Layer: L7-OSAT | Date: 2026-05-28 | Price at screen: $71.14
| Metric | Value |
|---|---|
| Price (2026-05-28) | $71.14 |
| 52-week High | $79.23 |
| 52-week Low | $17.79 |
| Distance from 52w High | -10.2% |
| YTD Return | +66.1% |
| 200-day MA | $43.64 |
| 50-day MA | $62.28 |
| Beta | 2.31 |
FULL SCREEN triggered: YTD +66.1% (below 80% threshold), within 10% of 52w high.
Amkor Technology (NASDAQ: AMKR) is the world's #2 OSAT (Outsourced Semiconductor Assembly and Test) provider with ~15% global market share (vs ASE at 44%). Founded 1968, HQ Tempe AZ. $6.7B revenue FY2025, 30,800 employees. Primary markets: mobile/consumer, compute/AI, automotive. Key facilities: South Korea, Japan, Vietnam, Philippines, Taiwan, and a new $7B Arizona campus (Peoria).
| Metric | Value | Note |
|---|---|---|
| Market Cap | $17.6B | Post-April rerating |
| EV | $17.7B | Net debt near zero ($67M) |
| TTM EPS | $1.75 | Q2+Q3+Q4'25 + Q1'26 |
| 2026E EPS | $2.08 | Consensus 7 analysts |
| 2027E EPS | $2.45 | Consensus 5 analysts |
| Trailing P/E | 40.7x | |
| 2026E P/E | 34.2x | |
| 2027E P/E | 29.0x | |
| EV/EBITDA | 14.8x | |
| P/B | 3.94x | Book $18.04/sh |
| P/S | 2.49x | |
| Div Yield | 0.47% | $0.336/yr, quarterly |
Context on rerating: Stock was at $17.79 in March 2025. It surged ~300% in 12 months. The April 2026 surge (+57% in one month, $45 → $71) coincides with: (a) Q1 2026 record results and (b) May 21 Investor Day setting $11B 2030 revenue target. The stock now trades at investor-day optimism multiples — this is the key risk.
Apple = 29.8% of FY2025 net sales (~$2.0B of $6.7B). Top 10 customers = 72% of revenue.
Key mechanics:
| Dimension | Amkor | TSMC In-House |
|---|---|---|
| CoWoS | Outsourced "oS" layer (substrate bonding) | Owns "CoW" die-on-wafer step |
| Advanced packaging | SiP, FOWLP, Fan-Out — strong | InFO (own), CoWoS (own high-margin step) |
| US footprint | Building $7B Arizona plant | Fab 21 Phase 1/2 in Arizona |
| Customer diversification | Apple, NVIDIA, AMD, Qualcomm | Internal (Apple, NVIDIA anchor customers) |
| Margin | ~14% gross (low-cost, volume) | 53%+ gross |
| Competitive threat | TSMC is a PARTNER via MOU, not direct OSAT competitor for the outsourced steps | TSMC keeps high-value die bonding; Amkor does substrate work |
Key insight: TSMC does not compete with Amkor directly at the OSAT level — their MOU co-locates Amkor to handle the packaging steps TSMC doesn't want to own. The real competitor is ASE Technology (44% market share, aggressively expanding with 6 new fabs in 2026). Amkor's Arizona plant + TSMC MOU gives it a structural moat in US-based advanced packaging that ASE cannot easily replicate.
