Amkor Technology (AMKR)
TL;DR
Amkor is the world's #2 semiconductor packaging-and-test company (largest US-based) — it assembles finished chips into the advanced packages AI accelerators, smartphones and cars need. Just reported: Q1 2026 revenue up 27% to $1.7B, led by premium smartphones (Apple) and recovering autos, with AI advanced packaging guided to triple this year. The stock roughly doubled to ~$70 on that AI story plus an ambitious Investor Day (~$5 EPS by 2030), but at ~36x it prices a lot for a cyclical, capital-intensive business. The risks: Apple is ~30% of revenue, free cash flow is negative during a $7B Arizona buildout, and TSMC packages the top AI chips in-house.
Summary
| Finding | Assessment |
|---|---|
| 5-Year Stock Return | $70.47 now vs $23.59 five years ago (+199%); a cyclical OSAT re-rated on the AI-packaging narrative. |
| Business Model | #2 global OSAT (semiconductor advanced packaging + test); Communications (Apple), Computing (AI), Auto/Industrial, Consumer. |
| Competitive Moat | MODERATE Real advanced-packaging tech + #2 scale + sticky qualifications, but TSMC-in-house, cyclical, Apple-concentrated. |
| Past 5-YR Drivers | The AI advanced-packaging pivot, the semiconductor-cycle recovery, and advanced-package content growth. |
| Recent Stock Move | $70.47, off an ~$86–93 June peak; +136% over two years. |
| Revenue Trend | DECELERATING YoY revenue growth 27.5%→22.5% (Q1'26→Q2'26E). |
| Margin Trend | Slight Up EBITDA margin 16.9%→16.9% (Q1'26→Q2'26E). |
| TTM OP Trajectory | DECELERATING TTM EBITDA QoQ +7.6%→+4.3% (Q1'26→Q2'26E). |
| Key Opportunities | AI advanced packaging (3×), the 2030 model (~$5 EPS), and US-packaging (Arizona). |
| Key Risks | Apple ~30% concentration + cyclicality, a capex surge + negative FCF, and TSMC-in-house + a rich multiple. |
| Catalysts | AI-3× delivery, margin inflection (2H), and Arizona milestones. |
| P/E(ttm) Valuation | 36x — re-rated from single digits; prices the 2030 AI vision optimistically. |
What Just Reported
- A record first quarter: revenue $1.7B (+27%), gross margin 14.2% (a beat), and EPS of $0.33. Factory utilization jumped to the low-70s (from the 50s a year ago) as the business climbed out of the semiconductor downturn.
- But the growth was smartphone- and auto-led, not yet AI. Communications rose +42% (premium smartphones, Apple-driven) and Automotive & Industrial +28% (ADAS) — while Computing grew a slower +19% (record AI data center, offset by soft PCs). The AI advanced-packaging surge is the forward story, guided to triple this year.
- Free cash flow was negative (−$71M). Capital spending is ramping hard toward the $7B Arizona buildout — the near-term cost of the long-term bet.
AI Packaging, Arizona & the 2030 Vision
- AI advanced packaging is guided to triple in 2026. Amkor's HDFO and 2.5D platforms (its answers to TSMC's CoWoS — "SWIFT" and "S-Connect") have a broad pipeline (11 customers on 2.5D, 5 on HDFO), with a data-center CPU program ramping mid-year.
- Arizona is the strategic US bet. A $7B campus (CHIPS-funded) will be the only scaled advanced-packaging plant in the US, packaging Apple/TSMC-Arizona silicon — but it doesn't start production until 2028 and dilutes margins ~1–2% from 2027 first.
- The May Investor Day set an ambitious model: ~$11B revenue, ~22% gross margin and ~$5 EPS by 2030 (more than triple 2025), driven by AI mix, operating leverage and the Arizona ramp.
Key Takeaways from Earnings Call
- "An advanced-packaging-led era is here." New CEO Kevin Engel frames the thesis around Moore's Law slowing and "More than Moore" advanced packaging becoming the critical path for AI/HPC chips — with Amkor positioned as the #2 global and largest US OSAT.
