Is Apple Losing Its Memory?

Tyler Crowe calls Apple’s results under new CEO John Ternus potentially more interesting than recent “same old” earnings. Matt Frankel says the notable surprise was memory: even Apple, the supply chain’s most powerful buyer, cannot obtain enough memory at reasonable prices, which Ternus called “unprecedented.” AI data centers are outbidding Apple for components, reinforcing how intense and supply-constrained the AI buildout has become. Lou Whiteman adds that Apple’s gross margin fell to 45% and says the company increasingly resembles Berkshire Hathaway: enormous, mature, and difficult to move.

The hosts ask whether reports of Apple’s new folding phones breaking represent a serious issue or a media-fueled headline. Matt says it is a real but anticipated, industry-wide problem: Samsung has made folding phones since 2019, and his neighbor’s Galaxy Flip has broken twice in a year. Apple says only a small number of units are affected, is replacing them through AppleCare, and still has demand ahead of supply. Lou says the problem is embarrassing for hardware veteran Ternus and weakens Apple’s old “it just works” halo, but remains fixable and not investment-moving.

Apple continues paying dividends, buying back stock, paying down debt, and generating excess free cash flow rather than spending like the hyperscalers on data centers. Lou says its share count has been cut in half over the past decade, but suggests a bold move: acquire Perplexity to give Siri a model-agnostic AI system rather than tying it to Google or another provider. Matt prefers supply-chain investment, arguing Apple could follow NVIDIA’s $2 billion investment in Lumentum by prepaying for memory or investing directly in a memory maker.

Lou says Palantir reported a strong quarter but deserved the market’s negative reaction because its valuation—about 110 times forward earnings—requires extraordinary growth. Commercial bookings grew only 11% quarter over quarter after a prior 153% jump, the lowest bookings gain since 2023, and Palantir added just 33 net new customers. Lou says government growth remains respectable but cannot justify the valuation because even a roughly $1 trillion Pentagon budget includes about $700 billion for personnel and healthcare; Palantir needs “crazy” commercial growth, not merely excellent growth.

Matt says Shopify’s report was also strong in isolation: revenue rose 32% year over year, merchant-solutions revenue 35%, operating income 66%, and free-cash-flow margin reached 22%, up four percentage points. It also repurchased more than $500 million of stock. However, gross merchandise volume and subscription revenue each missed expectations by roughly 1%, and the stock initially fell about 15% before recovering part of the decline. Shopify is still trying to place checkout and payment infrastructure across AI shopping agents, but management’s agentic-commerce strategy remains a work in progress.

Lou interprets reports that SpaceX may acquire or partner with Boost Mobile as revenue hunting rather than an attempt to become the country’s fourth major telecom operator. Starlink can provide adequate-to-mediocre cellular service, but needs a terrestrial component to compete effectively. By accessing Boost’s spectrum and other assets instead of building everything itself, SpaceX could tap part of the roughly $350 billion annual revenue generated by the major carriers and fund projects such as Mars missions or data centers. The arrangement remains only a rumor.

CVR Energy, controlled by Carl Icahn, agreed to buy two Marathon Petroleum refineries in Michigan and North Dakota for about $4 billion—nearly the size of CVR’s roughly $5 billion market value. Matt says refining margins are about two and a half times normal levels and could fall rapidly if the Iran war ends or idled capacity returns, making Marathon’s sale look well timed. Tyler notes the facilities represent less than 5% of Marathon’s refining capacity and wonders why Icahn is consolidating at expensive valuations; regulatory approval may also be difficult.