Pieter Slegers — Best Buys: August 2026

Mastercard does not lend money and does not take credit risk. It owns the wires between shoppers, shops and banks, and takes a very small slice of every payment that travels along them. Because the wires are already built, an extra transaction costs it almost nothing, so nearly all of the extra revenue drops through to profit.

Nobody can build a competing network, because of a circular problem: shops accept Mastercard because everyone carries it, and everyone carries it because all the shops accept it. A newcomer has to solve both halves at once, which is why this has been a two-company industry for decades.

The part that is growing fastest is not the payments. Having every bank and large merchant already plugged in, Mastercard now sells them fraud detection, cybersecurity and data analytics — extra services sold down an existing pipe, which is the cheapest kind of growth there is. Ranked the month's number-one candidate, though with no price or valuation attached.

TransDigm makes small, unglamorous aircraft parts — pumps, valves, ignition systems, cockpit hardware. The whole business rests on one fact: for about 80% of what it sells, it is the only manufacturer certified by the regulators to make that part. An airline cannot legally fly without certified parts, and cannot buy them anywhere else.

Most of the money comes not from selling parts to new aircraft but from replacing them for the next twenty-five to thirty years, which is far more profitable. And the company has an unusually explicit method for expanding: buy a small supplier that holds a sole-source certification, cut its costs, tie the managers' pay to the profit of each product line, then raise prices to reflect what the part is really worth to someone with a grounded aeroplane. It has done this hundreds of times and says there are hundreds more targets.

The current tailwind is a shortage elsewhere: Boeing and Airbus cannot deliver new aircraft fast enough, so airlines keep flying old ones, and old aircraft need replacement parts. As with the rest of this month's list, no valuation is offered.

When a company wants to borrow money in the bond market, it effectively must be rated by at least two of the three agencies that exist — Moody's, S&P Global and Fitch. That requirement, backed by regulation and investor mandates, turns Moody's into a tollbooth: it collects a fee on borrowing it does nothing to cause.

The weakness of that model is that bond issuance is lumpy — in some years companies borrow heavily, in others barely at all. Moody's answer is its analytics arm, which sells risk and compliance software on subscription, producing revenue that arrives whether or not anyone issues a bond.

The reason it appears now: the shares have gone roughly nowhere for a year while revenue and earnings kept growing, which mechanically makes it cheaper. The letter says this is the most attractive valuation since 2020, though it does not print the multiple; the previous week's spreadsheet showed about 28 times forward earnings against a five-year average of 33.

Canadian National runs 19,500 miles of track and is the only railway in North America touching the Atlantic, the Pacific and the Gulf of Mexico. The moat is that nobody will ever build another one: the land, the planning permission and the environmental approvals are effectively unobtainable today at any price. What exists, exists.

The economics are simple. Moving heavy goods long distances by rail uses about a quarter of the fuel of trucking, so for bulk freight there is no real alternative — which lets the railway raise prices at or slightly above inflation, year after year, through good times and bad.

The reason it is a candidate now is that the volumes are weak: freight traffic has been falling for several years. Management has been squeezing costs to compensate and raised its guidance at the last quarterly results. So this one is explicitly a cyclical entry — buy a business that cannot be replaced while the cycle it serves is depressed, and wait for volumes to recover. That is a different kind of bet from the other four on the list, and the letter is straightforward about it.

Hermès makes Birkin bags, silk scarves, watches and perfume, and its central business decision is to make fewer of them than people want. That refusal to meet demand means it never discounts, never needs outlet stores and never carries unsold stock — which is why its margins are the highest in luxury. The proof that the scarcity is real is the second-hand market, where used bags routinely fetch more than new ones.

The family still owns more than 65% of the shares, which is the strongest possible guarantee that nobody will be tempted to boost this year's sales by making more bags and quietly spending the brand.

The opportunity is that the shares have fallen nearly 30% after growth slowed in 2026. The letter's argument is that the slowdown is largely a currency-translation effect: measured in constant currencies the business is still growing. After years of being priced for perfection, "the valuation is finally coming down to more reasonable levels" — though as with the rest of the list, no actual multiple or fair value is given here.

Watsco is the middleman for air conditioning and refrigeration in North America: it buys equipment and parts from the manufacturers and stocks them close to the 120,000-odd contractors who install and repair them. Between 70% and 80% of what it sells is an emergency — a unit has failed and someone needs the part today. In that situation the contractor cares about availability, not price, which is why a distributor with local inventory has real pricing power.

It is also a consolidator. It sells about one in five residential systems in America, more than 2,000 regional distributors remain independent, and it does not yet operate in every state. Its ordering software, which now carries a third of revenue, makes contractors more profitable and therefore stickier. The business needs very little capital, has little debt and earns over 15% on the capital it does use; the shares are up roughly 12,000% since 1990.

The shares are nonetheless down more than 45% from their high, because revenue and profit have been falling. The letter's diagnosis is that three temporary distortions are unwinding at once: pandemic demand that was pulled forward from later years (plus dealers working off the inventory they over-ordered), a refrigerant regulation change that scrambled supply and briefly inflated demand, and manufacturers' post-pandemic price rises normalising after having flattered both sales and margins. None of those is a statement about the franchise.

The valuation is the only one in the issue, and it is worth seeing how it is built: management targets $10bn of revenue; assume an 8% net margin, so $800m of profit; assume the market pays 25 times that, so $20bn — 56% above today. Two of those three inputs are assumptions, and there is no date by which any of it must happen.