Apple FQ2 26 – Slow but steady

Apple struggles to justify its multiple. 

• Apple reported reasonable results, but the impact of memory and compute costs creates some uncertainty around profitability even as Macs become the AI machine of choice.
• FQ4 26 revenues / EPS were $114.5bn (up 12% YoY) / $2.10, slightly ahead of expectations of $113.5bn / $2.00.
• The star of the show was undoubtedly the Mac, where revenues grew by 24% YoY to $10.8bn, handily beating the consensus estimate of $9.7bn.
• This is the one business where Apple is benefiting from the boom in AI, and it looks like this is going to continue for a while.
• Revenue guidance for FQ1 27 (calendar Q4) was positive, with 7% to 10% expected compared to consensus at 6.1%, but there are problems elsewhere.
• Gross margins in FQ1 27 are expected to be 45% to 46% compared to consensus at 47% and well below the June 2026 high of 50%.
• Apple blamed this on the cost of memory (which I suspect it has largely been passed on) and higher compute costs for the upgraded Siri digital assistant.
• Here, I suspect that Apple is buying compute capacity in the spot market given how uncertain it is about how much compute it is going to need.
• This means that initially, it is probably paying around 2x the price that it would be paying if it signed a long-term agreement with one of the usual cloud providers.
• Hence, I think that this cost pressure will reverse itself in time, when Apple has better visibility and can book capacity long-term, and the crazy spot price normalises to something more rational.
• The market did not see this one coming, which, combined with the ridiculous negative sentiment around the iPhone Duo Crackgate problem, caused a 2% sell-off in after-hours trading.
• A few of the iPhone Duos have a hinge defect which causes the screen damage observed, but Apple is replacing all of these with no fuss, and so there is really nothing to see here.
• In fact, this is not unusual for a first-generation product and without Steve Jobs getting annoyed and claiming that users are holding or folding it wrong, this will quickly be forgotten.
• This is obviously not the start that John Ternus was hoping for, but it is not a disaster either, and so I think the outlook for Apple remains more of the same.
• The AI threat to its business remains very real, and this is the key challenge that Mr Ternus has to deal with to put his mark on the history of Apple.
• There is no real sign of a fix here, and so there remains a lot of work to do.
• Fortunately, Mr Ternus has time, as users have yet to buy smartphones because of the AI services that are running on them, but that time may come.
• For the moment, Apple is benefiting as Macs have become the place to run local AI services, which is why I expect that the Mac business will continue to be strong.
• This all adds up to a business that is trading on 35.4x 2027 and 31.9x 2028 PER, which is high when I am being offered 6.3% growth.
• I can have the King of AI (Nvidia) for less than 15x FY2028, which also thinks it can deliver me 70% EPS growth next year.
• No contest, in my opinion.