Fred Stein on This week's Apple trading strategies (7/13-7/17/26) - 'And thank you'
on This week's Apple trading strategies (7/13-7/17/26) - 'As regards predictions, there are too many black swans flapping around to know where things are going short term. But longer term, get your shades out, because things are looking pretty darned bright!'
on Silvant's Michael Sansoterra doesn't believe in Apple AI - '“can you show specific or estimated revenues and profits made from AI in those platforms?” “Mobile generative AI apps are on track to surpass $8 billion in annual consumer in-app purchase (IAP) revenue, but the actual platform profits are wildly unbalanced. While third-party developers generate billions in top-line revenue, infrastructure builders like Google face massive profit deficits, and Apple effortlessly collects billions in risk-free profit purely through ecosystem taxation.The breakdown of specific revenues and profit dynamics across these major platforms highlights starkly different financial realities: 1. Apple (The Profit King) Apple does not break out “AI platform revenue” explicitly, but it monetizes mobile AI more profitably than any other company on Earth through its App Store fee infrastructure. • Estimated Mobile AI Revenue: ~$3.0 billion to $3.5 billion annually (attributed to its 15% to 30% cut of third-party premium AI subscriptions). • Net Profits from AI: Nearly 100% profit margin. Because Apple is not running the cloud compute clusters or training the LLMs for apps like ChatGPT, its revenue from taking a platform cut is almost pure profit. • The ChatGPT Windfall: ChatGPT’s mobile app alone generated $1.35 billion in 2025. By Q2 2026, monthly mobile revenue climbed toward $275 million. Because Apple handles roughly 64% of global mobile app monetization, Apple effortlessly captures hundreds of millions of dollars from OpenAI’s mobile users annually without bearing the core computing costs. 2. Google & Android (The Revenue Engine with Capital Squeezes) Google makes massive revenues from AI across consumer subscriptions and cloud infrastructure, but its profits are heavily weighed down by immense data center capital expenditures. • Google One & Gemini Subscriptions: Google generated $1.2 billion directly from Gemini AI consumer subscriptions. Access to advanced Gemini features is bundled into Google One premium plans, which helped push total Google paid subscriptions past 350 million. • Google Cloud Enterprise AI: This is Google’s biggest AI revenue driver. Driven heavily by Gemini-powered enterprise demand, Google Cloud revenue reached $20 billion in Q1 2026 alone (a 63% increase year-over-year). Revenue from products built strictly on Google’s generative AI models grew nearly 800% year-over-year. • The Profit Reality: Despite this massive multi-billion dollar revenue scaling, tracking metrics show Alphabet (Google’s parent company) facing an estimated $227 billion deficit in pure AI profit tracking. Google spent roughly $287 billion building out massive AI data centers and TPU infrastructure while capturing about $60 billion in total cross-industry AI revenues. While highly subsidized by its legacy search advertising machine, Google’s direct consumer AI efforts are heavily focused on scale rather than short-term net profit margins. 3. Huawei (The Localized Enterprise Play) Huawei’s consumer mobile AI revenues are tightly shielded within China, but the company focuses heavily on monetizing AI as an end-to-end B2B infrastructure provider. • Estimated AI Revenue: ~$4.0 billion to $5.5 billion (heavily integrated through its Pangu Large Language Model enterprise cloud suites and local AppGallery transaction fees). • Monetization Mechanism: Because it lacks global Google Play billing systems, Huawei monetizes AI by selling “AI compute as a service” to Chinese automotive, mining, and government sectors via Huawei Cloud. For consumers, it takes localized platform revenue cuts from major third-party Chinese generative AI apps hosted in the Huawei AppGallery. • The Profit Reality: Huawei’s profit margins on mobile AI are constrained. It must reinvest a vast majority of its revenues into domestic supply chain R&D to bypass Western semiconductor equipment restrictions. Platform | Estimated AI Revenue Run | Estimated AI Profitability | Core Cost Burden Apple (iOS) | $3.0B – $3.5B (App Fees) | Near 100% Margin | Almost none (Leverages 3rd party cloud) Google / Android | $21.2B+ (Cloud + Sub) | Operating in Red (Infra) | Mass CapEx ($287B on Data Centers) Huawei | $4.0B – $5.5B (Cloud/App) | Modest / Reinvested | Heavy localized supply chain R&D Would you like to dive into: How OpenAI’s projected $14 billion net loss impacts the mobile app market? The specific ad-revenue metrics generated by free AI apps on Google Play? How much Google has managed to reduce its Gemini serving costs to improve profitability? And that’s where I stopped. Whereas Apple is generating ~$4B, it’s very profitable. For Google, $20+B revenue is a huge revenue generator against huge expenditures to run AI, profitless for now. An Amazon build out model if you will. Huawei is a hybrid model built on Chinese government subsidization and long term growth and control.'
