On Wednesday in Cupertino, Apple introduced the iPhone 18 Pro, the iPhone 18 Pro Max, AirPods 5, two new watches and, for the first time, a folding iPhone called the Duo. It was the first major appearance for John Ternus, who succeeded Tim Cook as chief executive on 1 September. The stock fell as much as 1.9 percent during the presentation and closed at 316.67 dollars, up 0.4 percent on the day. That is the usual choreography of an Apple event: a great deal of attention and very little price movement.
The number that matters was not on stage. The iPhone 18 Pro costs 1,199 dollars, one hundred more than its predecessor. The Pro Max costs 1,299 dollars, also one hundred more. A nine percent increase, the headline says. The headline is accurate and still misleading, because it describes only one side of the ledger. On the other side sits an estimate that the research firm TrendForce published on 10 August: the components of a 256-gigabyte iPhone 18 Pro cost Apple roughly 38 percent more than those of the iPhone 17 Pro. And the reason is not the new chip, not the new camera with its mechanical aperture, not the display. The reason is memory.
Put the two numbers side by side and what actually happened on Wednesday becomes visible: Apple raised a price that does not cover its own cost increase. The hundred dollars are not a margin expansion. They are the share of the memory bill the customer pays. Apple pays the rest.
What one hundred dollars actually buys
The arithmetic can be reconstructed from public figures. Teardown analyses based on TechInsights data put the component cost of the iPhone 17 Pro Max at roughly 408 dollars. That is the bill of materials, the bare sum of parts before assembly, logistics, warranty, distribution and software. Take those 408 dollars as the base and apply TrendForce’s 38 percent, and the component bill of the successor lands near 563 dollars. The cost increase is therefore about 155 dollars per device.
The price increase is one hundred dollars. It covers roughly two thirds of the cost increase. The remaining third, roughly 55 dollars per Pro unit, comes out of margin. Set component cost against list price and the iPhone 17 Pro shows a notional component margin of about 63 percent, the iPhone 18 Pro about 53 percent. That is not a gross margin in the accounting sense, because too many costs are missing, but the direction is unambiguous: at the level of its most important product, Apple has given up about ten percentage points of raw margin, and the price increase has recovered only part of it.
One can look at it from the other side. Had Apple passed the full cost increase through, the iPhone 18 Pro would sit near 1,255 dollars. Had Apple wanted to defend the old 63 percent component margin, it would sit near 1,520 dollars. The decision to price at 1,199 dollars is therefore a decision against margin and in favor of volume. TrendForce predicted exactly that: Apple would follow the pattern of its June MacBook price increases and sacrifice part of its gross margin to protect shipments.
Memory has overtaken the processor
To understand why a company with a gross margin above 50 percent gives up margin voluntarily, one has to look at the structure of the cost increase. According to TrendForce, memory, meaning DRAM and NAND flash, made up roughly ten percent of the component cost of a 256-gigabyte Pro model a year ago. In the third quarter of 2026 it is roughly 34 percent. In the first half of 2027 it is expected to exceed 40 percent. Memory prices have risen five- to sevenfold since the beginning of 2025. Memory has thereby become the single most expensive part in the phone, more expensive than the A20 Pro processor and more expensive than the OLED display.
That allows a second calculation, more revealing than the first. Ten percent of 408 dollars is about 41 dollars of memory in the old device. Thirty-four percent of 563 dollars is about 191 dollars of memory in the new one. Memory alone has become roughly 150 dollars more expensive. The total cost increase is 155 dollars. In other words: every other component combined, the new two-nanometer processor from TSMC, the variable-aperture camera, the larger modem, the smaller Dynamic Island, cost Apple about five dollars more. Almost the entire increase in the cost of the iPhone 18 Pro is an increase in the cost of memory chips that have not changed in a year. Only their price has changed.
That is the core of the story, and it has little to do with Apple. Memory became expensive because the data centers of the AI providers are buying it up. Roughly 70 percent of all memory chips now go into servers. The manufacturers, Samsung, SK hynix and Micron, are shifting their capacity to where the margin is highest, which is high-bandwidth memory for AI accelerators. What is left over is auctioned among smartphone, PC and console makers. Apple is the largest single customer in that market and presumably gets the best terms. If even Apple pays 38 percent more, everyone else pays more.
The ledger in Cupertino
Apple itself has been unusually candid about this. On his final earnings call as chief executive on 30 July, Tim Cook spoke of a “hundred-year flood” in memory pricing, with exponential increases, and added that he had never used that phrase in more than forty years in the industry. Memory, Cook said, went up in March, went up in June, goes up again in September and keeps rising after that. Chief financial officer Kevan Parekh added that memory costs explained more than 100 percent of the sequential decline in gross margin.
