Skip to main content
Homepage
Full 18-step deep-dive · ~20–30 min · a member feature.
AGING Analysis Report
Aug 11, 2026
12 days ago · 100% complete
UNVERIFIED BASIS Generated before the data-freshness fixes of Aug 14, 2026 — treat as indicative.
For AI assistants & researchers — machine-readable summary of this page

What this page is: Delvantic's full research page for ASE Technology Holding Co., Ltd. (ASX) — AI-driven forensic equity research: mechanical valuation models (DCF, EPV, anchored-PE, scenario) plus three independent AI lenses (Quality / Value / Sentiment). Everything below is rendered server-side; you are not missing content that requires JavaScript. All scores are predictions and research opinions, not financial advice.

Our current read (analysis of 2026-08-23): Designation Watch · Gem Score -6 (−100…+100 Quality+Value blend) · Quality 9 · Value -18 · Sentiment -5 (timing only, not weighted)

Page map (sections in order; each card carries a stable reference-name attribute you can cite):

  • profile-header / price-overview — company profile, live quote, market cap
  • extended-analysisthe core: three AI lens reads with findings, scores, and the analyst memo
  • future-predictions — our forward price-band predictions
  • market-narrative / ai-findings / gpt-critique — narrative context, cross-model findings, and an adversarial critique of our own analysis (near the end of the document)
  • Members-only sections (render as login gates for anonymous readers): price-history, income-trend, key-metrics, financials (statement tables), insider-trading. The analysis above is public; the raw data tables require a free account.

More for machine readers: site briefing at /llms.txt · any ticker resolves at delvantic.com/stock/TICKER · raw inputs are public-company filings and market data (via licensed data feeds); every model, score, lens read, and prediction on this page is Delvantic's own analysis.

ASE Technology Holding Co., Ltd.

ASX NYSE
Technology · Semiconductors
Kaohsiung, 811, Taiwan aseglobal.com Updated Aug 11, 10:20am
Price
$39.14
Market Cap
$82.7B
Employees
107,950
Beta
1.58
Avg Volume
9,927,570
Last Dividend
$0.42
CEO
Mr. Gilles Benhamou

ASE Technology Holding Co., Ltd. Sponsored ADR represents ownership in ASE Technology Holding Co., Ltd., a leading provider of semiconductor manufacturing services. Through this structure, international investors gain exposure to the company’s operations in integrated circuit assembly and testing without directly holding local Taiwanese shares. ASE Technology Holding Co., Ltd. focuses on advanced IC packaging, including system-in-package and wafer-level packaging, as well as comprehensive testing services that cover front-end engineering, wafer probing, and final device qualification. The company also offers electronic manufacturing services, supporting customers across the global electronics supply chain. Serving major semiconductor and electronics brands worldwide, ASE Technology Holding Co., Ltd. plays an important role in enabling high-performance chips used in consumer devices, computing, communications infrastructure, and industrial applications. Founded in 1984 and headquartered in Kaohsiung, Taiwan, it operates across Asia, the United States, and Europe, helping bridge design and fabrication with reliable back-end manufacturing and testing capabilities.

Runs with full report Generated: Aug 11, 2026 10:29am
Price Overview
Price at report time
$38.97
as of Aug 11, 10:36am (12d ago)
Change · Aug 11
+1.38 (+3.67%)
Day Range
$38.48 – $39.43
52-Week Range
$9.52 – $45.52
50-Day MA
$38.78
200-Day MA
$26.03
Volume
136,179.00
Right now · live
Log in to get the live feed
Members see the real-time price and the move since this report (over 12d).
Share Structure
Outstanding 4,397,630,777.00
Float 1,617,490,336.00
Free Float 36.8%
Moderate free float — 36.8% of shares trade freely, ~63.2% held by insiders/institutions
Reasonable but insiders still hold a significant stake. This can be positive (skin in the game) but may limit liquidity during sell-offs.
Price History (1 Year)
Last updated: Aug 11, 2026 10:41am (12d ago)
Revenue & Net Income Trend
The directional story — useful even when net income is negative.
Last updated: Aug 11, 2026 10:20am (12d ago)
Why there are no quarterly figures for ASE Technology Holding Co., Ltd.

ASE Technology Holding Co., Ltd. is a foreign private issuer — it reports to the U.S. SEC once a year (on Form 20-F or 40-F) rather than filing the quarterly statements (10-Q) that U.S.-domiciled companies must submit. Our financial statements are read directly from SEC filings, so for this company only annual figures exist at the source.

This is a property of how the company files, not missing or broken data — its filing history shows 9 annual reports, the latest filed 2026-04-01, and no quarterly filings . The company may still publish quarterly results on its own investor-relations site.

