They tripled capacity and quadrupled 2.5D revenue, yet gross margin sits at 14.6% with a guide down.
Thesis: Amkor is doing the work, but TSMC is getting paid. The 2.5D packaging revenue is real (quadrupling y/y), but it's not enough to lift the corporate average margin against a 60% utilization drag in legacy nodes. They are a CapEx servitor to the hyperscalers, taking on the depreciation risk while the foundry controls the pricing bottleneck. Until mainstream utilization crosses 80%, this is dead money.
Verdict: AVOID — Conviction: MEDIUM
Catalyst: Q1 2025 guidance confirming if the 'temporary' iOS headwind actually clears or if it was structural share loss.
Key Risk: Mainstream utilization stays at 60% through 2025, turning the new Vietnam capacity into a margin anchor rather than a growth engine.
The Tell: When asked about margin drivers, CFO Megan Faust admitted Vietnam start-up costs are hitting 100bps in Q4. They are expanding capacity into a downturn in their core business, counting on an AI ramp that hasn't yet proven it can carry the P&L.
Friction Level: HIGH_FRICTION — Bulls see a 2.5D growth engine masked by temporary phone seasonality. Bears see a structural lack of pricing power where TSMC captures the margin and Amkor takes the CapEx risk.
Report not found
The report data is no longer available. Please return to the archive.