A Risk-Off Tilt Beneath the Surface
The broader market closed with a defensive tone, and the divergence underneath the headline indices told the more interesting story. The S&P 500 slipped about half a percent to 7,139, the NASDAQ 100 absorbed sharper losses, and the Russell 2000 fell 1.2% as riskier corners of the market gave back ground. The Dow Jones, by contrast, lost only around 47 points, cushioned by solid earnings from blue-chip names. That asymmetry between the indices points to a rotation rather than a broad sell-off — money flowing out of the most aggressive growth pockets and into the steadier ones.
The trigger came from a Wall Street Journal report raising concerns about OpenAI's compute demand and spending. The market punished anything tethered to that narrative. AMD and Nvidia traded lower, and Oracle slid 4% even after issuing a public response to the report. Memory-linked names came along for the ride. Technology as a sector finished down 1.7%, leading losses. Industrials, which have effectively become a de facto AI trade through their data-center and infrastructure exposure, also got hit. Materials declined 0.7%, while consumer discretionary slipped 0.7% as well.
Defensive sectors absorbed the rotated capital. Consumer staples climbed, utilities and real estate finished higher, and financials nudged up fractionally. Energy was the standout, gaining 1.7% on triple-digit price gains in underlying commodities.
T-Mobile: Solid Numbers Eclipsed by Deal Speculation
T-Mobile's first quarter print was clean across the board, but the stock barely moved — a telling sign that the operating performance has become secondary to the M&A narrative swirling around the company.
The headline numbers beat on every line. Earnings came in at $2.27 per share against a $2.26 expectation. Revenue reached $23.11 billion versus a $22.98 billion estimate. EBITDA arrived at $9.24 billion compared to the $9.05 billion consensus. Capital expenditures landed at $2.62 billion, modestly above the $2.49 billion expectation, with full-year capex guidance of roughly $10 billion sitting right on top of the $10.01 billion estimate. Full-year core adjusted EBITDA was guided to $37.1 to $37.5 billion, with the midpoint matching the $37.36 billion consensus almost exactly. Free cash flow guidance of $18.1 to $18.7 billion was a reiteration.
The most important operating metric for any wireless carrier — postpaid net account additions — came in at 217,000 for the quarter, up 6% year-over-year and inclusive of both phone and broadband. Average monthly postpaid service revenue per account grew 3.9% year-over-year. Coming on top of similarly strong subscriber numbers from Verizon, the data confirms that demand for premium wireless plans and bundled streaming offerings remains durable.
Yet the stock seesawed around unchanged, drifting fractionally lower in the immediate aftermath. The reason is news flow. Rumors over the prior 48 hours pointed to a potential buyout involving the German telco parent, which would constitute a massive cross-border merger if it materialized. The company also announced a joint venture earlier in the day, and the elevated capex figure appears to reflect that investment. Markets effectively want to hear the conference call before committing capital — there is too much pending information for the print itself to drive the tape.
The technical picture remains constructive. The $180 level has historically served as a buying zone for institutional investors, and the stock is trading reasonably close to it. The analyst community had already turned more constructive heading into the print: Moffett Nathanson upgraded to buy from neutral, citing a reliable growth story anchored in rural America, and Key Bank moved to overweight from sector-weight with a $260 price target.
Seagate: A Real Read on the AI Buildout
If T-Mobile's print was a study in how news flow can drown out fundamentals, Seagate's was the opposite — a fundamental beat so clean it forced the stock through the upper bound of what the options market had priced in.
Fiscal third-quarter EPS came in at $4.10 against a $3.50 expectation. Revenue of $3.11 billion topped the $2.95 billion consensus. Adjusted operating margin of 37.5% was nearly three full points above the 34.8% estimate. The forward guide was even more striking: fourth-quarter EPS of $4.80 to $5.20 against a $3.97 consensus — potentially more than a full dollar ahead of the street. Revenue guidance of $3.45 billion handily eclipsed the $3.16 billion expectation. The stock surged more than 10%, exceeding the roughly 9.5% directional move the options market had implied.
The most striking fundamental data point is the year-over-year comparison on margins, which expanded by roughly 11 points. That kind of margin lift is unusual for a business with this much historical cyclicality. The reason is pricing power. Seagate is operating inside a secular trend where it can pass higher prices through to customers without resistance. The print is, in effect, one of the first real-time reads on whether the AI infrastructure buildout has genuine legs — and the answer, at least in the storage layer, is yes.
What makes the guidance particularly credible is what management chose to acknowledge in it. The company explicitly factored tariff impacts, Middle East conflict risk, and logistics headwinds into its outlook — and still delivered a guide a dollar above consensus. That is not a company papering over weakness; it is a company that has visibility into long-term partnerships and is comfortable forecasting through identified risks.
The Memory Cycle's Structural Story
Seagate's results sit inside a broader memory and storage narrative that has been gathering force across the supply chain. The relevant exposure here is hard disk drives and DRAM, not GPUs — a distinction worth holding onto when assessing AI-related earnings. Western Digital, the closest pure-play storage counterpart, has delivered a similar message. SK Hynix has posted incredible numbers. Samsung is expected to report record results. Micron has confirmed the same dynamic on the DRAM side.
The connective thread, repeatedly emphasized by analysts, is that names across the memory supply chain — DRAM, NAND, and storage — tend to move in sync. They share long-term partnerships with hyperscale customers, they face high and persistent demand, supply remains tight, and they retain pricing power. Bank of America noted earlier in the week that producers are deliberately not adding unit capacity, which is precisely how they can continue raising prices. Sell-side price targets have been getting boosted nearly daily on the thesis that this is a structural shift, not a cyclical bounce.
The after-hours reaction across the complex reflected that read. Micron rose roughly $5 on the back of Seagate's print. SanDisk gained about $20. These are the names retail and equity traders are most likely to pile into at the next open, and they may meaningfully cushion the day's losses in the NASDAQ and S&P 500.
Levels and What Comes Next
For Seagate specifically, a clean technical breakout requires a sustained move above the $640 area, where the call wall sits. The stock is trading at all-time highs and recovered convincingly from a 2.8% pullback earlier in the session. Whether Seagate offers anything truly differentiated at the product level is debatable — the answer is probably not — but the read-through to Micron and the broader memory complex is what equity flows are responding to.
The day's earnings, taken together, paint a market in transition. The original AI trade — concentrated in marquee chip names and the OpenAI orbit — is being questioned on spending discipline. Meanwhile, a quieter, more tangible expression of the same buildout is showing up in storage and memory, where pricing power is real, capacity is constrained, and visibility into demand is unusually long. Investors looking for evidence that AI capital expenditure converts into actual earnings are finding it — just not always in the names that have driven the narrative.