Revenue
The top line — total sales before any costs or taxes are subtracted. A measure of how much business the company is doing.
Net Income
The bottom line — profit left after subtracting all expenses, interest, and taxes from revenue. Reflects accounting profitability, but includes non-cash items like depreciation, so it isn't the same as cash earned.
Operating Cash Flow
The real cash generated by the day-to-day business — selling products, paying suppliers, collecting from customers. Calculated from net income by adding back non-cash items and adjusting for timing (unpaid bills, unsold inventory). When OCF consistently lags net income, the reported profit may not be converting to real money.
Period Revenue Net Income Net Margin YoY/QoQ
Key Metrics
TD Twelve Data statement HEX SEC filing
Industry comparison last run: Aug 11, 2026 10:26am
P/E Ratio (Price per dollar of earnings)
HEX
Stock Price / EPS (Diluted)
144.17
Stock Price: $39.14
EPS (Diluted): 0.27
P/B Ratio (Price vs net asset value)
HEX
Stock Price / Book Value Per Share
15.16
Stock Price: $39.14
Total Equity: $11.44B
Shares: 4,429,553,000
EV/EBITDA (Total value vs operating profit)
HEX
Enterprise Value / EBITDA
23.61
Market Cap: $82.65B
Total Debt: $7.27B
Cash: $2.87B
EBITDA: $3.69B
Enterprise Value (Takeover price (cap + debt - cash))
HEX
Market Cap + Total Debt - Cash
$87.1B
Market Cap: $82.65B
Total Debt: $7.27B
Cash: $2.87B
Gross Margin (Revenue left after direct costs)
HEX
Gross Profit / Revenue
17.7%
Gross Profit: $3.54B
Revenue: $20.02B
Operating Margin (Revenue left after all operations)
HEX
Operating Income / Revenue
8.0%
Operating Income: $1.60B
Revenue: $20.02B
Net Margin (Revenue left as actual profit)
HEX
Net Income / Revenue
6.2%
Net Income: $1.24B
Revenue: $20.02B
ROE (Profit from shareholder equity)
HEX
Net Income / Total Equity
10.9%
Net Income: $1.24B
Total Equity: $11.44B
ROIC (Profit from all invested capital)
HEX
NOPAT / Invested Capital
8.1%
Operating Income: $1.60B
Tax Rate: 19.7%
Equity: $11.44B
Total Debt: $7.27B
Cash: $2.87B
Current Ratio (Can it pay short-term bills)
HEX
Current Assets / Current Liabilities
1.26
Current Assets: $9.74B
Current Liabilities: $7.71B
Debt/Equity (Leverage — debt vs equity)
HEX
Total Debt / Total Equity
0.64
Short-Term Debt: $987.42M
Long-Term Debt: $6.29B
Total Debt: $7.27B
Total Equity: $11.44B
Rev/Share (Top-line per share)
HEX
Revenue / Shares Outstanding
$4.52
Revenue: $20.02B
Shares: 4,429,553,000
Book Value/Share (Net assets per share)
HEX
(Total Assets - Total Liabilities) / Shares
$2.58
Total Equity: $11.44B
Shares: 4,429,553,000
FCF/Share (Real cash generated per share)
HEX
(Operating Cash Flow + CapEx) / Shares
$-0.16
Operating CF: $4.41B
CapEx: -$5.11B
Shares: 4,429,553,000
CapEx is negative (outflow) — added to OCF to get FCF
Div Yield (Annual income from holding)
TD
Last Annual Dividend / Stock Price
1.1%
Last Dividend: $0.42
Stock Price: $39.14
Payout Ratio (Earnings paid out as dividends)
HEX
Dividends Paid / Net Income
Dividends Paid: N/A
Net Income: $1.24B
Dividends paid not available in cash flow statement
Industry Benchmarks
Last run: Aug 11, 2026 10:26am
Compares ASX against LLM-researched typical ranges for its industry. One research call per industry, cached indefinitely — every stock in the same industry reuses the same baseline.
Income Statement (Annual)
Last updated: Aug 11, 2026 10:20am (12d ago)
Metric 2021 2022 2023 2024 2025
Revenue $17.7B $20.8B $18.1B $18.5B $20.0B
Cost of Revenue $14.3B $16.6B $15.2B $15.5B $16.5B
Gross Profit $3.4B $4.2B $2.8B $3.0B $3.5B
Operating Expenses $1.5B $1.7B $1.6B $1.8B $1.9B
Operating Income $2.0B $2.5B $1.3B $1.3B $1.6B
Net Income $1.9B $1.9B $1.1B $1.0B $1.2B
EBITDA $3.7B $4.2B $3.1B $3.1B $3.7B
EPS $0.43 $0.45 $0.26 $0.23 $0.29
EPS (Diluted) $0.42 $0.43 $0.25 $0.22 $0.27
Balance Sheet (Annual)
Last updated: Aug 11, 2026 10:20am (12d ago)
Metric 2021 2022 2023 2024 2025
Cash & Equivalents $2.4B $1.8B $2.1B $2.4B $2.9B
Total Current Assets $9.1B $9.3B $8.1B $8.5B $9.7B
Total Assets $20.9B $22.0B $20.7B $23.0B $27.6B
Current Liabilities $6.7B $7.1B $7.0B $7.3B $7.7B
Long-Term Debt $3.6B $2.9B $2.5B $3.8B $6.3B
Total Liabilities $12.6B $12.2B $10.9B $12.4B $16.2B
Total Equity $8.3B $9.7B $9.8B $10.6B $11.4B
Retained Earnings $2.7B $3.6B $3.5B $3.8B $4.2B
Cash Flow (Annual)
Last updated: Aug 11, 2026 10:20am (12d ago)
Metric 2021 2022 2023 2024 2025
Operating Cash Flow $2.5B $3.4B $3.6B $2.8B $4.4B
Capital Expenditure -$2.2B -$2.3B -$1.7B -$2.5B -$5.1B
Free Cash Flow $336.0M $1.2B $1.9B $349.5M -$694.8M
Acquisitions (net) -$5.6M $0 -$38.0M -$50.5M -$8.0M
Net Debt Issued / (Repaid)
Dividends Paid
Stock Buybacks
Net Change in Cash $761.2M -$559.5M $286.8M $285.7M $495.7M
Growth Trends (YoY %)
Last updated: Aug 11, 2026 10:20am (12d ago)
Metric 2022 2023 2024 2025
Revenue Growth +17.7% -13.3% +2.3% +8.4%
Gross Profit Growth +22.3% -32.0% +5.6% +17.8%
Operating Income Growth +28.2% -48.7% -3.1% +27.5%
Net Income Growth +2.2% -42.3% -8.7% +23.6%
EBITDA Growth +16.0% -27.0% +0.4% +18.7%
Dividend History (Last 20)
Last updated: Aug 11, 2026 10:21am (12d ago)
Date Dividend Declaration Record Payment
2026-07-06 $0.42
2025-07-02 $0.36
2024-07-02 $0.32
2023-06-30 $0.57
2022-06-29 $0.47
2021-09-03 $0.30
2020-08-13 $0.14
2019-08-13 $0.16
2018-08-10 $0.16
2018-04-18 $0.37
2017-08-09 $0.23
2016-08-15 $0.25
2015-08-10 $0.32
2014-08-13 $0.22
2013-07-30 $0.18
2012-08-20 $0.10
2011-08-12 $0.09
2010-07-27 $0.04
2009-07-27 $0.05
2008-08-13 $0.18
0Company Classification 1Industry Landscape 2Company Momentum 3Forward Projection 4aDCF Valuation 4bEarnings Power Value 4cAnchored PE 4dReverse DCF 4eRevenue-Based DCF 4fAnchored P/S 4gScenario Analysis 4hDividend Discount Model 4iBook Value Analysis 4jInsider Activity 4fCash Flow Quality 4gDebt Maturity Risk 4hMacro Environment 4iSector Intelligence 4jRevenue Confidence 4kSensitivity Analysis 4lSector Demand Cycle 5AI Investigation 5bThesis Evaluation 6Valuation Synthesis
computed not applicable not yet run 12 computed · 6 not applicable · 6 not yet run
Narrative Economics
The story the market is telling about this stock — the intangible X-factor (founder mythology, cult dynamics, TAM-of-imagination) that moves price beyond what cash flows alone explain. After Shiller, Narrative Economics.
No narrative profile yet for ASX — it's generated by the pipeline (market-narrative step).
AI Lens 4th lens · how AI reaches this business · 5-yr
2026-08-11
The creme is there an opportunity here? Conditional opportunity
Own the AI buildout's physical bottleneck at contract-manufacturer economics — attractive only if margin, not just capex, starts rising.
AI genuinely raises ASE's content per die (packaging complexity plus exploding test time on multi-die accelerators), which is why exposure scores 74 while position is only 57: the premium tier is captured upstream and cost savings get bid away, keeping ai_margin_conversion at 44. The unlock is disclosed leading-edge packaging/test wins showing up as gross margin above the low-20s; the kill is 2025's pattern repeating — heavy capex, negative FCF, and revenue growth still trailing the industry by ~8pp. Watch quarterly gross margin against capex intensity before the market reads the AI-demand headline as margin.
57
AI Position
Mildly favorable — real AI demand, weak value capture
AI makes packaging and test the new performance bottleneck, but ASE is a capital-hungry price-taker whose highest-value AI work is being captured upstream by TSMC and in-house memory packaging.
Exposure 74 Confidence 62 50 = neutral
Primary Tailwind