- The recovery is broad-based. All four end markets grew, utilization is rising, and 2H gross margin is guided up to the mid-to-high teens on better utilization and AI mix.
- Capital intensity is the tension. Capex jumps to $2.5–3.0B (from $905M) for Arizona; the company issued a $1.2B convertible and its proforma leverage rose to ~2.1× — a heavy investment cycle ahead of the revenue.
- Competition and concentration are the debates. Analysts pressed on TSMC's in-house CoWoS and Apple concentration; management argues supply constraints make customers "very motivated" to qualify Amkor as a second source (a tailwind), and points to 30+ year customer relationships.
- Watch the AI mix vs the headline. Q1's +27% was Apple/auto-led; the bull case depends on the AI-3× advanced-packaging ramp actually landing through 2026 against material-supply constraints.
Key End Markets and Revenue Mix (FY2025)
| End Market | FY2025 Revenue | % · Growth | Key Details |
|---|---|---|---|
| Communications | ~$3.1B | 46% · +1% | Premium smartphones — Apple is ~30% of total revenue; seasonal. |
| Computing | ~$1.3B | 20% · +16% | AI data center + HPC (the growth engine) + PCs; the AI-packaging story. |
| Automotive & Industrial | ~$1.3B | 19% · +8% | ADAS, electrification; recovering from the downturn. |
| Consumer | ~$1.0B | 15% · +9% | Wearables, AR/gaming, connected home. |
| Total | $6.7B | +6% | Advanced products 83%; gross margin 14%; EPS $1.50; #2 global OSAT. |
Business Model
SUPPLIERS
Chip designers & foundries (ship wafers to Amkor)
Substrates, bonding materials, equipment
Advanced-packaging R&D (SWIFT / S-Connect / HDFO)
13M sq ft across 9 countries
Wafers +
materials
→
AMKOR
OSAT: advanced packaging (2.5D / HDFO / flip-chip / wafer-level) + test
Turnkey assembly; #2 global, largest US-HQ
~14% gross margin · ~17% EBITDA margin
Packaged &
tested chips
→
$
CUSTOMERS
Apple (~30%), Qualcomm (~11%)
NVIDIA, AMD, Broadcom, Marvell (AI)
Auto, industrial & consumer chipmakers
Top 10 = 72% of sales
Moat Assessment: MODERATE
- Real advanced-packaging technology + scale. As the #2 global OSAT and largest US-headquartered one, Amkor has proprietary 2.5D and HDFO platforms (its answers to TSMC's CoWoS), high capital and qualification barriers, and 30+ year relationships with 80% of its customers — stickier than a commodity contract manufacturer.
- A strategic US position. The $7B Arizona campus (CHIPS-funded) makes Amkor the only scaled advanced-packaging OSAT in the US, positioned to package Apple/TSMC-Arizona silicon as supply chains regionalize.
- But it's contestable, cyclical and concentrated. TSMC keeps the highest-value AI packaging (CoWoS) in-house, ASE is the larger #1 competitor, the business is deeply cyclical and capital-intensive (free cash flow is negative during the Arizona build), and Apple is ~30% of revenue — so we rate the moat MODERATE.
Stock vs. Earnings — 5-Year Performance
- Stock $70.47, up ~199% over 5 years and +136% over 2 — a cyclical packaging house that re-rated hard, now off an ~$86–93 June-2026 peak.
- The earnings base is cyclical, not compounding: trailing EBITDA reached $1,372M and FY2025 EPS was $1.50 — recovering from the 2023–24 downturn but well below the 2022 peak until the AI ramp scales.
- The multiple did the work: P/E ran from a historical ~5–17× to ~36x — a steep re-rating for a capital-intensive, cyclical business, on the strength of the AI-packaging story and the 2030 Investor Day model.
Past 5-Year Key Drivers
- Advanced-packaging content growth. Advanced products rose to 83% of revenue as chips increasingly need 2.5D, fan-out and flip-chip packaging — the structural "More than Moore" shift that favors OSATs like Amkor.
