Screened: 2026-05-28 | ADR price: USD 40.99 | Underlying: 3711.TW | Format: ABBREVIATED (YTD >80% + within 10% of 52w high)
ABBREVIATED REPORT TRIGGERED: YTD +154.6%, distance from 52w high -0.3%. Thesis question answered in the first paragraph — full screen not warranted.
| Metric | Value |
|---|---|
| Price (ADR, USD) | USD 40.99 |
| 52-week range | USD 9.23 – USD 41.10 |
| YTD return | +154.6% |
| Distance from 52w high | -0.3% (essentially AT the high) |
| 5yr return | +556% (from USD 6.25) |
| Return from 5yr low | +916% (USD 4.03, Oct 2022) |
| Market cap | ~USD 90B |
| Avg daily volume | 7.7M shares |
The thesis hook asked whether ASX had rerated like other AI-adjacent names (IMOS +340% from lows was the benchmark). The answer: yes, and then some. ASX is up +916% from its 2022 trough and sitting 0.3% below its all-time 52-week high. This is not a trough screen — it is the opposite.
| Metric | Value | Graham Limit | Verdict |
|---|---|---|---|
| P/E trailing | 63x | <15x | FAIL |
| P/E forward | 26x | <15x | FAIL |
| P/B | 8.1x | <1.5x | FAIL |
| P/E × P/B | 510x | <22.5 | EXTREME FAIL |
| Current ratio | 1.15x | >2.0x | FAIL |
| Dividend yield | 0.91% | Uninterrupted record | PARTIAL |
| FCF 2025 | NEGATIVE (-USD 700M eq.) | Positive | FAIL |
Graham filters: 2/7 PASS (adequate size, earnings not negative). Five filters fail. This is not a Graham value situation under any plausible earnings scenario.
| Growth | IV (trailing EPS USD 0.65) | MoS | IV (forward EPS USD 1.59) | MoS |
|---|---|---|---|---|
| g = 0% | USD 4.86 | -88% | USD 11.86 | -71% |
| g = 3% | USD 8.29 | -80% | USD 20.24 | -51% |
| g = 5% | USD 10.58 | -74% | USD 25.82 | -37% |
| g = 7.5% | USD 13.44 | -67% | USD 32.80 | -20% |
| g = 10% | USD 16.30 | -60% | USD 39.78 | -3% |
Even at g = 10% forward EPS, Graham IV ≈ USD 39.78 — essentially at today's price, with zero margin of safety. To get a 30% margin of safety on forward EPS requires ~g = 15% sustained growth. Trailing EPS (USD 0.65) produces IV of ~USD 5-16 across all scenarios: the stock is priced at 2.5–8x its Graham trailing value.
The trailing vs. forward P/E gap (63x vs. 26x, per shr-003) signals the market is pricing a profitability inflection — EPS expected to roughly double from 2025 to 2026 on CoWoS/advanced packaging ramp. The implied growth rates confirm this: trailing implies 27.3% sustained growth, forward implies only 8.75%. Exactly the inflection-pricing pattern shr-003 warns about.
What ASE does: World's #1 OSAT (Outsourced Semiconductor Assembly & Test) by revenue. ~31% global OSAT market share. Key services: wire bonding, flip chip, SiP (System-in-Package), and advanced packaging (CoWoS-adjacent via their own FO-WLP / fan-out technologies).
AI/HBM tailwind — already priced: ASE is a direct beneficiary of the AI packaging wave. CoWoS demand (TSMC-driven, for NVIDIA/AMD AI GPUs) has created overflow demand flowing to OSAT players for adjacent substrate and packaging work. The market has fully repriced this: USD 9 → USD 41 in 12 months.
Apple exposure: Apple is ASE's largest customer (~20-25% of revenue, primarily iPhone SiP modules). This creates a defensive floor but also a seasonal ceiling — iPhone replacement cycles directly drive ASE utilisation rates in H2.
ABF substrate: ASE is a net consumer of ABF substrates (not a producer like NOD or Ibiden). ABF squeeze (2021-2023) compressed margins; normalisation of ABF supply from 2024 onwards is partly what drove the earnings recovery. As a buyer, ASE benefits from easing ABF costs — but it cannot "pass through" substrate costs in the way a substrate manufacturer can; it absorbs them. With ABF now in mild oversupply, this is a tailwind that is already in 2025 numbers.
FCF problem: Capex tripled in 2025 (TWD 165B vs TWD 82B in 2024) as ASE builds advanced packaging capacity for CoWoS-S adjacent work. This flipped FCF negative despite operating cash flow growing. This is growth capex, not distress, but it means the balance sheet is leveraging up: total debt rose from TWD 201B to TWD 264B. Net debt/EBITDA is manageable but the capex cycle is not finished.
Not applicable — no buy contemplated. For reference if price corrects materially:
VERDICT: RED for satellite purchase at current price.
ASX is a fundamentally excellent business — world #1 OSAT, direct CoWoS/AI packaging beneficiary, Apple floor, ABF tailwind. But the market has already recognized all of this. The stock has repriced from a deep-value trough to a growth-premium. At USD 41:
Watchlist trigger for revisit: USD 18-22 range (representing ~50% drawdown and ~20-25% MoS on forward EPS at g=5-7.5%). That would require a semiconductor downcycle, geopolitical event, or loss of major customer — possible but not near-term imminent.
Comparison to IMOS: Both are OSAT ADRs that have fully rerated from AI tailwinds. IMOS scored RED for identical reasons on 2026-05-28. The window to buy OSAT at value was 2022-2023 (USD 4-9 range for ASX). That window is closed.
VERDICT LINE: RED — fully rerated AI/OSAT premium; 2/7 Graham filters, P/E×P/B=510x, no MoS at any trailing or forward scenario; analyst mean price USD 38.64 is BELOW current; watchlist only at USD 18-22 on deep drawdown.