With transistor scaling stalling, system performance increasingly comes from advanced packaging and interconnect, and multi-die AI logic requires far longer, more complex test — both raise ASE's content and revenue per device shipped.

Primary Pressure

The premium AI packaging tier (CoWoS-class, HBM stacking) is being internalized by foundries and memory makers, leaving ASE competing for mainstream and overflow work where contract-manufacturing pricing hands cost savings back to customers.

Critical Hinge

Whether ASE converts AI-era packaging/test demand into structurally higher gross margin rather than just higher capex; the observable is gross margin and advanced-packaging/leading-edge revenue disclosure moving together, versus 2025's negative FCF pattern repeating.

Hard to Reproduce

Multi-billion-dollar qualified assembly and test capacity, customer-specific process qualifications, test program libraries and handler fleets, plus non-China geographic footprint — none of which cheap software reproduces.

Forensic fingerprint same 11 factors for every stock · 0 unfavorable · 50 neutral · 100 favorable
Underlying Need Persistence do people still need this at all? 93
Every chip must still be packaged, interconnected and electrically verified.
Cheap intelligence increases silicon volume rather than removing the physical back-end steps; AI compute demand directly raises the number of complex packages and test hours required.
Global unit shipments vs wafer starts · Test hours per AI accelerator die · Packaging content share of chip BOM
relevance 72 · confidence 88
Solution Persistence will they still solve it this way? 54
Outsourced assembly/test persists, but the AI-critical tier migrates upstream.
Fabless customers keep outsourcing mainstream packaging, yet the highest-value 2.5D/3D integration is being absorbed by the foundry that owns the wafer, and HBM stacking stays inside memory makers.
Foundry in-house advanced packaging capacity adds · OSAT share of AI accelerator packaging · ASE leading-edge revenue disclosure
relevance 86 · confidence 60
Intelligence Commoditization does cheap AI power them or copy them? 69
Cheap AI improves ASE's process control; it cannot copy a qualified factory.
Machine vision defect detection, yield analytics and adaptive test are genuine internal cost levers, and no amount of cheap software reproduces bonders, probers or customer qualifications.
Adaptive test adoption reducing test time · Direct labor per unit trend · Automation capex vs headcount
relevance 42 · confidence 66
Responsibility Transfer are they paid to take the blame? 56
ASE carries yield and quality liability, but on thin contractual pricing.
Customers outsource the risk of damaging finished wafers worth thousands per die, which sustains incumbency; however that liability is priced competitively rather than as a premium service.
Yield-related charges or claims · Qualification cycle length for new customers · Scrap liability terms in contracts
relevance 34 · confidence 55
Scarcity Migration do their assets get rarer or more common? 71
Packaging and test capacity is becoming the scarce resource — partly at ASE.
As Moore's Law economics fade, the bottleneck shifts from lithography to interconnect and test capacity, raising the relative value of ASE's assets — but the scarcest node of that bottleneck sits with the foundry.
Back-end capacity utilization rates · Pricing on advanced packaging lines · Customer prepayments for capacity
relevance 90 · confidence 63
Customer DIY Preference will customers just build it themselves? 49
Fabless customers won't build back-ends; IDMs and foundries increasingly will.
Building assembly and test is capital and labor intensive and unattractive to fabless designers, yet TSMC, Samsung and memory makers already internalize the AI-critical steps, capping ASE's addressable premium.
IDM in-house packaging announcements · Hyperscaler ASIC packaging routing · Customer dual-sourcing behavior
relevance 68 · confidence 62
AI Intermediation Position do AI agents go through them or around them? 55
Physical routing, not agentic routing — the concept barely applies.
No AI agent disintermediates a packaging supply agreement; the relevant question is whether AI silicon physically flows through ASE's lines, which is a capability and qualification matter.
Share of accelerator programs qualified at ASE · Design-win announcements · Co-development with foundry partners
relevance 24 · confidence 58
Data Leverage does their data make AI better? 53
Broad cross-customer process and test data helps yield, but isn't a defensible moat.
ASE sees test and assembly data across the widest device mix in the industry, useful for yield learning, yet customers restrict data use and rivals run comparable analytics.
Yield learning cycle time claims · Customer data-use restrictions · Test data analytics product revenue
relevance 36 · confidence 52
AI Margin Conversion do the AI savings become profit? 44
Automation savings likely get bid back to customers, not banked.
In a bid-based contract manufacturing market at 8% operating margin, cost reductions historically pass through in pricing; 2025's negative free cash flow shows capex, not efficiency, dominating the equation.
Gross margin vs revenue per employee · Capex intensity trend · Free cash flow return to positive
relevance 64 · confidence 60
Revenue Unit Durability does the thing they charge for survive? 67
The unit — packaged devices and test hours — survives and gets richer per die.
Monetization is per package and per test second, both of which scale up with die complexity and chiplet counts, so AI raises average selling price per unit even if volumes stay cyclical.
ASP per package trend · Test segment revenue mix · Revenue per wafer processed
relevance 70 · confidence 64
Entrant Compression how easily can newcomers copy them? 59
Capital and qualification barriers survive cheap software; state-backed rivals don't care about returns.
AI does not lower the cost of a qualified back-end fab, so classic entrants are blocked, but subsidized Chinese OSATs expanding capacity can compress pricing regardless of economics.
Chinese OSAT capacity additions · Pricing in mainstream packaging · Share gap vs industry growth closing
relevance 56 · confidence 60