| # | Filter | Threshold | AMKR Result | Verdict |
|---|---|---|---|---|
| 1 | Adequate size | >$1.5B revenue | $6.7B FY2025 | PASS |
| 2 | Financial strength | Current ratio ≥ 2.0x | 2.27x | PASS |
| 3 | Earnings stability | Positive EPS 10yr | Positive all 4 yrs shown (2022-2025) | CONDITIONAL PASS (limited history available; peaked in 2022 at $3.11) |
| 4 | Dividend record | Uninterrupted | Quarterly since ~2021; ~5yr track | CONDITIONAL PASS (short record) |
| 5 | Earnings growth | EPS +33% over 10yr | 2022→2025: -52% (cyclical peak→trough); 2026E +39% recovery | FAIL (shr-017: cyclical trap — trailing data looks bad due to cycle, not secular decline) |
| 6 | Moderate P/E | <15x trailing | TTM 40.7x, 2026E 34.2x, 2027E 29.0x | FAIL (materially above) |
| 7 | Moderate P/B | <1.5x or P/E×P/B <22.5 | P/B 3.94x; P/E×P/B = 160x | FAIL |
Graham Score: 2/7 (or 3/7 with conditional passes) — Deep FAIL on value filters. This is NOT a Graham defensive value candidate.
shr-017 cyclical trap check: AMKR is genuinely a cyclical (semiconductor cycle driven). However, unlike Maersk, the forward earnings profile is improving (not deteriorating) — 2026E +39%, 2027E +18% growth from a valid fundamental driver (Arizona ramp + AI compute). The cyclical trap risk here is the inverse: buying on investor-day optimism at peak valuation, not buying on trailing peak metrics.
| Growth Rate | IV ($) | vs $71.14 | Margin of Safety |
|---|---|---|---|
| g = 0% | $14.88 | -79% | deeply negative |
| g = 3% | $25.38 | -64% | deeply negative |
| g = 5% | $32.38 | -55% | deeply negative |
| g = 7.5% | $41.12 | -42% | deeply negative |
| g = 10% | $49.88 | -30% | deeply negative |
| Growth Rate | IV ($) | vs $71.14 | Margin of Safety |
|---|---|---|---|
| g = 0% | $17.68 | -75% | deeply negative |
| g = 3% | $30.16 | -58% | deeply negative |
| g = 5% | $38.48 | -46% | deeply negative |
| g = 7.5% | $48.88 | -31% | deeply negative |
| g = 10% | $59.28 | -17% | negative |
| Growth Rate | IV ($) | vs $71.14 | Margin of Safety |
|---|---|---|---|
| g = 0% | $20.83 | -71% | deeply negative |
| g = 3% | $35.53 | -50% | deeply negative |
| g = 5% | $45.33 | -36% | deeply negative |
| g = 7.5% | $57.58 | -19% | negative |
| g = 10% | $69.83 | -2% | break-even |
Break-even growth rate using 2027E forward EPS = ~10%. The stock only reaches fair value if you believe Amkor can sustain ~10% EPS growth indefinitely from 2027. The Investor Day 2030 target of $5 EPS from $2.45 (2027E) implies ~27% CAGR 2027-2030, but that assumes full Arizona ramp execution — which is not yet proven.
| P/E Basis | P/E | Implied Graham g |
|---|---|---|
| Trailing TTM | 40.7x | 16.1% |
| 2026E consensus | 34.2x | 12.9% |
| 2027E consensus | 29.0x | 10.3% |
Gap between trailing and forward implied growth (16.1% vs 10.3%) = 5.8 percentage points. This gap reflects the market pricing a real earnings improvement (legitimate — 2026E is +39% growth driven by Arizona CPU ramp). But the market is also pricing in sustained ~10% long-run growth post-ramp, which is the optimistic scenario, not a margin-of-safety scenario.
| Year | Operating CF | CapEx | FCF | FCF Yield |
|---|---|---|---|---|
| 2022 | $1,099M | $908M | $190M | — |
| 2023 | $1,270M | $749M | $521M | — |
| 2024 | $1,089M | $744M | $345M | — |
| 2025 | $1,096M | $905M | $192M | 1.1% |
| 2026E | ~$1,200-1,400M | $2,500-3,000M | -$1,100 to -$1,800M | negative |
FCF will be deeply negative 2026-2027 during Arizona construction. This stock is in a heavy capex investment phase. Investors are paying for the 2028-2030 earnings stream, not current FCF. This is a growth capex story, structurally incompatible with Graham value criteria.
