Apple & Qualcomm – Gritted teeth

Apple grudgingly extends its contract.

• It turns out that cellular radio is difficult (5G particularly) as Apple has been unable to create a high-quality 5G radio of its own which has forced it to extend its contract to buy 5G modems from Qualcomm until at least 2026.
• Apple’s extension means that it will purchase modems from Qualcomm in 2024, 2025 and 2026 by which time Qualcomm is assuming that Apple will have its own modem and will be able to reduce Qualcomm’s share at Apple to 20%.
• This is a big disappointment for Apple as it has been working to create its own 5G modem since 2019 when it purchased most of Intel’s modem business but, clearly, it has not been successful.
• Apple and Qualcomm are not the best of friends and Apple is still smarting from the crushing defeat inflicted upon it by Qualcomm in 2019 when Apple’s urgent need for working 5G modems forced it to settle its IP dispute with Qualcomm over royalty rates (see here).
• This is why the new deal that it signed with Qualcomm in 2019 was only 6 years in duration as opposed to 10 years or more which is more normal practice in the industry.
• Since then, Apple has been trying to build a working 5G modem that it could then use internally, design out Qualcomm and then presumably, start the whole IP battle all over again.
• The problem is that cellular radio is difficult and in every generation that has followed 2G, Qualcomm has been first to market with high-quality working modems and it has taken everyone else at least a couple of years to catch up.
• A large part of this is because cellular radio involves the analogue domain, where getting products right is often more about one’s experience of building radios as opposed to one’s knowledge of physics.
• This is why analogue radio is often referred to as a black art and why so many companies who don’t have the requisite experience struggle to produce good working products.
• Furthermore, as the industry goes from one generation to the next, less and less radio spectrum is available which is why the 5G standard is flexible in the size of frequency blocks it can use and can also support frequencies above 28Ghz.
• This means that the radio environment has become more and more complex with every generation greatly complicating other components such as antennas, digital to analogue converters and power amplifiers.
• Qualcomm has been building all of these things for over 30 years and its abilities are evidenced by the fact that it has routinely been first to market with high-quality working products of the latest generation.
• Apple is a relative newcomer to this space and bought what I consider to be a second-rate asset from Intel in order to give itself a leg up.
• Unfortunately, this has not gone nearly as well as hoped and Apple is clearly still without a reliable 5G modem which has forced it to extend its contract with Qualcomm until at least 2026.
• This means that Apple will be unable to integrate the modem into the applications processor meaning that its devices will be slightly larger and consume slightly more power than they otherwise would.
• Fortunately, Apple’s applications processors are so good that they make up for this deficit meaning that battery life and size have not been a problem for the iPhone in any way.
• The net result is that Qualcomm’s short to medium-term revenues and profits are going to be higher than previously expected but the market is still not willing to give Qualcomm much credit for this.
• This news triggered a 3.9% rally in the shares which is pretty meagre given that Apple will now be a customer until 2026 and possibly much longer.
• The other issue is that cellular modems are constantly evolving and before long everyone will be talking about 6G meaning that to get rid of Qualcomm, Apple has to hit a moving target.
• I think that this will be very difficult, and I suspect that there is a pretty good chance that Apple will be forced to continue buying modems from Qualcomm after 2026.
• This also delays the inevitable resumption of hostilities between Apple and Qualcomm as Apple can’t risk losing access to Qualcomm’s 5G modem as a result of refusing to pay IP royalties as it did in 2017.
• When this happens it will cause great volatility in the shares, but I think that this is now very unlikely to occur before 2027 which is when the 2-year extension to its 6-year patent license (signed in 2019) will expire.
• Apple will almost certainly exercise the 2-year option in 2025 as it will still be needing to buy 5G modems from Qualcomm in 2025.
• Hence, there is now a good 3-year window with little uncertainty from Apple although there is still the lawsuit with Arm to contend with.
• Consequently, once the current inventory correction in smartphone chips has stabilised, Qualcomm will have a good base from which to build and the outlook for it in sectors outside of smartphones is good.
• Stable smartphones and expansion in other device categories represents a credible growth story that is currently being ignored.
• The shares are now trading on an 18% discount to value peers such as MediaTek and TSMC and a 75% discount to high-flyers like Nvidia.
• This is unusually good value and so I have taken a half position in the shares with the expectation of increasing should the shares weaken further.