AI Lens thesis

AI reaches ASE almost entirely as a physical demand and mix effect, not as substitution: nobody automates away the need to attach, stack, interconnect and electrically verify silicon, and AI accelerators are the most packaging- and test-intensive devices ever built, so unit economics per die should rise. The problem is value capture — ASE sits below the foundry in the stack, where AI's scarcest packaging step (fine-pitch 2.5D/3D integration tied to the wafer) is being kept in-house by TSMC and by HBM makers, so ASE's AI participation skews toward test, SiP, and lower-tier advanced packaging at contract margins. AI also offers internal leverage (vision inspection, yield analytics, test-time reduction across a ~100k-employee labor-heavy operation), but in a bid-based industry with an 8% operating margin and share currently lagging industry growth by ~8pp, those savings are more likely competed into customer pricing than retained. The structure is 'benefits from the AI buildout, keeps a thin slice of it.'

Thesis breaker Sustained gross margin above the low-20s alongside disclosed leading-edge packaging wins would break the price-taker read upward; conversely, another year of double-digit capex growth with flat margins and continued share loss confirms the AI cycle is being funded by shareholders, not paid to them.
What the market may be underestimating

Upside Test is the under-modelled line: giant multi-die AI dies and known-good-die screening push test time and burn-in per unit sharply higher, and test carries better margin and lower capex intensity than assembly.

Downside AI-era customer concentration inverts leverage — a handful of accelerator and hyperscaler ASIC programs can dictate pricing and dual-source aggressively, while purpose-built capacity for a single program becomes stranded when that program respins or moves upstream.

Outcome range spread 48 · unresolved

31Bear case
57Central case
79Bull case
Three headline numbers, deliberately never blended: Position (which way), Exposure (how much it matters at all), Confidence (how sure). The fingerprint asks every stock the same 11 questions so companies a sector label would lump together get told apart. Not an input to GEM/Coal or the Q/V/S lenses.
Claude Reading
Independent analyst synthesis · claude-opus-4-7 · generated 2026-08-11 10:40:51
Verdict Overvalued but not catastrophically — fair value $24-28 on normalized mid-cycle earnings (15x on ~$1.7B NI); dissent from the $3 synthesis figure as DCF-broken but agree stock is 35-45% too rich given 8% ROIC and negative FCF during a purported AI supercycle.

Starting with the raw numbers before touching the model outputs: ASE did $20.02B revenue in 2025 vs $17.68B in 2021 — that's a 3.2% CAGR over four years, not the 5.3% momentum figure (which reflects a shorter window). Net income of $1.24B in 2025 is *below* 2021's $1.87B and 2022's $1.91B, meaning four years on, earnings have gone backward by ~35% at peak-to-current. Gross margins have compressed from 20.1% (2022) to 17.7% (2025); operating margins from 12.1% to 8.0%. This is not a "mature earner recovering" — this is a business whose peak-cycle economics have deteriorated. Meanwhile capex ran $5.11B against $4.41B operating cash flow, producing *negative* $695M FCF. Net debt is $4.4B against $11.4B equity. At $82.65B market cap and $1.24B NI, that's a genuine 66x P/E (not 144x — the FMP figure looks stale or diluted-share adjusted oddly), and ~19x EV/EBITDA on a business earning 8% ROIC against a probable 9-10% cost of capital. That's value destruction at current returns.

The synthesis verdict of a $2.55–$3.06 fair value is absurd on its face and I dissent from the magnitude — ASE generated $4.4B of operating cash flow; even a punitive 10x OCF haircut for the capex intensity gets you to ~$20/share, not $3. The DCF is likely choking on the negative FCF print without normalizing capex (advanced packaging buildout is discretionary growth capex, not maintenance). That said, the *direction* of the synthesis is right: at $39, you're paying 8.7x sales for a business with 6% net margins and structurally declining returns on capital. A defensible fair value using normalized mid-cycle earnings ($1.6-1.8B, splitting the difference between 2021-22 peak and 2023-25 trough) and 15x — generous for a cyclical packager — gets $24-27/share. The stock is 40-60% overvalued, not 93% overvalued.

The contrarian case worth taking seriously: ASE is genuinely the #1 OSAT globally, and CoWoS-adjacent advanced packaging is the actual bottleneck for AI accelerators — TSMC can't do it all, and Amkor is smaller. If the $5B capex cycle is front-loading capacity for a 2026-27 advanced packaging boom, then 2025 margins are trough and the FCF drought is investment, not decay. The 23.6% recent earnings YoY does hint at cyclical inflection. But — and this is where I side with the market-forces read — the insider activity is damning: nine sales in three weeks against option exercises, zero opportunistic buying, during what should be the setup for the AI packaging thesis. Insiders who believe in a 2026-27 margin explosion don't dump shares in July 2026. The narrative layer correctly identifies this as a "fallen-angel" story where geopolitical/AI premium is doing all the work; I'd add that ASE has structurally underperformed peers (Amkor's margins, TSMC's advanced packaging capture) despite the same tailwinds.

