Latest / Business Strategy Talks with Fexingo: Competitive Advantage, Market Positioning, and Growth Planning / How a 50-Year-Old Coatings Patent Built an Unbreakable Moat
Transcript
- Lucas: So there's this coating — it's called titanium aluminum nitride, or TiAlN — and it's applied to nearly every industrial cutting tool you've ever used, from drill bits to end mills to the inserts that machine engine blocks. Luna: And I'm guessing this isn't just any coating. Lucas: Right. There's a specific formulation — a particular ratio of titanium to aluminum, plus a patented layering technique — that was filed as a patent in 1974 by a small German chemical company named Hertel. Luna: 1974. That's over fifty years ago. Lucas: Exactly. And that patent, with some clever extensions and related filings, has essentially blocked every competitor from using that exact chemistry for half a century. The company that owned it was eventually acquired by Sandvik, the Swedish engineering giant, and Sandvik still holds the core intellectual property. Luna: So how does a fifty-year-old patent still enforce a moat? Patents expire after twenty years in most jurisdictions. Lucas: That's the clever part. The original composition of matter patent expired in the mid-1990s. But Hertel filed a series of process patents — methods of deposition, specific temperature and pressure ranges, thickness profiles — that created a web of protection. Each one expired at a different time, and some were extended through regulatory delays and divisional applications. Luna: So it's not one patent, it's a family. Lucas: Exactly. And the final key patent in that family — the one covering the most commercially relevant deposition method — expired in the European Union in 2016 and in the United States in 2018. Luna: Wait, so the moat ended nearly a decade ago? Lucas: That's what most people assume. But here's the thing: the know-how required to actually apply the coating at industrial scale — the precise control of the vacuum chamber, the pre-cleaning steps, the quality metrics — was never fully patented. It was kept as trade secret. And Sandvik has been iterating on that process for fifty years. No competitor has matched the consistency. Luna: So the patent expired, but the tacit knowledge locked the door. Lucas: Exactly. And the numbers bear it out. Sandvik's cutting tools division — which is about forty percent of their revenue — has consistently reported gross margins between 60 and 65 percent. That's higher than most software companies. Luna: Right, because they're not just selling a coating. They're selling the performance guarantee that comes with it. Lucas: And that's the real moat. A Japanese competitor, Mitsubishi Materials, tried to develop a competing TiAlN coating in the late 2000s. They spent something like two hundred million dollars on R&D and pilot production. By 2009, they had a product that they claimed was equivalent. Luna: And was it? Lucas: Independent tests by a German automotive manufacturer showed that the Mitsubishi coating had a tool life about 18 percent shorter on average than Sandvik's equivalent grade. For a high-volume production line, that difference translates into millions of dollars in downtime and tool-change costs. Mitsubishi eventually pulled the product. Luna: So even with a clean room and the chemical formula, you couldn't replicate the results. Lucas: That's the moat. It's not the patent. It's the fifty years of process refinement that never got written down in a patent application. And Sandvik has been investing that 60-plus-percent gross margin right back into process R&D. Luna: Let me ask a question that might be on listeners' minds: does this apply to the current market environment? Mid-2026, we're seeing a lot of industrial companies struggle with demand softness. Lucas: It actually makes the moat more relevant. When end-market demand weakens, customers don't stop cutting metal — they just become more sensitive to tool costs. But Sandvik's coating extends tool life by 3x to 5x compared to uncoated tools. So even at a premium price, the total cost per part is lower. Luna: So the moat actually widens in a downturn. Customers trade down to cheaper consumables, but in cutting tools, the premium product is the cheaper option on a per-part basis. Lucas: Exactly. And that's why Sandvik's cutting tools revenue only dipped about 5 percent during the 2020 downturn, while many competitors saw 15 to 20 percent drops. Luna: I want to zoom out a bit. You mentioned earlier that the strongest moats often come from materials science, not software. Can you elaborate? Lucas: Sure. Software moats are real — network effects, switching costs, data advantages — but they tend to erode faster. A new framework, a new API, a new regulation can disrupt them. Materials science moats are built on atoms, not bits. The physical properties of a coating, the exact crystal structure — those are incredibly hard to reverse-engineer and even harder to manufacture at scale with consistency. Luna: And they're often protected by decades of accumulated learning. Lucas: Right. Think about it this way: the original TiAlN coating was discovered by accident. A researcher at Hertel was trying to improve wear resistance and tried a different ratio of titanium to aluminum. The resulting film had a hardness of about 30 gigapascals — nearly