on Jony Ive's AI project draws a lawsuit from Apple - 'Robert, I would imagine this is a nightmare scenario for OpenAI. Much of their recruiting and talent acquisition would have been on the basis of the payout when the IPO happened. That IPO is now resting on very shaky foundations – leaving aside the increased competition, the growing realisation that the money may not lie so much in the “frontier models” as much as in the “frontier ecosystem” (See Nadella and Cook and their language use around community and human team collaboration) and the fact that edge compute and AI capabilities rising (which Apple will only drop the hammer on in the next year), it leaves OpenAI in a very precarious position. Anthropic may have their challenges, but they have been booking revenue as opposed to subsidising consumption, Every partner of theirs (Microsoft included) is now investing in their own stack rather than building on theirs and practically speaking, there is zero lock in with the ChatGPT product. It’s easy for consumers to switch, especially as Agentic rises and the surfaces expand, which will further abstract away the actual model you are using. This looks like a completely unforced error by OpenAI and increasingly stellar bad judgement by Jony (who having some sense of the scale and resources required to build consumer tech may have believed OpenAI to be the best chance of acquiring them – no way was he going to get the same “freedom” from Microsoft, Google, Nvidia or others he might have contemplated as alternatives. But then, we assume that investors are getting solid analysis, knowledgeable and informed scenario modeling and technically informed takes on the prospects of these firms. Judging from the SpaceX IPO, it would appear not and that it doesn’t matter. If you get away with it, you get away with it. Someone else holds the bag.'
on Silvant's Michael Sansoterra doesn't believe in Apple AI - 'When it comes to a battle of wits concerning Apple/AAPL Sansoterra, like most analysts, are unarmed.'
on Apple dodged chip tariffs by doing a deal with Intel - 'Government should not be in the business of picking the winners.” There is a big difference between picking the winning companies, and picking the winning principles by which companies can win. The present government wants the principle “made in America “ to be the winning principle. Footnote: I wish I was on my way to the bay area for dinner'
on Jony Ive's AI project draws a lawsuit from Apple - 'With the disclosure of many of the facts involved in the lawsuit, what are the implications for Jony Ive? How does OpenAI approach its IPO with this major issue now in play?'
on Silvant's Michael Sansoterra doesn't believe in Apple AI - 'When I watched this on live TV I considered his last line in the clip shown here. He hasn’t seen the killer app for AI on Apple devices and wants to see that first. So if Apple has a popular AI incorporation in Iphone, IOS Mac etc and sales take off he will miss that move on the portion of his investment which he is underweight on Apple. Not an uncommon position for managers who have trouble being equal weight Apple stock. When I first started investing managers wanted steady income growth over quarters before they approved of investing in a particular story. This meant they missed very large moves in turnaround stories and other changes in a stocks story which didn’t have a smooth rise. By the time they said a stock was a reliable investment they likely missed huge profit gains. If Apple is a prove me story at this point in his mind then I think he hasn’t paid attention for the last 20 years. At least he has an underweight position.'
on Jony Ive's AI project draws a lawsuit from Apple - '$6 billion for io is a LOT. Jony Ive will be remembered as someone who backstabbed Apple for money because — like many of you have pointed out — he, of course, knew and was part of the process of poaching many of those 400 Apple employees. This is his fall from grace.'
on Apple dodged chip tariffs by doing a deal with Intel - 'How things change. The conservative position used to be a favorite saying of Maine’s junior senator, Angus King: “Government should not be in the business of picking the winners.” Fast forward 13 years since he became Senator: China is ascendant in no small part because of its industrial planning, and the conservative position today is to go beyond picking winners by also pressuring their customers to support them. With so many American workers cheering on “made in America”, even Trump’s detractors are forced to admit it’s good politics. For those who want to see Trump and Vance out, the question becomes in part, how do we do a better job than Trump has done of creating more manufacturing jobs? There’s tons of justified criticism of Trump, but the question remains: what’s the better plan that many Trump voters will buy? Oddly, with Trump’s chaos, the bar for answering that question is pretty low–so what is it?'
on Jony Ive's AI project draws a lawsuit from Apple - 'My hypothesis: Apple’s purpose is not an eventual payout years from now. OpenAI has been considering an IPO, seeing how successful the SpaceXAI one went. It would be the big payout they have all been hoping for. That hope may be gone. This type of lawsuit, from such a large corporation, with connections to virtually every company OpenAI would want to work with (including the financial ones who would be responsible for the IPO) , could have a huge impact on the valuation of the IPO. It makes OpenAI a bit more toxic to work with. And who would want to piss off Apple by working with OpenAI? One part of this that is telling about culture – Apple claims that OpenAI lied to one of Apple’s suppliers, telling them Apple had given permission to reveal trade secrets when it had not. In an industry where trust remains a big part of the interactions, this could well poison OpenAI’s relationship with every supplier.'