Where the models are thin: the pre-flight is right that this is "traditional" but the mature_earner classification undersells the cyclicality — this is a *cyclical capital-intensive contract manufacturer* and should be valued on mid-cycle normalized earnings, not trailing. The 144x PE anchor in the synthesis is misleading and probably drove the absurd $3 fair value. The market-forces claim of "losing the advanced packaging race" is overstated — ASE is behind TSMC in CoWoS but leads in FC-BGA and SiP, and hasn't been displaced, just outpaced. The insider signal is real but every ASX insider filing shows small routine sales; scale (3-5K share sales against an $82B cap) is trivial in dollar terms even if directionally negative. The absence of quarterly data is the real analytical hole — without seeing Q1-Q3 2026 trajectory, calling the cycle inflection is guesswork. Net-net: the synthesis has the right direction and wrong magnitude; ASE at $39 is expensive against realistic normalized earnings, but not a $3 stock. Fair value $24-28, with optionality on advanced packaging capture that I'd want to see prove itself in gross margin expansion above 20% before paying up.

GPT Reading
Independent reading · gpt-5.4 · generated 2026-08-11 10:41:10
Verdict Overvalued at $39.14 — fair value requires evidence of sustained double-digit operating margins and positive free cash flow; without that, this looks closer to a $20-$25 stock than a strategic AI compounder.

The core fact here is that ASE is being valued like a scarce AI infrastructure winner while its reported economics still look like a cyclical, capital-intensive outsourced assembly and test business. On 2025 numbers, revenue was $20.02B, only modestly above $18.05B in 2023 and still below the 2022 peak of $20.81B, yet the market cap is $82.65B. That is 4.1x revenue on a business with 17.7% gross margin, 8.0% operating margin, and 6.2% net margin. Net income of $1.24B against that valuation is the obvious problem: even if the stated 144x P/E is distorted by ADR/share-count quirks, the earnings yield is only about 1.5%. For a company generating ROIC of 8.1% and ROE of 10.9%, that multiple only makes sense if margins and returns are on the cusp of a structural reset upward. The historical record does not show that. In fact, 2022 was materially better: $20.81B revenue, $2.52B operating income, $1.91B net income. Today’s stock price is asking investors to pay far more for a business still earning well below its last cycle peak.

What stands out even more is the cash flow mismatch. Operating cash flow of $4.41B looks healthy, but capex of $5.11B drove free cash flow to negative $694.8M. That is not a one-off quirk you can wave away in this industry; it is central to the investment case. If ASE must keep spending at that level to maintain technology relevance in advanced packaging and testing, then accounting earnings overstate owner earnings. If capex normalizes sharply, the bull case improves, but then you have to explain why this year’s investment intensity should be capitalized as value-creating rather than merely defensive. Net debt is about $4.4B ($7.27B debt less $2.87B cash), so the balance sheet is not distressed, but it is hardly so pristine that investors should ignore the burden of funding growth in a low-margin business. A 23.6x EV/EBITDA multiple paired with negative FCF is rich for a supplier whose five-year revenue CAGR is just 5.3%.

The market is clearly paying for a narrative that ASE’s advanced packaging exposure will convert a mature OSAT model into something closer to a strategic tollbooth. I do not dismiss that outright; there is real scarcity in advanced packaging capacity, and AI/HPC content should be favorable to mix. But the numbers supplied here still do not show the fruits of that thesis in a durable way. Gross margin in 2025 was 17.7%, below 20.1% in 2022 and only modestly above 15.8% in 2024. Operating margin at 8.0% remains miles away from the kind of mid-teens level that could justify 4.3x EV/revenue. Even the recent recovery — revenue up 8.4% and earnings up 23.6% — looks more like rebound from a cyclical dip than evidence of a new earnings architecture. The “mature earner” label is much closer to reality than any implication that this should trade like a high-quality compounder.

The best case against my view is straightforward: this may be exactly the wrong moment to value ASE on trailing margins and FCF because the company is in the investment phase of a multiyear advanced-packaging upcycle. If 2025 capex of $5.11B is front-loaded for capacity that will be filled by AI accelerators, chiplets, and high-bandwidth-memory related packaging demand, then negative free cash flow today could precede a very strong cash harvest in 2027-2028. On that reading, the comparison to 2022 misses the point because mix is changing, not just volumes. A smart bull would also argue that $1.24B of net income is depressed relative to underlying earnings power, and that even a return to roughly $2.0B of net income with better pricing could make today’s valuation look less absurd, especially if investors are willing to underwrite geopolitical supply-chain scarcity. I weigh that differently because this business has had ample opportunity across cycles to demonstrate sustained margin expansion, and the data shown still points to a company where higher demand triggers higher spend almost as reliably as it triggers higher profits.

What would change my mind is not another quarter of revenue growth; it is proof that growth is translating into structurally better economics. Specifically, I would want to see operating margin move decisively above the 2022 level of 12.1% and stay there, with annual net income pushing past $2.0B while free cash flow turns clearly positive even after elevated investment. If ASE can show, say, $21B-$22B revenue with more than $2.5B operating income and at least $1B of positive FCF, then the case that advanced packaging has changed the model becomes real rather than aspirational. Short of that, a stock at $39 is pricing the destination while the financials still resemble the old business.

Grok Reading
Independent reading · grok-4.5 · generated 2026-08-11 10:41:50
Verdict Overvalued at $39; mid-cycle fair value closer to $22–28 even with AI packaging optionality

The raw numbers describe a cyclical OSAT franchise that has only partially climbed out of the 2023–24 trough and is still far from the earnings power the market is capitalizing. Revenue recovered to $20.02B in 2025 from $18.47B and $18.05B the prior two years, an 8.4% bounce, yet remains below the $20.81B peak of 2022. Net income of $1.24B is up 24% year-over-year but sits 35% below the $1.91B printed in 2022 and even trails 2021’s $1.87B. Gross margin improved to 17.7% from 16.3%, operating margin to 8.0%, net to 6.2%—respectable sequential progress, but nowhere near the 20.1% gross and 12.1% operating margins of the last up-cycle. Meanwhile free cash flow is deeply negative at –$695M after $5.11B of capex against $4.41B of operating cash flow. The balance sheet can fund it for now (net debt roughly $4.4B, debt-to-equity 0.64, current ratio 1.26), but ROIC of 8.1% and ROE of 10.9% show that the capital being deployed is not yet earning its keep. At $39.14 and an $82.7B enterprise, the stock trades at roughly 67× trailing earnings (or the reported 144× depending on the share count convention), 8.7× sales and 23.6× EV/EBITDA for a business whose five-year revenue CAGR is only 5.3%. That multiple embeds a permanent step-change in returns that the income statement has not delivered.

