It's no secret that Apple (AAPL +0.11%) has put a large emphasis on having some of the world's finest chip design engineers. The most recently released Apple A7 chip, which was a technical tour de force among mobile apps processors, is a testament to the quality of the work that Apple is doing with its chip teams. The next question, then, is whether it would be in Apple's best interests to own and operate a semiconductor manufacturing plant.
What kind of gross margin savings are we talking about?
The gross margin profile of the world's leading semiconductor foundry, Taiwan Semiconductor (NYSE: TSM), typically hovers between the 47%-49% range when utilization is good (and given the mobile boom, utilization has been quite good). So, let's assume that Apple currently pays on average $15 per chip from its current foundry partners. Let's be super conservative and assume that Apple gets a great deal and the gross margin percentage on this business is about 45%.
This implies that across 150 million iPhones and about 70 million iPads sold per year, Apple is handing over roughly $3.3 billion to Samsung (NASDAQOTH: SSNLF) for chips per year and Samsung collects about $1.5 billion in gross profit. This, in turn, implies that Apple would -- if it could bring the foundry in-house -- save about $1.5 billion in cost of goods sold. This would have raised gross margins by about 2.3% during fiscal 2013.
OK, that's great, but what about operating expenses?
While the gross margin savings are decent, the next question worth asking is whether the incremental operating expenses required to support this business would more than offset the gross margin savings. If Apple were to take over the manufacture of its own chips, it would need to spend the following:
- Research and Development of next-generation manufacturing technology;
- Human costs to operate the factories; and
- Capital expenditures to actually build the factories and then the depreciation of the factories that Apple builds over the lifetime of those factories (impact on net income of those purchases).
To put it in perspective, TSMC -- the world's leading foundry -- spends about $1.5 billion per year in R&D and $10 billion in capital expenditures. Of course, Apple doesn't need anywhere near the kind of capacity that TSMC needs, and TSMC's R&D isn't spent only on process for mobile chips, so Apple could probably get away with capital expenditures of about $1 billion per year and incremental R&D of $750 million-$1 billion.
Then there's the whole risk element of it all
Assuming that Apple is able to come out ahead in this equation financially (which seems rather unlikely), there's still the whole element of risk. If Apple's process development team can't get its manufacturing recipe to outperform foundry offerings, yield well enough to be economical, and be ready on time for yearly iPhone/iPad launches, then it will put its cash-cow iPhone/iPad products in serious jeopardy in order to end up with what amounts to, at best, a savings that works out to be well under $1 billion when the incremental R&D is accounted for.
To put it plainly and simply, it's not worth it.
Foolish bottom line
With Samsung/Global Foundries looking to compete head-on with Taiwan Semiconductor, and with Intel (NASDAQ: INTC) entering the foundry game, Apple will have its pick of partners to build its chips and -- in true Apple fashion -- it will probably be able to get a great deal given how important its orders are to the foundries trying to fund leading-edge process and attendant capacity. There is no need for Apple to build its own foundry, and it is unlikely that it will do so anytime soon.
