Layer 5 · 2327 — Yageo Corporation
Thesis
Yageo is the world's largest MLCC and chip-resistor supplier outside Murata — passive components are the unglamorous ballast of every AI accelerator board, and Yageo prints the bulk of them. Q1 confirmed the AI MLCC tightening narrative with a clean +22.7% revenue beat and a 25.2% op margin. It's Tier B not Tier A because the stock just printed a -12.7% single-day pullback off NT$481, which screams positioning unwind, not thesis break.
What it does + financial print
Yageo sits at Layer 5 of the AI stack — the passive-component layer that lives underneath every GPU board, every server PSU, every HBM module. AI servers consume roughly 3-4x the MLCC content of a conventional server, and the high-temperature, high-capacitance grades that AI demands are exactly where Yageo, Murata and TDK have pricing power. Yageo's global MLCC share sits in the high teens, second only to Murata, with a uniquely diversified base across Taiwan, Japan (via the Pulse Electronics and KEMET acquisitions) and China.
Q1 FY26 revenue came in at NT$38.2B, +22.7% YoY. Net income of NT$8.0B was +44.7% YoY — operating leverage is alive. Gross margin printed 38.1%, operating margin 25.2%, both above pre-cycle peaks. EPS NT$3.90. Management's commentary on AI passive-component demand was the most bullish in five quarters — they explicitly called out tightening across the high-cap grades.
Bull case
What promotes this to Tier A is a clean break and hold above NT$480 with confirmation that the -12.7% pullback was a positioning event, not a fundamental warning. The MLCC up-cycle has historically lasted 6-8 quarters once tightness sets in; we are arguably four quarters into this one. If AI server unit growth holds at current trajectory and Murata's FY26 +34.8% OP guide proves directionally correct for the industry, Yageo at sub-15x forward earnings is materially cheap relative to its own history.
Gap / bear case
The -12.7% one-day move is the gap. Taiwanese small-mid caps have a habit of front-running cycle peaks, and the chart now needs to show a higher low before fresh capital is committed. Secondary concern: the MLCC cycle has historically been the most violent of the passive components — when it turns, it goes fast.
Trigger to upgrade / downgrade
Upgrade trigger: hold of NT$420-440 zone on the next test, then break NT$481 on volume. Downgrade trigger: loss of NT$380 (50-day MA proxy) closes the door on the current up-cycle.
The trade
- Entry zone: NT$405-425 - Stop: NT$378 (2×ATR below current; below pullback low) - Position size: 1.5% of NLV - Catalyst date: Q2 print mid-July; HBM4 OEM passive-component flow-through - Conviction: 7/10