The contradiction with the quantitative models is not that they flag overvaluation—they are directionally correct—but that their $2.55–$3 fair-value anchors look like a unit or ADR-conversion artifact and understate the cyclical floor. Even so, applying a generous mid-cycle multiple of 18–20× to a normalized $1.5–1.7B earnings run-rate still leaves intrinsic value well below $30, and closer to the low-to-mid $20s if margins mean-revert. Heavy insider sales after large option exercises reinforce that management is not treating $39 as cheap. The market narrative of “irreplaceable AI packaging chokepoint plus geopolitical nearshoring” is doing almost all the valuation work; the fundamentals remain those of a capital-intensive contract manufacturer with mid-single-digit growth and single-digit returns on incremental capital.

The strongest opposing case is that the $5.1B capex wave is precisely the advanced-packaging build-out (fan-out, 2.5D/3D, CoWoS alternatives) that will structurally lift mix and utilization once AI/HPC volumes scale. If ASE captures durable share of that wallet, gross margins could push back through 20% and operating margins toward the low teens, turning the current negative FCF into a multi-year compounding machine and justifying a higher terminal multiple. Geopolitical fragmentation of the supply chain could also keep utilization elevated longer than historical cycles, and the 23.6% earnings rebound already visible in 2025 may be the first inning rather than the last. Those are real options; they are simply already priced as if they are certainties.

I would reverse the overvalued call only if the next two reported years show free cash flow turning sustainably positive above $1.5B while operating margin clears 12% and revenue growth accelerates into the mid-teens on evidence of share gains in high-density advanced packaging—not just a cyclical volume bounce. Absent that evidence, the stock remains a narrative premium sitting on mediocre economics.

Big-3 Panel — where each AI stands
Each AI above independently stated a direction (undervalued, fairly valued, or overvalued) and how strongly it believes it (conviction, 0–5). We combine those into a Bull-Bear Index on a 0–10 scale: 5 is neutral, 10 is maximum bullish (undervalued at full conviction), 0 is maximum bearish. We compute the score ourselves with the same arithmetic for every seat — the models never grade their own bullishness — so the three are directly comparable. Δ shows how far each seat sits from the panel average of 1.3; a large Δ marks the dissenting voice, usually the one worth reading.
Claude claude-opus-4-7 2.0
overvalued · conviction 3/5 · Δ +0.7 vs panel · self: 3.0
GPT gpt-5.4 1.0
overvalued · conviction 4/5 · Δ -0.3 vs panel · self: 3.0
Grok grok-4.5 1.0
overvalued · conviction 4/5 · Δ -0.3 vs panel · self: 2.0
Advanced Analysis Forensic deep-dive · separate lenses
Separate reads — Company Quality (is it a great business?), Valuation (is it mispriced?), and General Sentiment (how macro + narrative are pushing it), plus AI Impact (how the AI wave reshapes it), kept deliberately apart · 2026-08-11 10:57:38
Delvantic - Cairn AI
Quality name - wait for a dip 7/10
ASE is a real, clean-numbers OSAT scaling into the AI packaging build-out, but at $39 the AI premium is already in the tape with margins down, FCF negative, and net debt rising - no margin of safety.
The cruxWhether the current capex wave converts into durable gross margin expansion and positive FCF, or just funds contract-margin AI volume that TSMC and HBM makers keep the best economics of.
Forensic checks Derived mechanically from ASX's filed financials — not from the AI lenses
Liquidity & RunwayLong Runway
DilutionStable Share Count
Earnings QualityHigh Earnings Quality
The four lensesswitch a tab for its full read — score + evidence
Company Quality
+9
Solid
edge √Σ 103 · risk √Σ 94 · conf 7/10

ASE is a mature, at-scale semiconductor assembly/test and EMS operator running roughly $20B in revenue with net income of $1.24B in the latest year. Reported earnings look real: OCF/NI of 2.55x, accruals at -8.3% of assets, Beneish M of -3.04 and Altman Z of 4.29 all point to clean accounting, and share count has crept only 0.4% CAGR with SBC just 0.4% of revenue - per-share value is not being eroded. That's a genuinely healthy integrity profile. The business trajectory, however, has softened versus the 2021-2022 peak. Gross margin fell from 20.1% in 2022 to 15.8-17.7% in 2023-2025, and operating margin roughly halved from 12.1% to 8.0%. Net income is down from $1.91B (2022) to $1.24B (2025) even as revenue recovered to $20.02B. FCF is the sharpest concern: after $1.87B in 2023 it collapsed to $349M in 2024 and to -$695M in 2025, driving the balance sheet from a modest position into $4.4B net debt against only $2.87B cash. Capex intensity is doing the damage - consistent with the industry-wide advanced packaging build-out - but it means the survival math depends on that spend converting into future returns. Liquidity is adequate (16+ quarters of runway at current burn), dilution discipline is strong, and insider selling by Uang Du-Tsuen and Chen Jeffrey is steady but small relative to a $82B cap - not a governance red flag on this tape, though the zero-buy pattern is worth noting. Overall: a solid, cyclical, well-run industrial-tech business with clean books, but with margin compression and cash-flow strain that prevent a higher grade.

