps the Buy anyway, on the strength of the succession.

Brookfield owns physical infrastructure — power generation, grids, ports, property, renewables — partly with its own money and partly with other people's, charging fees to run it.

The news in this update is a new business called Radiant: a cloud service aimed at competing with Amazon's AWS. That sounds implausible until you see the angle. The binding constraint on AI data centres is electricity and land, and Brookfield already owns both. It intends to build capacity for its own use, sell the surplus, and undercut competitors precisely because it is not renting the inputs. The target customers are corporations and governments wanting their own sovereign AI capacity.

Management's stated goal is to grow the company's underlying value 15% a year, which would double it every five years. And here is the honest wrinkle: on Slegers' own multiple test this is the one position in the group that fails — 20.2 times distributable earnings against a bar of under 12 — and he rates it Buy regardless, calling it "the ideal cornerstone for every Portfolio." That is a judgment overriding a checklist, and it is worth seeing it stated plainly.

OTC Markets runs the trading and data platform for American shares that are not listed on the New York Stock Exchange or Nasdaq, charging companies and brokers for listings, data and compliance tools.

This is the most interesting entry in the update because everything measurable says buy and the verdict is not. The shares are cheaper than their own five-year average, the return model clears the hurdle, and the price implies less growth than the company is expected to deliver. Profitability is excellent — 27 cents of every revenue dollar is profit — and more than 5% of the share price comes back each year in dividends and buybacks.

And Slegers writes: "OTC Markets is the company we own that I'm the least sure about… growth has stalled and it looks like it's a more structural problem than initially thought." A cheap price does not help if the earnings behind it stop growing permanently — the value trap in its purest form. His conclusion is about opportunity cost rather than valuation: "there might be better investment opportunities in the market today."

Seven days later he names one. The position is sold and the money goes into Zoetis.