Semiconductor Materials

Alright, let me break down the numbers for these two quarters from Yue Qingning's perspective. Let me start with the core call—2026Q1 is already flashing early recovery signals: revenue did slide noticeably from Q4, but gross margin rebounded from 27.4% back to 31.2%, and operating profit was pulled up to 355 million yuan, essentially flat with Q3. This tells me that Q4 gross margin collapse was most likely a one-off inventory write-down / temporary product-mix disruption, not a systemic deterioration of the core business. Two key assumptions underpin my forecast. First, **customer inventory destocking across the two main pillars—precursors and high-purity specialty gases—should wrap up by late Q2**, and combined with H2 being the typical peak procurement window for domestic wafer fabs, revenue will recover sequentially off the Q1 trough—but the pace will be moderate, not a vertical takeoff, given that key customers (YMTC, SK Hynix Wuxi, etc.) are still keeping capex pacing on a tight leash. Second, **product mix keeps tilting toward high-end precursors and advanced-node specialty gases**—pricing power and gross margin moats are concentrated in these segments, so gross margin stabilizing around 32% is a reasonable equilibrium, and we shouldn't see an extreme print like that 27% Q4 figure anywhere in the full year. On the revenue line, I'm sketching a gradual ramp—roughly 2.03 billion in Q2, 2.07 billion in Q2 (note: original text shows both as Q2, appears to be Q2/Q3). Net profit tracks accordingly, landing at 252 million in Q2 and 260 million in Q2, pacing broadly in sync with revenue. I'm not assuming any "operating leverage outpacing revenue" upside—mainly because I assume opex ratio stays rigid with no meaningful compression. But the biggest risk I have to flag explicitly: **memory customer capex falling short of expectations**. A meaningful chunk of Yoke's electronic chemicals business rides on memory fabs. If overseas majors like Samsung and SK Hynix continue to push out equipment orders, and domestic YMTC's expansion cadence undershoots its earlier guidance, the H2 pull-in season could get pushed back by a quarter or two, forcing gross margin into a defensive crouch. Separately, the LNG plate business has its own delivery-timing variable on the overseas project side—order recognition timing is genuinely hard to pin down precisely. Bottom line in one sentence: the direction is gradual repair, but don't bet on a V-shaped reversal.