as hard as cubic boron nitride. That discovery was serendipitous. And the subsequent fifty years of optimization — how to deposit it uniformly, how to control the columnar grain structure — that's not something you can replicate by reading a paper. Luna: So for an investor, the lesson might be: when you find a company with a materials science moat, pay attention to the vintage of the original discovery and the investment in process R&D since then. Lucas: Exactly. And check the gross margin stability. If it's been above 55 percent for a decade, and the company is still spending 4 to 5 percent of revenue on process R&D, that's a signal that the moat is being actively maintained. Luna: It's interesting — we often talk about 'moats' as if they're static, like a castle wall. But this example shows they require constant investment. Lucas: Absolutely. The moat is a process, not a thing. Sandvik's coating division files about 40 new patents per year, mostly on deposition methods and quality control. They're not resting on the 1974 patent. Luna: And if they ever stop innovating, the moat starts to erode. Lucas: Right. But that's a pretty good problem to have when you've got a fifty-year head start. Luna: It reminds me of something we talked about in episode 108, with Vanguard's moat on low cost. There too, the moat is maintained by constant reinvestment in efficiency. Lucas: Good connection. But the difference is, Vanguard's moat is built on scale and organizational structure. Sandvik's is built on physics. And physics doesn't change. Luna: Unless someone discovers a new coating that makes TiAlN obsolete. Lucas: And that's the risk. There's research into aluminum oxide coatings and diamond-like carbon coatings that could, in theory, outperform TiAlN in certain applications. But so far, no one has matched the combination of hardness, thermal stability, and cost that Sandvik's TiAlN offers. And even if someone does, Sandvik has the manufacturing scale and customer relationships to adapt quickly. Luna: So the moat is sticky. Lucas: Very. And that's why, even in a market like mid-2026 where industrial production growth is slowing, Sandvik's cutting tools business continues to generate strong cash flow. It's a reminder that the best moats are often the ones that are hard to see — hidden in the chemistry of a coating, not in a brand logo. Luna: Before we wrap, I want to come back to something you mentioned earlier about the R&D investment. Can you put a number on what Sandvik spends on process R&D for coatings? Lucas: Roughly 4.2 percent of cutting tools revenue goes to R&D, which in 2025 was about 280 million dollars. Of that, the majority is process development — not new coatings, but better ways to apply the existing coating more consistently, faster, and with less waste. Luna: So they're not trying to invent the next TiAlN. They're trying to make the current one even harder to replicate. Lucas: Exactly. And that's a smart strategy. Because the incremental improvement in tool life — even 2 or 3 percent per year — compounds over time. Over a decade, that's a 20 to 30 percent performance gap that a new entrant can never close unless they also invest for ten years. Luna: And if they try, Sandvik has the cash flow to outspend them. Lucas: And the patent portfolio to sue them. It's a triple moat: legal, technical, and economic. Luna: It's a good reminder that when we look for durable competitive advantages, we should look beyond the usual suspects — brands, network effects — and into the nitty-gritty of how things are actually made. Lucas: Absolutely. And speaking of things that are made, I should mention that if today's episode added a little clarity to your thinking about moats, especially the less-visible ones, and you'd like to support the show staying ad-free, you can find us at buy me a coffee dot com slash fexingo. Luna: Yeah, it's a small gesture that goes a long way in keeping these conversations going. Lucas: Okay — back to the coating world. One more angle I want to touch on: the supply chain implications. Because Sandvik's coating isn't just a competitive advantage for them — it's also a strategic dependency for their customers. Luna: Meaning what? If Sandvik raises prices, customers can't easily switch? Lucas: Right. In the aerospace industry, for example, a major engine manufacturer might use Sandvik's coated tools for 70 percent of their machining operations. The tools are qualified through a lengthy process that involves testing thousands of parts. Switching to a different coating, even one that's 90 percent as good, would require requalification that could take 18 months and cost millions. Luna: So the moat extends beyond the product itself into the customer's own processes. Lucas: Exactly. And that's what makes it so durable. The customer is locked in not just by performance, but by the time and cost of switching. Luna: That's a powerful combination. And it's a great example of a moat that's been built patiently over decades, not something that can be replicated overnight. Lucas: And that's the key takeaway for today. The next time you hear about a company with a 'moat', ask yourself: is it based on a patent that will expire, a brand that can fade, or is it based on fifty years of accumulated process knowledge that literally cannot be written down? The latter is the one that lasts.