Skip to content

Dossier · UMC · Dormant

UMC · United Microelectronic Corp.

MEDIUM Special situation Catalyst · ai-chip-infra-memory

Last analysed ·

Current thesis

Stale "sleepy ADR" tag broken: UMC re-rated ~3x off its $6.56 low to $21.63 (6/12) as the tape reprices an advanced-node/AI-photonics pivot. Q1 EPS doubled YoY (beat), monthly sales accelerating +10.8%→+17.78%, while sell-side targets ($7.40–$10.36) sit 50%+ below spot. Accelerating momentum; June sales print ~7/8 the next read.

Invalidation trigger

Weekly close back below the rising 50-day average (~$15.50), or June/July monthly sales YoY decelerating under ~+10% (reversing the +10.8%→+17.78% acceleration); an RSI>90 blowoff into the ~late-July Q2 print would also cap the leg.

Thesis status

Open commitment catalyst in 24dscored if the trigger above fires How this is scored →

Current Thesis

The "sleepy mature-node foundry" tag is stale. The UMC ADR has re-rated roughly 3x off its $6.56 52-week low to a $21.63 close (2026-06-12), sitting 8.7% under its $23.69 high after a +35.9% month and a +9.8% week, as the market reprices a legacy specialty foundry into an advanced-node and AI-photonics pivot. The fuel is fundamental, not just tape: a Q1 EPS beat that doubled YoY (NT$1.29 vs NT$0.85 consensus, reported ~2026-05-06), monthly sales accelerating from +10.8% YoY in April to +17.78% in May (NT$22.94B, 2026-06-05), and an Intel 12nm Arizona JV targeting late-2027 production. Sell-side targets ($7.40–$10.36) sit 50–66% below spot the narrative is running weeks ahead of the analysts who cover it, which in this regime is confirmation, not a warning.

Bull Case

  • Q1 2026 EPS NT$1.29, more than double YoY, beat NT$0.85 consensus on revenue NT$61.04B (+5.5% YoY), record 22nm-node revenue (reported ~2026-05-06) EPS doubling on mid-single-digit revenue growth is margin and mix leverage, the cleanest signal the 2024–25 mature-node glut has turned.
  • Monthly sales re-accelerating: April +10.8% YoY → May +17.78% YoY (NT$22.94B, reported 2026-06-05) two consecutive prints accelerating, with the recurring catalyst trending the right way into the June read (~2026-07-08).
  • Intel 12nm Arizona JV reaffirmed on track for late-2027 production (cited at the 2026-05-27 52-week high) capital-light advanced-node exposure plus AI-photonics investment lets UMC climb the value chain without funding leading-edge capex alone.
  • 14nm embedded high-voltage FinFET display-driver platform launched 2026-05-14 first FinFET process for smartphone display drivers, shifting mix away from commodity 28/22nm into defensible specialty silicon.
  • Planned H2-2026 wafer price increases to support $1.5B capex (flagged on the Q1 call) returning pricing power inverts the price-war narrative that capped the stock through 2025.
  • Price $21.63 against sell-side targets of $7.40–$10.36 consensus is 2–3x behind; the cluster of semis ripping alongside (Micron's UBS target tripled, 2026-05-26) marks sector-wide repricing rather than a one-stock spike.

Bear Case

  • Sell-side targets $7.40–$10.36 vs $21.63 spot, consensus "Hold," 4 of 6 analysts hold-or-sell either the Street is badly behind or the ADR is 100%+ over fair value; a single mark-to-reality downgrade could open a sharp air-pocket.
  • Zero-coupon convertible bond raise approved 2026-06-03 (~NT$16B, dual tranche) it funds expansion but adds a dilution overhang the stock must absorb, and convertible arbitrage desks typically short the underlying.
  • Dividend yield compressed from ~6% to ~2% as the ADR tripled the old income and valuation-floor argument is gone; this is now a pure growth re-rate with no payout cushion underneath.
  • Core revenue is still commodity mature-node SMIC and Hua Hong keep adding 28/22nm capacity in China; if the H2 wafer price hikes fail to stick, gross-margin compression resumes.
  • The Intel JV payoff is late 2027 a meaningful slice of the bull case is a 2027 option being paid for at a 52-week-high multiple, and any timeline-slip headline cuts both ways.
  • Extension and structure risk +35.9% in a month, 8.7% below the high, ~40% above the 50-day average; ADR liquidity is thinner than US-domiciled peers and carries a Taiwan-Strait geopolitical tail.

Setup & Price Structure

  • $21.63 close (2026-06-12), +4.59% on the day, $21.67 after-hours pressing the top of the range with the $23.69 high the only overhead resistance.
  • +35.9% 1-month, +9.8% 1-week, RSI 65 a strong trend, but mid-60s RSI is not a blowoff; there is room before the >88 zone where momentum names typically crack.
  • Price ~40% above the 50-day (~$15.50) and ~113% above the 200-day (~$10.15), market cap $53.1B a stage-2 advance with both averages rising and stacked beneath price, and enough float to absorb institutional flow.
  • Beginner-trap check: this is institutional semis repricing, not WSB/StockTwits retail froth; Q1 already printed and Q2 is ~late July, so no earnings inside three trading days; price is well above support, so it is not averaging-down terrain. The single live caution is chasing a 52-week-high name after a 35% month moderate RSI and fundamental confirmation argue the leg is not yet exhausted, but a fresh entry here is buying strength, not a base.

