Sanmina Ships Helios Into a Sell-the-News Tape
Sanmina Ships Helios Into a Sell-the-News Tape
A blowout fiscal 2026 and an $18.5 billion midpoint guide are real. The stock's 45% drawdown says the AI infrastructure trade is no longer paying for promises.
Sanmina's fiscal fourth quarter was not subtle. Revenue rose 84% to $3.86 billion, non-GAAP operating margin hit 9.0%, and non-GAAP EPS jumped 166% to $4.43. For the full year, “Revenue grew 79% year-over-year to $14.53 billion.” — Jure Sola, Chief Executive Officer · 2026-11-09 The guide was bigger still: fiscal 2027 revenue of $17.5 billion to $19.5 billion, with $7.5 billion to $9.0 billion from the newly named cloud and AI systems bucket. Management even began shipping the Helios program in small quantities.
Yet the equity sits 45% below its June peak and is down 24.5% over the last 90 trading days. That is the report's central tension: the income statement is accelerating, while the market is de-rating the stock.
The Beat Is Real. The Multiple Is Not.
Sanmina is not a broken story. Q4 core Sanmina revenue was $2.97 billion, up 42%, and ZT Systems contributed $884 million, in line with the outlook. Book-to-bill was 1.22:1, and management sees “strong demand for AI hardware in the second half of fiscal '27” — Jure Sola, Chief Executive Officer · 2026-11-09. The core industrial, energy, medical, defense and automotive businesses grew 19% to $1.49 billion in the quarter.
But the margin math is less clean. Q4's 9.0% operating margin included about $30 million of tariff recoveries, worth roughly 75 basis points and $0.42 of EPS, none of which was in the outlook. The fiscal 2027 guide asks for 7.0% to 7.5% operating margin, below the Q4 print, which management attributes to mix and the ramp of new programs. Through the June quarter, operating margin was 6.4%, and capital expenditures had reached $101 million. Q4 was a step change; fiscal 2027 is a test of whether that step change holds without one-time help.
The New Thing Is Not ZT. It Is the Platform Portfolio.
The top keyword remains ZT Systems, but the more important change is structural. Sanmina is now reporting a cloud and AI systems category that includes ZT plus additional platforms and related products. That is a new lens on the business, and it matters because it admits the old core-versus-ZT split no longer describes how the company wins work. The July call was still a transition story: management said “we expect revenue to ramp over time throughout 2027” — Jonathan Faust, CFO or Finance Executive · 2026-07-27. Three months later, the tone is operational: “we've now begun to ship the Helios product, albeit in small quantities” — Jonathan Faust, Executive Vice President and Chief Financial Officer · 2026-11-09.
some of that could happen in the ZT plants, but some will be done in core Sanmina plants. So the lines are starting to blur
That blurring is the real pivot. The second platform is now included in fiscal 2027 guidance and ramps through the year. A third platform is in discussions but excluded from the guide, leaving upside optionality rather than a promise. The company is also leaning harder on vertical integration, selling mechanical racks, printed circuit boards and other subassemblies into non-Helios customers. None of this is a repeat of the ZT acquisition narrative; it is the first evidence that the acquired capability is becoming a portfolio.
Cash, Consignment and the Cost of Scaling
Growth is consuming working capital. Q4 inventory was $2.6 billion net of customer advances, up 131% year-over-year, and inventory turns fell to 5.2x from 6.7x. Fiscal 2027 capex is guided to $450 million. Net leverage was only 0.33x at quarter-end, below the 1.0x to 2.0x target range, but management expects it to rise as working capital builds. That was not a surprise: in July, Jon Faust said “we do expect working capital to build” — Jonathan Faust, CFO or Finance Executive · 2026-07-27.
The funding model helps, but it also complicates the revenue optics. Management still assumes a net accounting basis with a consignment model, as it did in July, when Faust said “the vast majority of that business will be consignment” — Jonathan Faust, Executive Vice President and Chief Financial Officer · 2026-04-27. The balance sheet has room: $3.83 billion of total liquidity and a refinanced Term Loan A. But the June filing already showed effective net cash at negative $285 million, a swing from positive $666 million a year earlier. The market may be discounting cash conversion, not demand.
Why the Tape Is Right to Be Suspicious
The global tape is already voting against AI infrastructure. Over 90 days, AI data centers is among the market's biggest decliner themes, alongside high performance computing and data center capacity. Sanmina's AI system revenue sits directly in that wake. The stock's drawdown from 282.72 in June to the recent 229.04 peak-then-drop is a valuation event, not an order event.
But Sanmina is not a pure-play. The core businesses are growing, and management expects low double-digit growth there. The bear case is that cloud and AI systems carry a lower margin and heavier working capital than the legacy mix, and that the third platform is not in the number. The bull case is that an $18.5 billion midpoint with 7.25% operating margin and $16.50 EPS is still a lot of earnings power for an $8 billion market cap. The next 90 days will settle it: Q1 cloud and AI systems revenue of $1.3 billion to $1.5 billion, Helios shipment cadence, second-platform ramp, capex and inventory. If those land, the gap between operating momentum and the tape closes. If margin slips below 7%, the market will treat the guide as a peak rather than a base.