- The semiconductor cycle. Revenue swings with the chip cycle — a "higher highs, higher lows" pattern ($4.3B/2018 → $7.1B/2022 peak; $4.1B/2019 → $6.3B/2024 trough) — now recovering on AI and premium smartphones.
- The AI-packaging + US-manufacturing pivot. HDFO/2.5D platforms for AI accelerators plus the CHIPS-funded Arizona campus repositioned Amkor as a strategic AI-infrastructure and US-supply-chain play.
Quarterly Key Metrics
| USD | Q2'24 | Q3'24 | Q4'24 | Q1'25 | Q2'25 | Q3'25 | Q4'25 | Q1'26 | Q2'26E | Q3'26E |
|---|---|---|---|---|---|---|---|---|---|---|
| Quarter End | Jun 2024 | Sep 2024 | Dec 2024 | Mar 2025 | Jun 2025 | Sep 2025 | Dec 2025 | Mar 2026 | Jun 2026 | Sep 2026 |
| Stock | $38.58 | $29.85 | $25.38 | $18.08 | $20.99 | $28.41 | $40.32 | $41.25 | $86.23 | $70.47 |
| Stock QoQ | 24% | -23% | -15% | -29% | 16% | 35% | 42% | 2% | 109% | -18% |
| Rev YoY | 0% | 2% | -7% | -3% | 3% | 7% | 16% | 27% | 22% | 20% |
| OM% | 17% | 17% | 19% | 15% | 17% | 17% | 20% | 17% | 17% | 17% |
| R40 | 17% | 19% | 12% | 12% | 21% | 24% | 35% | 44% | 39% | 37% |
| TTM OP | $1,139M | $1,115M | $1,091M | $1,055M | $1,067M | $1,098M | $1,165M | $1,253M | $1,307M | $1,372M |
| TTM OP QoQ | 0% | -2% | -2% | -3% | 1% | 3% | 6% | 8% | 4% | 5% |
| FCF/OP | 44% | 38% | 32% | 21% | 26% | 21% | 16% | 13% | -25% | -39% |
| Debt/EBITDA | 0.7x | 0.7x | 0.7x | 0.7x | 1.0x | 1.1x | 0.8x | 0.8x | 0.8x | 0.8x |
| P/E | 29x | 23x | 21x | 17x | 20x | 28x | 31x | 27x | 49x | 36x |
Recent Stock Performance & Drivers
- Stock $70.47, up ~136% over two years and ~199% over five — a cyclical OSAT that re-rated hard, now off an ~$86–93 June peak.
- What drove it: the AI advanced-packaging narrative (guided to triple in 2026), the recovery from the 2023–24 semiconductor downturn (utilization 50s → 70s), and a May Investor Day targeting more than triple earnings power (~$5 EPS) by 2030.
- At ~36×, expensive for a capital-intensive, cyclical business: the market is paying for the 2030 AI vision, which requires flawless execution through a heavy Arizona capex cycle.
Opportunities
- AI advanced packaging. HDFO and 2.5D platforms (rivals to TSMC's CoWoS) guided to triple in 2026, a data-center CPU program ramping, and a broad pipeline (2.5D: 11 customers, HDFO: 5); OSAT server advanced-packaging demand ~3× by 2030.
- The 2030 model. Investor Day targets ~$11B revenue, ~22% gross margin and ~$5 EPS (>3× 2025) as AI mix, operating leverage and the Arizona ramp compound.
- US-packaging optionality. The Arizona campus — the only scaled US advanced-packaging OSAT, CHIPS-funded — is a strategic asset as supply chains regionalize (~$1B revenue at >30% gross margin at full scale).
Challenges
- Extreme customer + cyclical concentration. Apple is ~30% of revenue and the top-10 are 72%; the business is deeply cyclical (revenue was negative year-over-year through 2024), and the current recovery is smartphone-led as much as AI-led.
- Capital intensity + negative FCF. Capex jumps to $2.5–3.0B (from $905M) for the $7B Arizona build, turning free cash flow negative and diluting margins ~1–2% from 2027 before Arizona contributes revenue (2028+).
- The TSMC threat + a rich multiple. TSMC packages the highest-value AI chips (CoWoS) in-house, ASE is the larger competitor, and ~36× is a steep price for a cyclical packaging house executing a multi-year, capital-heavy turnaround.
