ASE Technology Locks in 30% Gross Margin as New Baseline, 2026 Capex Raised to $11.5 Billion
nashnova research
ASE Technology has set 30% as the new gross-margin floor for its packaging-and-testing business, with quarterly improvement expected through 2027; 2026 capex has been raised for the third time to $11.5 billion, as AI-driven advanced-packaging demand reshapes the world's largest OSAT's profit structure.
What keeps the 30% margin floor intact?
Q3 assembly-and-test (ATM) gross margin hit 30.6%, driven by a rising share of LEAP — ASE's advanced-packaging brand — and near-full capacity utilization.
This means → the profit engine is shifting: margins used to depend on "good cycle + high utilization"; now they depend on product-mix upgrades — the more high-end work in the mix, the more resilient the margin.
Both advanced and mainstream packaging capacity remain tight, keeping pricing power on ASE's side; management expects ATM margins to rise quarter by quarter through 2027.
Why does LEAP's revenue target keep going up?
The 2026 LEAP revenue target was raised from $3.7 billion to over $4 billion; the 2027 target is to double that again.
Within LEAP, full-turnkey packaging — including CoWoS, a high-end process that stacks chips like a sandwich — is targeted at $300 million in 2026 and expected to top $1 billion in 2027.
This reflects AI-chip demand for advanced packaging growing faster than ASE itself forecast at the start of the year — three target raises in one year signal steadily improving order visibility.
What does CPO mass production mean?
Co-packaged optics — embedding optical communication modules right next to the chip for faster data transfer and lower power — entered mass production in Q4 2026, with 2027 revenue expected to exceed $500 million.
The market worries that CPO and near-packaged optics (NPO, placing optical modules near but not directly on the chip) will cannibalize each other. ASE's view: both architectures will coexist long-term, serving different performance needs.
In plain terms = ASE is betting on both tracks — not on which one wins, but on the idea that no matter which wins, the packaging step still runs through ASE.
Where does $11.5 billion in capex go?
ASE raised 2026 capex for the third time this year, adding $1 billion to bring the total to $11.5 billion: $500 million for buildings, $500 million for equipment; LEAP takes 60% of the new equipment spend, mainstream tech 40%.
22 building projects are under way simultaneously in Taiwan; more than half are expected to start contributing revenue before 2027. ASE is also looking to acquire existing facilities to speed expansion.
This means → the bottleneck is no longer "whether to buy equipment" — it is whether buildings go up fast enough. ASE's expansion pace is effectively gated by construction timelines.
What do Q4 guidance and the risks look like?
Q4 ATM revenue is guided at 11%-13% quarter-on-quarter growth, with gross margin up another 1-2 percentage points; consolidated revenue is guided at 6%-8% QoQ growth, gross margin 23%-24%.
EMS is the drag: AI-server component shortages pushed up pass-through costs, pulling Q3 EMS gross margin down to 8.3% and Q4 operating margin to roughly 2%.
Mainstream packaging-and-testing is expected to grow over 20% in 2026, with auto and industrial recovery plus AI-chip demand offsetting soft consumer electronics; whether the 30% margin floor holds quarter after quarter in 2027 will be the market's key test of ASE's profit-structure transformation.
市场有风险，内容仅供研究参考，不构成投资建议。