Catalyst Calendar (next 30 days)

  • ~2026-07-08 (est.): June 2026 monthly sales print the near-term needle-mover; the read is whether the +10.8%→+17.78% acceleration extends. A third accelerating month confirms the trend; a sharp deceleration is the first crack.
  • 2026-06-03 → ongoing: convertible bond pricing and placement watch the terms and any conversion-price overhang as the two tranches settle.
  • ~2026-07-29 (est.): Q2 2026 earnings just outside the 30-day window; the next real margin and utilization read, and the test of whether the H2 wafer price hikes are landing.
  • Late 2027: Intel 12nm Arizona production ramp structural, not tradeable now, with headline risk both directions on any schedule change.

What Would Change Our Mind

  • Trend break: a weekly close back below the rising 50-day average (~$15.50) ends the momentum leg and argues to stand aside until the name bases.
  • Fundamental break: the June or July monthly sales print decelerating back under ~+10% YoY reverses the acceleration that underwrites the re-rate.
  • Blowoff exhaustion: RSI pushing >90 into the ~late-July Q2 print on thinning volume and stalling price marks mania rather than trend and argues for patience over chasing.
  • Pricing-power failure: any sign the H2-2026 wafer price increases are being walked back peer commentary, customer pushback removes the margin lever doing the heavy lifting.
  • Confirmation to press: a third consecutive accelerating sales month plus the first sell-side target revisions toward spot would signal the narrative going mainstream with room still left before saturation.

Correlation Notes

  • Trades as high-beta foundry/SOX-SMH semis exposure, moving with TSM, SMIC and GlobalFoundries across the mature-node complex. The 2026-05-26 session (Micron's tripled UBS target, alongside RDW/MOD/ASTS) shows UMC riding sector repricing, not idiosyncratic alpha alone.
  • The specialty mix display drivers, RF, BCD power, automotive ties UMC to broad silicon-content growth more than to AI-GPU capex, a partial hedge against an accelerator-digestion scare that would hit Nvidia-levered names harder.
  • The Intel JV adds an INTC-headline correlation: Intel foundry-roadmap news, positive or slipped, moves UMC's 2027 narrative leg directly.
  • ADR structure and Taiwan domicile layer in an NT$/USD currency factor and a Taiwan-Strait geopolitical tail absent from US-domiciled foundry peers.

Notes

  • Low-beta dividend ADR (~5-7% yield) slow vehicle, sizes as a probe not a conviction position even when theme tag reads ACCELERATING.
  • Intel 12nm Arizona JV payoff is 2027 structural, not tradeable near-term; watch for timeline-slip headline risk both ways.
  • Reports monthly Taiwan sales ~8th-10th of each month the primary recurring catalyst; April 2026 was NT$22.6B +10.8% YoY (2026-05-08).
  • Q2 2026 earnings ~late July (est. ~2026-07-29) outside 30d window; next real margin/utilization read.
  • Theme tag overstates AI linkage core revenue is mature-node specialty (display/RF/power), not AI-GPU. Treat as foundry-cycle/value-rotation beta.
  • CORRECTION (2026-06-13): prior 'low-beta sleepy ~$8 ADR / 5-7% yield floor' framing was a no-price-feed guess and is WRONG. ADR re-rated ~3x off the $6.56 52-week low to $21.63 (2026-06-12 close), market cap $53.1B, +35.9% 1mo / +9.8% 1wk, RSI 65. Yield compressed to ~2%; trades as high-beta semis momentum, not an income vehicle.
  • Monthly Taiwan sales print (~7th-10th each month) is the primary recurring catalyst. April +10.8% YoY, May +17.78% YoY (NT$22.94B, reported 2026-06-05). Next: June print ~2026-07-08 watch whether acceleration extends to a third month.
  • Q1 2026 was a genuine earnings inflection: EPS NT$1.29 (>2x YoY) beat NT$0.85 consensus on revenue NT$61.04B (+5.5%), record 22nm revenue (~2026-05-06). H2-2026 wafer price hikes flagged to fund $1.5B capex margin lever to track.
  • Sell-side targets $7.40-$10.36 are 50-66% below the $21.63 spot narrative running well ahead of analysts. First PT revisions UP toward spot = narrative going mainstream; that is the saturation tell to watch for later.
  • Convertible bond raise approved 2026-06-03 (~NT$16B zero-coupon, dual tranche) for capacity expansion dilution/conversion overhang; arb desks typically short the underlying.
  • Q2 2026 earnings ~2026-07-29 (est.) outside the 30d window; next margin/utilization read. Intel 12nm Arizona JV is late-2027 production: structural, not tradeable near-term, timeline-slip headline risk both ways.

Related · shared themes