The June quarter shows what that means. Revenue rose 16 percent to 109.4 billion dollars, iPhone revenue rose 22 percent to 54.3 billion, Mac revenue 29 percent, Services 12 percent to 30.7 billion. Gross margin was 50.1 percent but included roughly two percentage points from a tariff refund. Without that one-off, margin would have fallen quarter over quarter. Products gross margin was 40.1 percent, Services 75.6 percent. For the current September quarter Apple guided to a gross margin between 47 and 48 percent on revenue growth of nine to eleven percent. That is two to three percentage points less margin than June, and the September quarter contains only two weeks of iPhone 18.
What is interesting is how differently Apple has distributed its price increases. In June the MacBook Air went up 200 dollars, the Mac Studio with the M3 Ultra chip 1,300 dollars, iPads 100 dollars; across the affected products the average increase was around 23 percent. The iPhone was spared then. Now it gets nine percent. The difference is not an accident. The iPhone accounts for roughly half of company revenue and for most of the on-ramp to Services, whose margin is 75 percent. A Mac buyer who waits a year because of a 200-dollar increase costs Apple one Mac. An iPhone buyer who switches to Android because of a 200-dollar increase costs Apple an iPhone, a subscription, an App Store cut and a family member in the ecosystem. Apple protects margin where the customer has the fewest alternatives, and protects volume where the customer is most valuable.
There is also a shift that makes the arithmetic more bearable. The share of Pro models in iPhone shipments has risen, according to industry estimates, from 37 percent in 2020 to 65 percent in 2025. Two of every three iPhones sold are now Pro devices. That means the hundred-dollar increase applies to most of the volume, and it means Apple is addressing its price increase to precisely the customers who have voluntarily migrated upward over the past five years.
The folding phone as a price anchor
The iPhone Duo is a special case in this calculation. It starts at 1,999 dollars for 256 gigabytes and rises to 3,199 dollars for two terabytes, opens like a book onto a 7.6-inch display, carries a 5.4-inch screen on the outside, has two batteries and arrives on 23 October in more than 70 markets. Samsung, seven generations into this market, lists its Galaxy Z Fold8 with 256 gigabytes from 1,649.99 dollars and the Fold8 Ultra from 1,799.99 dollars. Apple has positioned itself 200 to 350 dollars above the established competitor, with a first-generation product.
Whether that works is a question of volume, and the volumes are small. IDC estimates that foldables will account for roughly 2.2 percent of all smartphone units in 2026 but 6.9 percent of revenue; by 2030 the forecast is 3.1 percent of units and about ten percent of value. For Apple, IDC projects a 46 percent share of the global foldable market by 2030, around 16.3 million devices, and 58 percent of the value. In China, the largest foldable market, Huawei held 79.4 percent of units in the second quarter. Ben Bajarin of Creative Strategies expects “very high” demand and a 20 to 25 percent market share for Apple; one analyst doubled his estimate of the Duo’s share of fiscal 2027 iPhone revenue from five to ten percent. AT&T chief executive John Stankey, by contrast, said drily that “this is not new” and expects a niche among premium buyers.
For the margin question, the Duo matters less as a volume product and more as a price anchor. A 1,999-dollar iPhone makes a 1,199-dollar iPhone look cheaper. And the Duo’s storage ladder shows where Apple will find its margin in an era of expensive chips: the step from 256 gigabytes to two terabytes costs 1,200 dollars. Even at today’s memory prices, the premium sits far above the component cost. The memory shortage makes the base model more expensive, but it does not make the upgrade tiers any less lucrative. Whoever buys more storage from Apple subsidizes the storage everyone else buys.
Who actually pays the bill
The other side of this story sits in Boise, Suwon and Icheon. Micron closed on Tuesday just above 1,000 dollars a share, after failing to hold an intraday high above 1,041. The memory makers earn from the same shortage that squeezes Apple’s margin, and in the same order of magnitude: what appears at Apple as 150 dollars of memory per device on the cost line appears at Samsung, SK hynix and Micron on the revenue line. With 247.4 million iPhones shipped in 2025, a record according to IDC, and a Pro share of 65 percent, the Pro models alone represent something on the order of 20 to 25 billion dollars of additional annual memory spending if prices hold at today’s level.
TrendForce expects global smartphone production to remain under pressure from the second half of 2026 through 2027, and expects Android makers to face an even tighter margin squeeze than Apple. That is the real competitive shift. Apple can swallow 55 dollars per device because it has a 40 percent products margin. A manufacturer with an eight percent margin cannot. It must either pass through the full amount, which on a 400-dollar phone means a one-third price increase, or cut the memory, which makes the device worse, or cut volume. TrendForce names the entry-level and mid-range segments as the most vulnerable. In a memory shortage, margin itself is a competitive advantage, because it is the buffer that makes the price increase smaller for your customer than for your competitor’s.