Next 12 Months
- The AI-3× delivery + FY2026. Whether AI advanced packaging actually triples and the data-center CPU ramp (meaningful in Q3) lands, against material-supply constraints.
- Margin inflection. 2H gross margin guided to the mid-to-high teens on utilization + AI mix — the near-term proof point for the 2030 model.
- Arizona milestones + the balance sheet. Construction progress and CHIPS funding, and how the capex cycle strains leverage (proforma ~2.1× after a $1.2B convertible).
Next 12-24 Months
- The Arizona ramp (2028+). First production in 1H 2028, break-even 2029, full 2030 — the biggest swing factor for the long-term margin story.
- AI-packaging share vs TSMC/ASE. Whether Amkor's SWIFT/S-Connect win durable AI-accelerator packaging as a "second source" or remain secondary to TSMC's in-house CoWoS.
- Cyclical durability. Whether the "higher highs, higher lows" pattern holds as advanced-package content per system rises across all four end markets.
Core Metrics Scorecard
| Q4'21 | Q1'22 | Q2'22 | Q3'22 | Q4'22 | Q1'23 | Q2'23 | Q3'23 | Q4'23 | Q1'24 | Q2'24 | Q3'24 | Q4'24 | Q1'25 | Q2'25 | Q3'25 | Q4'25 | Q1'26 | Q2'26E | Q3'26E | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Dec21 | Mar22 | Jun22 | Sep22 | Dec22 | Mar23 | Jun23 | Sep23 | Dec23 | Mar24 | Jun24 | Sep24 | Dec24 | Mar25 | Jun25 | Sep25 | Dec25 | Mar26 | Jun26 | Sep26 | |
| R40 > 40% | ✅ | ✅ | ✅ | ✅ | ||||||||||||||||
| R40 accelerate | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ||||||||||
| TTM OP QoQ > 10% | ||||||||||||||||||||
| TTM OP QoQ accelerate | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ||||||||||||
| TTM FCF/OP > 60% | ❌ | ❌ | ❌ | ❌ | ❌ | ❌ | ❌ | ❌ | ❌ | ❌ | ❌ | ❌ | ❌ | ❌ | ❌ | ❌ | ❌ | ❌ | ❌ | ❌ |
| Debt/EBITDA < 3.5 | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ |
| Total passes | 3 | 2 | 1 | 4 | 1 | 1 | 3 | 1 | 3 | 2 | 2 | 2 | 1 | 1 | 3 | 3 | 3 | 4 | 1 | 1 |
| Stock QoQ | -2% | -8% | -26% | -1% | +45% | +8% | +15% | -24% | +48% | -3% | +24% | -23% | -15% | -29% | +16% | +35% | +42% | +2% | +109% | -18% |
| P/E ttm | 9x | 8x | 6x | 5x | 8x | 10x | 12x | 14x | 24x | 23x | 29x | 23x | 21x | 17x | 20x | 28x | 31x | 27x | 49x | 36x |
- Amkor screens as a cyclical semiconductor packager, not a compounder. It sits at 1–4 passes with a Rule-of-40 that swings with the cycle (single digits in the 2023–24 downturn, ~44% now) — and R40 crosses 40 only because the sheet uses the ~17% EBITDA margin, not the ~6% GAAP operating margin.
- The re-rating is a narrative, not a scorecard breakout. P/E went from ~5–17× to ~36× on the AI-3× advanced-packaging story and the 2030 Investor Day vision — the multiple, not the passes, drove the ~2× move.
- Cash conversion is the flag. Free cash flow turned negative as capex tripled for the Arizona build; the scorecard's cash test fails in the forward quarters, and the $7B campus won't contribute revenue until 2028 — a long, capital-heavy bet.
Chart: R40 vs Stock Price (Since 2020)
Chart: Scorecard Total Passes vs Stock Price (Since 2020)
Source: SEC EDGAR annual and quarterly filings, company earnings conference calls and presentations.
Disclaimer: This report is for educational purposes only. NO investment advice.