IC Packaging & Test (OSAT)

Let me walk you through my outlook for the next two quarters. Before I do, let me lay out the key pillars supporting my view. The first pillar I call "Advanced Packaging Taking the Baton." Look at our gross margin trajectory over the past four quarters — climbing from 17.3% all the way to 20.3%. That move can't be explained by pricing alone; it's driven by the rising revenue mix from our high-end processes, namely Fan-Out, 2.5D/3D, and CoWoS. Demand from AI accelerator cards and HBM-related packaging and testing is the biggest underpinning of my forecast. So I'm modeling Q2 2026 gross margin pushing up to roughly 20.8%, with Q3 hitting 21.2% — the ceiling on our product mix is still moving higher. The second pillar is "Capacity Allocation Supporting Utilization." We've poured serious capex into expanding our packaging and testing capacity over the past several years. For 1H 2026, I'm assuming utilization at our Taoyuan, Zhongli, and Kunshan facilities stays above 90%. The modest Q1 revenue dip from Q4 I read as seasonal inventory adjustment layered on top of the Lunar New Year effect; Q2 returns to the growth track and Q3 enters the traditional peak season. So I'm estimating Q2 revenue at roughly TWD 179.5 billion and Q3 at TWD 188.0 billion, with sequential momentum building each quarter. On the bottom line, operating margin expands in step with the improving product mix. I'm modeling Q2 operating profit at TWD 18.9 billion, breaching TWD 20.0 billion in Q3. Net income attributable to the parent comes in at TWD 15.7 billion and TWD 16.9 billion respectively, translating to EPS of roughly TWD 3.65 and TWD 3.93 — a measured, step-up trajectory. Of course, I have to be straight with you on the biggest risks — AI demand is a double-edged sword. If hyperscaler capex pacing slows, or if TSMC's CoWoS capacity ramps and crowds out our room on the advanced packaging order book, the pace of gross margin expansion may fall short of expectations. FX is also a hidden variable: every 1% appreciation in the TWD hits our USD-denominated revenue translation in a meaningful way. The near-term cyclical sweet spot is still intact, but as a veteran of this OSAT game, I always keep a healthy wariness toward the scenario of demand suddenly freezing up.