Pieter Slegers — Buy-Hold-Sell List: July 2026

Dino Polska runs mid-sized supermarkets in Polish towns, close to where people actually live, and it builds them itself. There is nothing clever about the business — that is the argument for it. Groceries near your house are hard to disrupt, and the company opened 345 new shops in the past year while growing revenue 15%.

Over the same year the shares fell more than 40%, and they now trade at the lowest valuation in the company's history. Nothing in the numbers explains that: profits per share have grown 19.3% a year for a decade and the returns on the money it invests are comfortably above 15%. The explanation offered is attention — as the article puts it, nobody wants a boring Polish grocer while a rocket company is available at ninety times its sales.

The arithmetic of the case is given in full, which is rare here. Expect about 2.4 zloty of earnings per share in 2028; put a normal 20 times multiple on that and the shares would be worth 48 zloty against 28.7 today — roughly 70% higher, or more than 20% a year. The risk is contained in the same sentence: it depends both on the earnings arriving and on the market being willing to pay a normal multiple for them again.

Adobe makes the software professionals use to create images, video and documents — Photoshop, Premiere, Acrobat — sold as subscriptions. The market has decided artificial intelligence will make that kind of creative work cheap and therefore make Adobe less valuable, and the shares have been priced accordingly: about 8.5 times expected profits, against an average of nearly 31 over the past five years, and the cheapest the company has ever been.

Two things are offered against the fear. The first is that Adobe is using AI inside its own products rather than being displaced by it, and independent estimates still show double-digit revenue growth ahead. The second is a purchase: in late June a director of Adobe — who is also the chief executive of Eli Lilly — bought 10,000 shares at $194.50 of his own money. Insiders sell for many reasons and buy for one.

On the firm's own models this is the single most attractive name in the whole universe this month, with an expected return of nearly 26% a year, and the market is currently pricing in a business that shrinks slightly. It is also explicitly labelled higher-risk: the AI concern is real, and the case depends on it being wrong.

Ares manages money in things that are not listed on a stock exchange — direct loans to companies, private equity, property and infrastructure. It earns fees on the money it looks after, and as that pile has grown so have its profits and its dividend, which currently pays 4.6%.

It appears here because the firm's screen ranks it among the most attractively priced businesses on the watchlist: about 18 times expected profits against a "fair" 25, which together with the growth and the dividend implies roughly 20% a year. It has no formal Buy rating yet.

The context is worth noting: this is raised three weeks after the archive bought KKR, which does the same thing at a larger scale with an insurance company attached. Two names from the same industry appearing at the same time usually means the industry, rather than the company, is what has become cheap.

KKR appears in the portfolio table for the first time, three weeks after being bought at a limit of $98. It is already down: the shares are 27% lower than they started the year and the position is under water almost immediately.

What the table adds to the purchase case is the third model. Working backwards from today's price, the market is implying KKR grows profits by less than 1% a year, against an expectation of nearly 10% — the widest such gap of any of the nineteen holdings. In plain terms, you are being asked to pay a price that assumes the business essentially stops growing.

Inter Parfums licenses famous fashion names and makes and sells fragrances under them. It matters here less for what it does than for what it did this month: the shares rose 35.1% in four weeks, turning a badly negative year into a strongly positive one, and it is now the fifth-best performer in the whole watchlist.

That is the archive's first concrete evidence for a claim it has been making all summer — that when a style of investing comes back into favour, it does so suddenly rather than gradually. The caution is that three good weeks in three owned names is a very small sample, and the article presents it as encouragement rather than proof.

Schneider Electric makes the equipment that manages electricity inside buildings, factories and data centres — switchgear, controls, power management. It is added to the watchlist this month with no rating, no valuation and no argument, so there is nothing yet to agree or disagree with.

It is worth flagging for one reason: this is the first name in the archive's universe that benefits directly from the electricity demand of the artificial-intelligence build-out, which the same letters treat as the thing being dangerously overpaid for elsewhere. Owning the picks and shovels rather than the miners is a different position from avoiding the mine altogether.