AI Laggards Q3 25 – Bounce but no Recovery

Still not much AI going on.

Amazon Q3 25 – Still behind.

• Amazon beat low expectations but, in my opinion, remains behind in the AI race, which leaves AWS at risk of losing market leadership to Azure.
• Q3 25 revenues / EPS were $180.2bn / $1.95. well ahead of estimates of $178bn / $1.57 where the high-margin AWS business made most of the difference.
• AWS posted Q3 25 revenues of $33bn, up 20% YoY, marking a reacceleration from the 17% YoY it posted in Q2 2025, which is badly needed.
• However, I don’t think it is anything like enough, as Azure is growing at 39% YoY and GCP at 33% YoY, having seen a strong acceleration as demand for AI compute has taken off.
• AWS remains larger than these two, which makes fast growth more difficult, but nowhere are there mentions of capacity constraints or compute shortages that would imply that customers are beating down the door like they are at Azure and GCP.
• Amazon’s excuse is that it had already anticipated this, but if this were the case, the acceleration should have been greater.
• The signs of this have been apparent for some time now as Amazon remains a laggard in AI, and this is damaging AWS’s brand as a place to come and train and run AI workloads.
• Azure is growing twice as fast as AWS, and it is not inconceivable that, unless Amazon really pulls its finger out, market leadership is surrendered to Microsoft in 2027.
• Elsewhere, the news was better, as profitability in the core e-commerce business remains more stable than it has ever been, and guidance for Q4 2025 is comfortably in the range.
• Here, Q4 25 revenues are expected to be $206bn – $213bn ($209.5bn) just ahead of consensus of $208.5bn with operating profit of $21bn – $26bn ($23.5bn), which was also in line with forecasts.
• The shares have jumped 13% in after-hours trading as the view is that AWS is finally catching the AI wave, but I am yet to be convinced.
• Hence, I think it will continue to struggle, and as such, I have no inclination to own it.

Apple FQ4 25 – Short-term bounce

• A reasonable set of results were held together by the services business and the promise of a revenue bounce next quarter as iPhone replacement demand is more robust than it has been for several years.
• FQ4 revenues / EPS were $102.5bn / $1.85, just ahead of expectations of $102.2bn / $1.78 as Services continued to outpace expectations and made up for weakness in China.
• Apple put the weakness in China down to supply chain issues and stated that demand for the iPhone 17 both inside China and elsewhere was stronger than it had seen for some time.
• This has been verified by Counterpoint Research, whose research uncovered a fast start for the iPhone 17 in September and 29% YoY growth in China in the first two weeks of October.
• This is all down to the hardware refresh, where the iPhone 17 looks meaningfully different from previous versions, even if the performance and the user experience are pretty much the same.
• This is enough to trigger a mini-upgrade cycle as consumers get the feeling that they are getting something new rather than an incremental update to last year’s model.
• This is why Apple is guiding strongly into FQ1 26 with 10% – 12% YoY revenue growth powered by the iPhone, where revenues are expected to grow by double digits YoY.
• However, this is not a return to growth in the market generally but a short-term blip as the fan base buys a new phone sooner than they otherwise would have done.
• This means that Apple’s growth problems remain, as does the black hole that is its AI strategy.
• Apple has time here, as AI is not even close to the point at which it affects the purchase decision of a smartphone, but that time may come.
• This is why Apple must fix this problem, and I suspect that acquisition is the only way that it will be able to do that.
• Should the AI magic money tree stop showering money, then most of the model builders will be in serious trouble, and Apple will be able to pick one and make it its own for a much more reasonable price than it can today.
• In the meantime, I see a company struggling to grow that trades on a valuation much higher than Google, TSMC, MediaTek, Samsung or Qualcomm and I would own any of these over Apple any day of the week.
• I continue to hold both Samsung and Qualcomm and remain very comfortable with both.