Lego

Lego makes bricks and sells them in boxes, and the reason that sentence is not the whole case is that the bricks are a system. Every element made since 1958 fits every other, which means each box a child owns raises the value of the next one. That compatibility is the product. It is also why the near-death of 2003–2004 happened: under pressure from the expiry of its patents and the rise of video games, Lego spent the late 1990s inventing things that were not the system — Galidor action figures, Clikits jewellery, a television studio, clothing, theme parks, software — and doubled the number of unique elements it moulded to roughly 13,000. The company lost DKK 1.9 billion in 2004 on revenue of about DKK 6.3 billion and was, by its own account, weeks from breaching its covenants.

Jørgen Vig Knudstorp's turnaround was a return to the system with discipline added. The element count was cut roughly in half, colours were cut, the parks were sold, and every product had to earn a 13.5% return on sales or be dropped. Then the company did what it had failed to do in the 1990s: it found new customers for the old product. Licensed themes — Star Wars from 1999, Harry Potter from 2001, then Marvel, Minecraft and the rest — put someone else's story on Lego's bricks. Lego Ideas turned fans into designers. And adults, who had always bought Lego quietly, were given a line of their own; the 18+ sets are now one of the fastest-growing parts of the range and the most profitable per box.

The result in 2024 was revenue of ≈ DKK 74.3 billion and operating profit of ≈ DKK 18.7 billion, a 25% margin, from a product whose core has not changed in sixty-seven years. Kirkbi, the family holding company, owns three-quarters of the group and the Lego Foundation the rest; the company has no shareholders to answer to and publishes full annual reports anyway, which makes it the most transparent private company in this library.

The instructive thing about Lego's moat is that the company tried to defend it in court and lost, and the moat held anyway. The stud-and-tube patent filed in 1958 expired in the late 1970s and 1980s. Lego then spent twenty years arguing that the brick's shape was a trademark, and in 2010 the European Court of Justice ruled against it in a case brought over Mega Bloks: a shape that is necessary to achieve a technical result cannot be a trademark. Anyone may now make a brick that fits a Lego brick, and many do, at a fraction of the price. Lego's share of the construction-toy market has not moved.

That is because the moat was never the brick. It is the installed base — every brick ever sold fits every brick sold today, so a child's collection is a reason to buy Lego rather than a clone — and it is the manufacturing tolerance, which the company measures in thousandths of a millimetre and which the clones have never quite matched. On top of that sits a brand that parents trust with the same certainty that they trust the tolerance, and a licensing position that gives Lego, not its imitators, Star Wars and Harry Potter. A compatible brick from a rival is a substitute for the plastic. It is not a substitute for the box.

Where it is weaker: the licences are borrowed, and the licensors know what they are worth. The adult market is a real growth engine but also a fashion, and a fashion can turn. And Lego's cost base is Danish; the company has held its margin partly through scale and partly through pricing, and it has raised prices repeatedly. A parent who notices that the same number of bricks costs half as much from the clone is the one customer the trust has to keep winning.

Lego's allocation record has two halves and the line between them is 2004. Before it, the company under Kjeld Kirk Kristiansen spent the late 1990s buying and building things that were not bricks: theme parks in Windsor, California and Germany, a video-game studio, a clothing line, a television venture, action figures with no bricks in them, and a range that doubled its unique elements. Each one was a reasonable answer to a real fear — that the patents had gone and the screens were coming — and together they nearly ended the company.

After it, under Knudstorp and then Niels B. Christiansen, the rule has been to invest in the brick and its system and almost nothing else. Capital expenditure goes into moulding plants close to the markets they serve — Hungary, Mexico, China, Vietnam, Virginia — and into Lego's own stores, which passed a thousand in 2024 and let the company keep the retail margin. The parks were sold in 2005 to fund the recovery and licensed back; the family bought a stake in the operator, Merlin, and took it private with Blackstone and a Canadian pension fund in 2019 for roughly £6 billion. That sequence — sell what you cannot run, keep the option, buy it back when you can — is the most instructive piece of the record.

The standing criticism is that Lego has let a large amount of capital accumulate at Kirkbi rather than reinvesting it in the toy business or returning it. That is a fair point about a private company with no obligation to answer it. What the family would say is that the cash is the reason 2004 will not recur.