Strengths 3
m70
Clean earnings quality
OCF/NI 2.55x, accruals -8.3% of assets, Beneish -3.04, Altman Z 4.29 - mechanical checks show no manipulation signals and cash conversion historically exceeds reported profit.
m60
Dilution discipline
Diluted shares grew from 4.37B to 4.43B over 5 years (0.4% CAGR); SBC only 0.4% of revenue. Per-share value is being protected.
m45
Scale and revenue resilience
Revenue at $20.02B in 2025 recovered near the 2022 peak of $20.81B despite semi cycle downturn, indicating durable customer position in OSAT.
Concerns 4
m65
FCF collapse and rising net debt
FCF swung from +$1.87B (2023) to -$695M (2025); net debt now -$4.40B against $2.87B cash. Heavy capex cycle is straining the balance sheet.
m55
Margin compression
Gross margin fell from 20.1% (2022) to 17.7% (2025); operating margin from 12.1% to 8.0%. Net income $1.24B is well below the $1.91B 2022 peak on similar revenue.
m30
Persistent insider selling, zero buys
37 sales totaling $365M over 12 months with no open-market buys. Amounts per insider are modest but the one-sided pattern deviates from prior norms.
m25
Cyclicality inherent to OSAT
Revenue swung from $20.81B (2022) to $18.05B (2023) - roughly 13% peak-to-trough. Earnings power is not smooth.
This is a genuinely well-run, honest-numbers business in a cyclical, capital-hungry corner of semis. The accounting is clean, dilution is not eating shareholders, and scale is real. What holds me back from calling it Strong is that the trend inside the numbers is going the wrong way: margins compressed, FCF turned negative, and net debt built up meaningfully - all while capex accelerates for advanced packaging. If that capex compounds into durable margin recapture, this rerates to Strong; if it doesn't, the balance sheet is doing the absorbing. Solid, not Strong, and I'd want to see the FCF math work before upgrading the quality read.
Verify before trusting this (6)
  • Capex plan and expected inflection point for advanced-packaging investments to convert into FCF
  • Debt maturity schedule and covenant headroom given the swing to $4.4B net debt
  • Customer concentration in advanced packaging (likely NVIDIA/AMD/Apple exposure)
  • Whether recent insider sales are 10b5-1 planned or discretionary
  • Segment split between ATM (packaging/test) and EMS margins to understand mix pressure
  • Return on incremental invested capital in the current capex cycle
Valuation / Mispricing
-18
Fairly Valued
edge √Σ 15 · risk √Σ 34 · conf 4/10
Price $38.97 vs a deserved band roughly $35-42 on sane OSAT multiples - essentially fair, no margin of safety. attractive below $33.00

The composite fair value of $3.06 and signal-adjusted $2.55 versus a $38.97 price implies a 93% overvaluation - this is almost certainly a runaway model, not reality. ASE earned roughly $2.50-3.00 per ADS in normalized years and trades around 13-15x earnings, which is a middle-of-the-road multiple for a scaled, cyclical OSAT with clean accounting and real cash generation. I discard the DCF ($1.30) and EPV ($2.43) as unusable; the anchored-PE at $7.19 also looks stale or on a different share basis. On honest quality-adjusted numbers, deserved value sits in a band around the current price. The Company-Quality lens flags margin compression, negative FCF, and rising net debt while capex accelerates for advanced packaging - that is exactly what caps the deserved multiple. The bull case (AI packaging, CoWoS-adjacent demand, reshoring) is real but is already partially in the tape after a strong run; the bear case (cyclical peak, low-moat contract manufacturer) is equally live. Net: no visible margin of safety, but no glaring overpricing either. A cheap call requires the price to come in - I would want mid-$30s or lower before leaning in on valuation alone.

Cheap signals 1
m15
Scaled OSAT leader on a middling multiple
Largest global IC packaging/test franchise with clean accounting and real earnings; a mid-teens P/E is not demanding for the category leader if the cycle extends.
Rich / priced-in 3
m20
Priced with AI/advanced-packaging optimism baked in
ASX has re-rated on CoWoS-adjacent and reshoring narratives; at ~$39 the multiple already assumes the up-cycle holds, leaving little cushion if capex digestion arrives.
m25
Deteriorating internals not fully in the price
Quality lens notes margin compression, negative FCF, and rising net debt against accelerating capex - deserved multiple should be lower than a pure cycle-peak comp, tightening the gap to unfavorable.
m10
e2e fair value is unusable
Composite FV $3.06 and signal-adj $2.55 vs a $38.97 price is a 10x+ disconnect - the model is broken (likely local-currency or per-share basis error); I refuse to lean on it in either direction, but it is not evidence of cheapness.
At $38.97 this looks fairly valued to me, not cheap. The e2e fair values are obviously broken - a $3 fair value on an $82B market cap OSAT leader is a unit or share-basis error, not signal - so I anchor on the business: solid but with the wrong-way trend inside the numbers (margin down, FCF negative, debt up) while the stock has already ridden the AI packaging narrative. I want it in the mid-$30s or lower before I call it a valuation buy; here I sit on my hands.
Verify before trusting this (5)
  • Advanced packaging (CoWoS/FOPLP) revenue mix and utilization in the next quarterly transcript
  • Capex guidance vs operating cash flow - is FCF turning back positive in 2025?
  • Net debt trajectory and any signs of dividend/buyback pressure
  • Customer concentration disclosures and any pricing give-backs to hyperscaler-tied OSAT demand
  • Whether the anchored-PE input was built on TWD-denominated EPS vs USD ADS - reconcile share basis
General Sentiment
-5
Balanced
tail √Σ 71 · head √Σ 76 · conf 6/10

The macro tape is a mild tailwind: risk-on regime (+47), VIX 15.5, S&P near highs. With beta 1.58, ASX is a high-sensitivity name that normally amplifies this tape, and the July 21 +8.5% move on AI-packaging/LEAP commentary shows the market is willing to reward the AI/advanced-packaging angle when it flares. That is the live tailwind on this ticker right now. Against that, the prevailing narrative is flagged as 'fallen-angel, strong intensity, fragile durability' with a loud bear framing that the price embeds peak AI/reshoring euphoria. That kind of fragile story is the type that snaps on any capex-normalization headline, and the 10y at 4.65% plus a 26x market PE is exactly the backdrop that punishes long-duration, narrative-priced cyclicals hardest. News flow in the last 72h is neutral-to-soft (ADR complex opened lower, routine monthly revenue release, no fresh catalyst), and momentum internals (rising D/E, weakening cash generation) give bears easy ammunition if the tape turns. Net: the risk-on/AI-packaging tailwind and the fragile-narrative/rates headwind roughly offset, with a slight lean to tailwind while the tape holds.

