Great example, Intel is the De Facto standard CPU for all computers (even servers). It is tough to port software and system from one CPU vendor to another. Is SWKS in the same position? Not Yet
Microsoft Windows and Office have a huge switch moat, I wish SWKS had the same switch moat too.
Well, let’s unpack that a bit.
First of all, I was not suggesting that Skyworks’ moat is as deep as either Windows or Intel in PC’s. The point was that products with a very deep moat can exist inside of a commodity product. Therefore the implication that just if/because smartphones become commodities in no way necessitates that Skyworks’ solution there would be as well.
Secondly, I think they may be more secure in their position than many suppose. In particular, with Apple, who last year accounted for 44% of their business (see page 62 of 10-K), I think they have a very secure position. Looks at this from Apple’s perspcective: first of all, clearly Apple has already set-up Skyworks for the iPhone 7. Well, it would be simply stupid to switch to a competitor’s solution for 7S. So that puts us out to iPhone 8, which is already into the design and development stage. Apple has to assume they’ll produce about a quarter of a billion iPhone 8’s in a year. The solution that Skyworks is critical to a quality smartphone, and is a tiny fraction of the bill of materials. Maybe most importantly, Apple has to be able to count on it’s supplier to be completely trustworthy to supply a quarter of billion product in a year.
People seem scared because Apple accounts for 44% of Skyworks sales. Yet, iPhone accounts for nearly 75% of Apple’s sales. Apple simply cannot and would not risk the health of their goose that lays the golden eggs to save a buck or two by switching to a Skyworks competitior. The only reason they would even consider switching (assuming Skyworks continues to perform well) is for a dramatically superior solution (not for cost…it’s already tiny). Yet given the complexity of the solution being provided, it’s not like someone can simply invent a better widget. They would have to build many better widgets, and have a superior assembly package…and they would have to be able to provide with no doubt a quarter of a billion in a year. Good luck with that.
Now with other Skyworks customers, particularly smaller use cases, this entrenched position may not exist. Yet now that they have such scale (with more dollars going to R/D), their competitive position should steadily grow. And if you’re a business trying to break into an IoT field, or wearables, or self-driving cars, I suspect there will be an element of “no one ever got fired for hiring IBM” existent now, and likely growing in the future. Again, the percentage of the bill of material for Skyworks’ solution is low, so just going with the best and most proven provider makes sense in this cutthroat industry.