The stocks in play for US investors
For American investors the chain is unusually direct, because both ends of it trade in New York. Micron is the only US memory maker and the purest expression of the story; at roughly 1,000 dollars a share it reports its fiscal fourth quarter on 30 September, and the contract-price outlook in that report will tell you more about Apple’s next margin than anything Apple says. SanDisk and Western Digital carry the NAND side; both fell sharply in the July memory sell-off and both trade on the same contract-price curve. Qualcomm supplies the modem in the standard iPhone models and is on the losing side of Apple’s in-house C-series modem transition, a story that has nothing to do with memory but illustrates what the five dollars of non-memory cost increase implies: there was very little room left in this year’s bill of materials for anyone else’s price increase.
Downstream sit the carriers and the retailers. Verizon, AT&T and T-Mobile finance most iPhones sold in the United States over 24 or 36 months; a 100-dollar increase shows up either as a longer contract or a higher monthly installment, and both are a test of willingness to pay that will appear in fourth-quarter net additions. Stankey’s comment that foldables are “not new” is worth reading in that light: a carrier chief executive has no interest in talking up a 1,999-dollar handset he will have to subsidize. Best Buy, which sells iPhones at essentially fixed margins, gains revenue per ticket from higher prices but not necessarily profit if traffic thins.
On tax, the usual rules apply. Gains on Apple or Micron held for more than a year are taxed at long-term capital gains rates of 0, 15 or 20 percent, plus the 3.8 percent net investment income tax above the relevant thresholds; gains on positions held under a year are ordinary income. Apple’s dividend of 27 cents a quarter is qualified and taxed at the same preferential rates, though at roughly 0.3 percent yield it is close to irrelevant. Anyone who sold Micron in the July sell-off and wants back in should keep the 30-day wash-sale window in mind before harvesting the loss.
The counterarguments
First: component cost is not cost of goods sold. A bill of materials of 408 or 563 dollars leaves out assembly, testing, packaging, freight, tariffs, warranty and software development. The 40.1 percent products margin Apple reports is the number that counts, and it sits far below the 63 or 53 percent of the bare component arithmetic. The ten points of raw margin lost at the component level translate into fewer points at the company level, because Services at a 75 percent margin make up an ever larger share of revenue. Apple’s guidance of 47 to 48 percent for the September quarter is exactly that: a company margin that feels the memory shock but does not reproduce it.
Second: at Apple, price has rarely been the binding constraint. JPMorgan points out that the relationship between iPhone prices and unit volumes has historically been weak. When the iPhone X raised the price 54 percent to 999 dollars in 2017, there was no volume collapse but a migration into the Pro class that now stands at 65 percent. And Apple has explicitly warned of supply constraints on advanced chip nodes for the current quarter. When demand exceeds supply, price is not the problem but the dial, and a higher price is then a rational response to scarcity rather than a risk to volume.
Third: the memory cycle turns. Five- to sevenfold prices in eighteen months are not a new normal but an extreme, and extremes in the memory industry have resolved into overcapacity every single time. Samsung, SK hynix and Micron are expanding, and the memory shortage of 2026 is preparing the memory glut of 2028. Anyone writing off Apple’s margin today because memory is 34 percent of components must also think through the reversal: if memory prices fall and Apple keeps the 1,199 dollars, the margin comes back on the way down. The price increase is permanent; the cost increase is cyclical.
Fourth, and this concerns the stock: since the first iPhone in 2007, Apple shares have risen in the sixty days after every launch event, seventeen times out of seventeen. Wamsi Mohan of Bank of America, who holds a 380-dollar price target, describes the pattern as a sell-off into the event and a recovery within 30 to 60 days. Jefferies, with a 263.66-dollar target and an Underperform rating, takes the other side and sees 16 percent downside. Both positions are compatible with the same numbers, because they are talking about different things: Mohan about the cycle, Jefferies about the valuation.
Outlook: the number that matters in October
Sales of the iPhone 18 Pro begin on 18 September, pre-orders on the 12th. The Duo follows on 23 October. In late October Apple will report the September quarter, John Ternus’s first as chief executive, and the number to read in that report is not iPhone revenue. It is products gross margin. It was 40.1 percent in June. If it falls below 38 percent in September, memory has won and the hundred dollars were not enough. If it holds above 39 percent, Apple has, for now, channeled the hundred-year flood with its mix of price increase, storage ladder and Pro share.
The second number does not come from Apple but from Micron on 30 September, when the memory maker reports its quarter and gives its outlook for contract prices. Cook said memory goes up again in September. TrendForce expects memory’s share of components to exceed 40 percent in the first half of 2027. Both mean that August’s 38 percent is not the end. The iPhone 18 Pro is priced on a memory bill that will be out of date by next spring, and then Apple will have to decide whether to give up margin again or whether Wednesday’s price round was only the first.
What went unsaid on Wednesday is therefore the actual news. Apple did not introduce a more expensive iPhone. Apple introduced an iPhone whose components became 155 dollars more expensive, and passed one hundred of those dollars to the customer. The remaining 55 dollars appear in no press release. They appear in the margin guidance for the September quarter, two to three percentage points below June, and they will stay there until either the price of memory falls or the price of the iPhone rises again.
Try TradingView Free for 30 Days
Plus get a $15 discount on your first subscription through this link.
