Pieter Slegers — Best Buys April 2026

KKR manages money for large institutions — buying companies, infrastructure and property with their capital — and collects an annual fee on everything it manages plus a share of the profits when investments are sold well. It also owns an insurance business, which supplies capital that never has to be handed back on a schedule. Its assets under management have tripled in five years, and reputation is the barrier to entry: a fifty-year record is what makes a pension fund pick you.

The stock has fallen close to 50% for three reasons, all of them about the future rather than the fee base: worries about a recession, the fact that KKR owns a lot of software companies at a moment when the market fears AI will damage them, and concern that the companies it lends to privately may struggle to repay. Against that, 2025 was its largest fundraising year ever at $129 billion, and $126 billion of it is still uninvested cash — which in a falling market is an asset, not a problem, because it lets KKR buy cheaply. And insiders keep buying the stock, which Slegers treats as the confirming signal.

3i is a UK investment company whose value is dominated by one holding: Action, a European discount retailer that sells cheap household goods and now makes up about three quarters of its private-equity portfolio. Action works the way Costco and Amazon do — every efficiency it gains from buying in bulk is handed to the customer as a lower price, which brings more customers, which allows even bigger bulk buying. The investor Nick Sleep gave this loop a name: Scale Economies Shared. The store count has been doubling every four to five years.

What separates 3i from an ordinary private-equity firm is that it does not sell its winners to collect a fee — it behaves like a holding company and keeps them for decades, which is how its net asset value doubled in three years.

The stock has nearly halved in six months for reasons Slegers states rather than skips: Action's sales growth has slowed from about 10% to about 5%, competitors in France (its second-biggest market) are forcing price cuts so sales there grew just 2%, Europe is filling up with Action stores, and 3i is committing €350–400m to an unproven US expansion. The result is that the shares now trade at a clear discount to the value of the assets they represent. The row uses the London ticker (III.L) because bare "III" belongs to Information Services Group.

MSCI gets paid twice for the same work. It builds the indexes that ETFs track, collecting a small percentage of the money invested in every fund with an MSCI name on it — more than $2.3 trillion and rising, because money keeps moving from active funds to passive ones. And it sells data and analytics software to banks and asset managers by subscription, growing about 10% a year with over 90% of customers renewing, because once its data is wired into a firm's systems, pulling it out is a project nobody wants.

Investors fear AI will erode the analytics business, or that firms will build their own custom indexes free. Slegers answers with two observations rather than an argument: MSCI is buying back its own shares consistently, so each remaining share owns more of it, and insiders have been buying recently while the stock sits near its cheapest valuation in ten years.

FICO owns the credit score used across American lending. Every mortgage, car loan and credit card application generates a small fee, and it also sells the software banks use to automate lending and fraud decisions. Ninety percent of top US lenders use it, and the fee is a couple of dollars against a loan worth thousands — which is why it has been able to raise prices for years without losing anyone.

This is the same company Slegers passed on a month earlier for being too expensive. It has fallen further, to near its cheapest valuation in a decade, and now makes the list. He dismisses the AI panic — no bank swaps a legally accepted score for an unproven model to save a few dollars a loan — but is careful to state the risk that is real: VantageScore, a rival built by the three credit bureaus, which the government-backed mortgage agencies Fannie Mae and Freddie Mac have been encouraged to accept. His judgement is that displacement of something this deeply embedded takes many years, and that the fear in the meantime is what creates the price.

MercadoLibre is the Amazon of Latin America, and increasingly its PayPal too. A little over half its revenue comes from taking a cut of everything bought and sold on its marketplace and from delivering the parcels through its own logistics network; the rest comes from Mercado Pago, its payments arm, which people across the region now use for groceries, fuel and bills as well as online shopping, plus lending and credit cards.

The pieces reinforce one another: shopping generates payments, payments generate data about who repays, that data lets it lend safely, and credit gives customers more to spend on the platform. Building its own delivery network in a region where shipping is genuinely hard is the part a rival cannot copy quickly, and e-commerce there is still so under-penetrated that the market is expected to grow 20% a year to 2033.

The stock is down more than 35% because the company chose growth over reported profit three times over: it slashed shipping prices to beat back Temu (more volume, thinner margins), announced a $14 billion investment plan for 2026 when investors wanted earnings, and is expanding lending quickly at a time when high inflation makes investors worry about defaults. For someone measuring in years rather than quarters, Slegers argues, that combination has produced one of the lowest valuations the company has ever traded at.

Cintas rents uniforms and supplies the mats, mops, restroom products, first-aid kits and fire-extinguisher inspections that over a million American businesses need to operate. Slegers calls it "a tollbridge on the physical workplace." Its advantage is unglamorous and very hard to copy: route density. With more than 12,100 delivery routes, the more customers it has on a given street, the less each stop costs in fuel and labour — a cost advantage a smaller rival can never match. Contracts run three to five years and over 90% of customers renew.

It is also trying to buy UniFirst, the number-three player, for $5.5bn, which would add 300,000 customers and $2.4bn of revenue at Cintas' much better margins — if regulators allow it.

And yet this is a pass, with a price attached. The stock is down over 25% from its high and still trades above 30 times forward earnings. "You rarely get it at a bargain price." He names the level he would act at: a forward P/E of 25, which means about $132 a share against $174 today. Until then it stays on the watch list — a good illustration that a quality verdict and a buy decision are separate things.

UniFirst appears only as the company Cintas is trying to buy for $5.5 billion — the third-largest player in uniform rental, bringing 300,000 customers and $2.4 billion of revenue. The point Slegers makes about it is uncomplimentary in a useful way: its margins are well below Cintas', which is precisely where the value of the deal would come from, since Cintas would apply its own operating model. Whether regulators allow the number one to buy the number three is, he notes, the open question. No view is offered on UniFirst as an investment in its own right.

Vestis is named once, as a measure of how lopsided the uniform-services industry would become: a combined Cintas and UniFirst would generate about $11.2 billion of revenue, and Vestis — the next largest competitor after them — generates roughly one fifth of that. It is used to size the competitive gap, not as a recommendation, and no view on the company is given.