The MoatWide moat

Broadcom is a quiet toll collector on the plumbing of the modern world, and it is one of the more interesting moats to study precisely because so few ordinary people have ever heard of it. It makes the chips and software that sit at the heart of networks, smartphones, and data centers — the unglamorous, essential components that a customer designs into a product and then depends upon for years. These are not parts anyone swaps out lightly, because they are woven into the very heart of how a device works, and the risk and cost of changing them is out of all proportion to what the part itself costs. That is the foundation of the whole enterprise.

The mechanism is the design win. When an engineer selects Broadcom's chip for a new product, that decision does not last for one sale but for the entire life of the product — often many years — because the component is baked into both the hardware and the software around it. Switching to a rival's part later would mean re-engineering and re-testing the whole design, a costly and risky undertaking that customers avoid unless they have a compelling reason. So each design win is, in effect, an annuity, and the accumulated base of them across countless products is a moat a competitor must dislodge one design at a time.

Under its current management, the company added a second act that surprised many observers: it began buying established enterprise¹-software franchises — businesses with deeply entrenched customers and sticky, subscription-like revenue — and running them for cash. This has turned Broadcom into an unusual hybrid, part semiconductor maker and part software company, with the recurring revenue of the software smoothing the natural cyclicality of the chip business. It is a strategy that has few precedents and that has compounded shareholder wealth at an impressive rate.

The playbook driving all of it is unsentimental and remarkably consistent: acquire a business with a genuine moat, focus ruthlessly on its most profitable customers and products, cut away everything that does not earn its keep, and raise prices where the entrenched position allows. Applied again and again to carefully chosen targets, it has proven to be one of the most effective value-creation engines in all of technology — closer, in truth, to shrewd capital allocation than to ordinary chip-making.

What ties the two halves together is the common thread of mission-critical franchises whose customers cannot easily leave. Whether the product is a networking chip essential to how the internet moves data or an enterprise software system a large company has built its operations around, the source of the moat is the same: indispensability, and the enormous switching cost that comes with it. That is what gives Broadcom pricing power out of all proportion to the size of the components it sells.

The strategy carries genuine risk, and it has to be weighed. The same aggressiveness that makes the playbook so effective — the relentless focus on the best customers and the willingness to raise prices on captive ones — can, if pushed too hard, strain the very relationships the business depends upon and eventually drive customers to seek alternatives. And the acquisition machine runs on considerable debt, which demands both discipline and a steady hand. These are real considerations, not to be dismissed.

But taken as a whole, Broadcom is a business built on some of the stickiest revenue in technology — hardware embedded so deeply it cannot be cheaply removed, and software so entangled with a customer's operations that switching is scarcely contemplated — run by a management team that thinks like owners and allocates capital with unusual rigor. It is a less famous moat than a great consumer brand, but in its own quiet, technical way it is a formidable one, and it has rewarded those who understood it handsomely.