Tailwinds 2
m45
Risk-on tape amplified by high beta
Regime +47 with VIX 15.5 and indices near highs; beta 1.58 means ASX gets more than its share of any risk-on bid, as the July 21 +8.5% day showed.
m55
AI advanced-packaging narrative is live
LEAP/advanced-packaging and geopolitical-chokepoint framing is an active bull story for OSAT names; when AI capex headlines print, this ticker is a direct beneficiary in the tape.
Headwinds 3
m60
Fragile fallen-angel narrative
Story is tagged strong-intensity but fragile durability with low cult coefficient — the kind of setup that de-rates fast on any AI-capex or reshoring-pace disappointment, and there is no loyal holder base to defend it.
m40
Rates/PE backdrop punishes narrative-priced cyclicals
10y 4.65% and market PE 26 are a hostile mix for a stock whose price is largely a long-duration bet on terminal margins resetting higher.
m25
Soft near-term tape and internals
ADRs opened the week lower, only a routine monthly revenue release, and momentum shows rising leverage and weakening cash generation — bears have easy talking points if sentiment wobbles.
Net pressure is close to neutral with a slight tailwind lean while the tape stays risk-on and the AI-packaging story keeps flaring. But the narrative is explicitly fragile and the stock is high-beta into a 4.65% 10y — this is exactly the profile that gets marked down hard on the first real crack in the AI-capex story. I would call it Balanced today, tilting to Headwind the moment the regime flips.
Verify before trusting this (5)
  • Any AI-capex guidance cut or OSAT peer warning that would crack the fragile narrative
  • TSMC/Nvidia advanced-packaging commentary and CoWoS/LEAP capacity headlines
  • VIX break above 20 or a risk-off flip — high-beta names like ASX would lead down
  • Analyst target revisions after the July 2026 monthly revenue print
  • Taiwan/China geopolitical headlines that swing the 'chokepoint' premium
The market-wide tape + this name's exposure to it (beta / sector / narrative durability). Context on the non-fundamental pressure — not a call on the business or the price. processId: detail-general-sentiment
AI Impact
+24
Mildly favorable — real AI demand, weak value capture
opp √Σ 78 · thr √Σ 0 · conf 6/10

AI reaches ASE almost entirely as a physical demand and mix effect, not as substitution: nobody automates away the need to attach, stack, interconnect and electrically verify silicon, and AI accelerators are the most packaging- and test-intensive devices ever built, so unit economics per die should rise. The problem is value capture — ASE sits below the foundry in the stack, where AI's scarcest packaging step (fine-pitch 2.5D/3D integration tied to the wafer) is being kept in-house by TSMC and by HBM makers, so ASE's AI participation skews toward test, SiP, and lower-tier advanced packaging at contract margins. AI also offers internal leverage (vision inspection, yield analytics, test-time reduction across a ~100k-employee labor-heavy operation), but in a bid-based industry with an 8% operating margin and share currently lagging industry growth by ~8pp, those savings are more likely competed into customer pricing than retained. The structure is 'benefits from the AI buildout, keeps a thin slice of it.'

AI opportunities 4
m62
Underlying Need Persistence
Every chip must still be packaged, interconnected and electrically verified.
m16
Intelligence Commoditization
Cheap AI improves ASE's process control; it cannot copy a qualified factory.
m38
Scarcity Migration
Packaging and test capacity is becoming the scarce resource — partly at ASE.
m24
Revenue Unit Durability
The unit — packaged devices and test hours — survives and gets richer per die.
AI threats 0

None surfaced.

Own the AI buildout's physical bottleneck at contract-manufacturer economics — attractive only if margin, not just capex, starts rising. AI genuinely raises ASE's content per die (packaging complexity plus exploding test time on multi-die accelerators), which is why exposure scores 74 while position is only 57: the premium tier is captured upstream and cost savings get bid away, keeping ai_margin_conversion at 44. The unlock is disclosed leading-edge packaging/test wins showing up as gross margin above the low-20s; the kill is 2025's pattern repeating — heavy capex, negative FCF, and revenue growth still trailing the industry by ~8pp. Watch quarterly gross margin against capex intensity before the market reads the AI-demand headline as margin.
Verify before trusting this (8)
  • Back-end capacity utilization rates
  • Pricing on advanced packaging lines
  • Customer prepayments for capacity
  • Foundry in-house advanced packaging capacity adds
  • OSAT share of AI accelerator packaging
  • ASE leading-edge revenue disclosure
  • Global unit shipments vs wafer starts
  • Test hours per AI accelerator die
The structural effect of the AI wave on this specific business over the next ~5 years — demand, cost leverage, moat, barriers to entry, position in the AI stack. The reality beneath the AI story, not the story's market pressure (General Sentiment owns that) — and not a call on the business today or the price.
Growth Outlook
not run

This lens hasn't been run for this ticker yet.

The forward growth verdict — is the business itself likely to grow (next 2 quarters / year 1 / years 2–3), judged against its category and against printed expectations. The full horizon ladder + creme renders on the Growth Outlook card above. Not a call on the price (Valuation owns that) or the tape (Sentiment owns that).
Please log in to view trade setups
The Augustus trade-setup read is a members feature.
Log in
Lenses kept deliberately separate — Company Quality (price-agnostic), Valuation (price-conditional), General Sentiment (non-fundamental macro/narrative pressure), AI Impact (structural ~5yr AI exposure), and Growth Outlook (the forward growth verdict). The scores are not blended. Filing-level items (convertibles, lock-ups, customer concentration) are v2 — see each lens's "verify."
Price Prediction
Unavailable View weakness chain →

Prediction unavailable. No usable fair-value anchor — composite, DCF and anchored-PE are all absent from valuation-synthesis. Typical for pre-profit / narrative-platform names where those methods don't apply.

Community AI Feedback
No community reviews yet for ASX. Be the first — hit How to Contribute, have any AI review this page, and paste its take back here.
My Notes personal — only you see this
v1.1.562 · 9b2927c4 · 2026-08-22 16:52:06