SBC: $20M in FY2025 (low, ~0.1% of market cap). Not a meaningful adjustment.
| Category | Signal | Status |
|---|---|---|
| Recent price action | +300% in 12 months; +57% in April 2026 alone on Investor Day run-up | RED — potential euphoria premium |
| Earnings vs expectations | Q1 2026: EPS $0.33 vs $0.24 est (+39.8%), Revenue $1.68B vs $1.44B est (+27.5%) | GREEN — genuine beat |
| Q2 guidance | $1.75-1.85B revenue, $0.42-0.52 EPS — above prior consensus | GREEN |
| Analyst rating | 2 Strong Buy, 3 Buy, 5 Hold, 0 Sell. Mean PT $75.50 vs price $71.14 (+6%) | AMBER — analysts only ~6% above current price |
| Insider activity | Multiple open-market SALES in May 2026 (CFO, GC, insiders; $750K-$3.7M) | RED — no open-market purchases; selling into the rally per shr-002 |
| Short interest | 7.45M shares short, 12.2% of float. Short ratio 1.43 days. Declining (from 8.7M prior month) | AMBER — moderate, declining, not squeeze-level |
| Dividend status | $0.336/yr, current, growing slowly (recent large 2024 Q4 was special div $0.488 one-time) | GREEN |
| Balance sheet | D/E 0.35x, current ratio 2.27x, net debt ~$67M | GREEN |
| Arizona plan risk | CHIPS Act still preliminary; $2.5-3B FY2026 capex = negative FCF | RED |
| Apple concentration | 29.8% revenue single customer | RED — structural |
| Competitive context | ASE accelerating 6-fab buildout; TSMC CoWoS oversupply from rival capex | AMBER |
| AI packaging claim | CPU/AI packaging tripling in 2026, concrete (NVIDIA, AMD programs) | GREEN — not "AI washing" per shr-032 |
Active REDs: 4 (euphoria premium, insider selling, Arizona execution risk, Apple concentration)
AMKR is a high-quality growth/capex story that has rerated from deep-value territory ($17 in March 2025) to growth-premium territory ($71 today). The thesis of "cheap OSAT with Apple and AI tailwinds" was valid at $17-25. At $71 it is a growth stock priced for near-perfect execution of a multi-year construction project. Graham defensive methodology is structurally the wrong tool here — this is a DCF/EV problem, not a P/B or P/E screen.
The satellite EUR ~3,000 portfolio already holds cyclical value (AKE.PA, AIG, AGN.AS, RI.PA). Adding AMKR adds: (a) USD currency concentration, (b) semiconductor cycle beta, (c) single-stock AI capex execution risk. VWCE already provides AMKR exposure (NASDAQ constituent). Per shr-019, the case for concentrating satellite capital here vs VWCE is weak when the stock has already rerated and offers no Graham-style margin of safety.
Watchlist candidacy: If AMKR pulls back to $45-55 range (the Dec 2025 / early Jan 2026 prices, before the investor-day run-up), the IV math becomes more interesting at 2027E EPS and the execution story would be partly de-risked by Arizona progress. That would be a 25-35% drawdown from current levels.
VERDICT LINE: 🔴 NO-BUY at $71.14 — Graham IV deeply negative at all growth scenarios using trailing/forward EPS; implied growth rate 10.3-16.1% required for fair value; 4 active REDs including euphoria premium post +300% rerating, systematic insider selling, negative FCF 2026-2027 from $7B Arizona build-out, and 30% Apple concentration; VWCE already provides exposure; revisit at $45-55 if Arizona ramp progresses and stock retraces to pre-investor-day levels.
Data sources: yfinance (prices, financials, insiders — 2026-05-28), SEC 10-K FY2025, SEC 8-K Q1 2026 earnings, Amkor Investor Day May 21 2026. Short interest from yfinance. Apple concentration from FY2025 10-K via SEC EDGAR.