NYSE ADR (1 ADR = 5 TWSE ordinary shares) · operating company in Hsinchu, Taiwan · ISIN US8740391003 · CEO: C. C. Wei · Mkt Cap: $2.21T · Lifecycle: Growth (dominant, cash-generative)
$425.83
−$15.57 (−3.53%) on the day
Jun 16, 2026 · Signal v6
DISCLAIMER: This is a quantitative framework for educational purposes only. It is not financial advice. Always do your own research and consult a licensed financial advisor before making investment decisions.
Horizon
Signal
Composite Score
Confidence
Key Driver
Short-term (1-3 mo)
HOLD
58
60%
Great business but extended/overbought near the 52-week high; risk-off macro + FOMC/earnings ahead
Medium-term (6-12 mo)
BUY
67
62%
Quality + AI driver offset a full valuation; near-term tech is still a mild macro headwind, so no amplification
Long-term (3-5 yr)
STRONG BUY
74
65%
Dominant moat + AI secular driver (88) + long-horizon tech/EM tailwind → BUY amplified to STRONG BUY
Override chain per horizon: Base Matrix → Amplification → Hard Gates → Do-Not-Buy. Hard-gate state: CAUTION — Taiwan/China geopolitical binary tail risk flagged (not fired); valuation extended near 52-week high; Q2 earnings approaching.
Five independent scores — Business Quality, Valuation Attractiveness, Entry/Exit Timing, Underlying Drivers, and Economic Alignment — each 0–100 with confidence. The base BUY/HOLD/SELL comes from the three fundamental pillars (Quality/Valuation/Timing) via the Decision Matrix; the two context pillars (Drivers, Economic Alignment) then amplify it toward STRONG BUY / STRONG SELL.
Business Quality
88
Best-in-class foundry: 62% gross margin, 36% ROE, net cash
Confidence: 82%
Valuation Attractiveness
57
Fair — cheap vs AI peers / forward EPS, rich on trailing & FCF
Confidence: 78%
Entry/Exit Timing
55
Strong uptrend but overbought near highs; rolling over intraday
Contrarian short · Trend-Following long · Headwind→Tailwind
Confidence: 70% · Macro report 2026-06-13
2
Hard Gates & Do-Not-Buy Status
Binary safety checks — financial distress, earnings-event risk, valuation ceiling, dilution/accounting, and regulatory/binary events. Any triggered gate caps the signal regardless of composite score; a Do-Not-Buy trigger overrides everything. State: CAUTION — no gate fired, but the Taiwan/China geopolitical tail and valuation extension are flagged.
✅
Financial Distress Net cash, interest coverage 176×, current ratio 2.49, D/E 0.17. Clear.
Regulatory / Binary (Taiwan) China-Taiwan invasion/blockade is a genuine >20% binary tail — persistent & structural, not a dated ruling, so flagged as caution, not fired.
✅
Severe Driver Collapse Driver score 88 (Strong Tailwind) — nowhere near the ≤15 collapse threshold. Clear.
Do-Not-Buy triggers: none fired. Leverage+rising-rates — N/A (net cash). Valuation-at-extreme — disarmed: the multiple sits high in its range but growth is accelerating (revenue +35% YoY, EPS +58%). Negative earnings revisions — none (estimates rising). Insider-selling spike — none observed. Structural threat — the Taiwan tail is a known, persistent risk, not a newly-emerged un-priced threat, so it is not invoked as an Analyst Override.
3
Pillar Detail: Business Quality
Why Quality scored 88: sub-signals vs the Semiconductors profile (cycle-aware), competitive moat, ROIC & capital allocation, and the industry benchmark. Scored relative to sector norms, with cycle context (peak vs trough).
Business Quality — Pillar Score
Among the highest-quality businesses on the watchlist: a near-monopoly in leading-edge foundry with 62% gross margins (66% last quarter), 36% ROE, ~47% net margin, and a net-cash balance sheet. The only quality nuance is cyclicality — current margins are near cycle-peak.
Heavy capex ($40B+/yr) consumes ~58% of operating cash
Reinvestment, not weakness
Balance sheet
Net cash; D/E 0.17; coverage 176×
Pristine
Downturn-proof
Capacity utilisation / book-to-bill
Advanced node >90%; CoWoS sold out
>80% = healthy
Demand outstrips supply
Competitive Moat — 82 / 100
Pricing Power
90
Raising leading-edge wafer prices; no real alternative
Network Effects
60
Design ecosystem (OIP), but limited two-sided network
Switching Costs
85
PDKs, multi-year qualification, tape-out cost
Cost Advantage
90
Scale + yield/learning-curve leadership
Intangibles
85
Process IP; ~67% leading-edge share, >90% at advanced node
Industry Benchmark — Gross Margin + Capacity Utilisation: 92/100 (STRONG). Gross margin 62% TTM / 66% latest quarter (threshold >55%); advanced-node utilisation >90% with CoWoS advanced-packaging sold out (threshold >80%). Both well above the "strong" band. Cycle caveat: these are near cycle-peak readings — a demand digestion would pull margins toward the high-50s, still excellent but off the peak.
ROIC & Capital Allocation — 86/100. ROIC well above cost of capital and stable/rising; disciplined capex tied to customer pre-commitments; a steadily-growing dividend (payout ~26%) plus a flat share count (no dilution). FMP financial-health rating: B+ (overall 3) — ROE and ROA score the maximum 5, dragged only by P/E (2) and P/B (1), i.e. the model agrees the business is excellent and the price is full. That cross-check lines up exactly with the Quality-high / Valuation-fair split below.
4
Pillar Detail: Valuation Attractiveness
Why Valuation scored 57: cycle-adjusted multiples vs semis peers and TSMC's own history, reverse-DCF implied growth, FCF-yield anchor, analyst consensus targets & grades, and the FMP ratings cross-check.
Valuation Attractiveness — Pillar Score
Fair, at the Attractive/Fair boundary. Genuinely cheap versus the other AI mega-caps and on forward earnings (21.7× / PEG ~1.3); rich on trailing P/E, price-to-book and FCF yield, and trading at ~90% of its 52-week range. The forward lens, not the trailing one, is what keeps this from scoring lower.
57
Confidence 78%
Lens
Reading
Score band
Sector / peer median (fwd P/E)
21.7× vs NVDA/AVGO/AMD at 25–35×
Attractive vs peers
Own historical decile (trailing P/E)
~35× — high end of TSMC's 5-yr ~13–35× range
Expensive on its own history
Growth-adjusted (PEG)
~1.3 (fwd); 0.68 trailing
Fair / reasonable
Reverse-DCF implied growth
~low-teens priced in vs 35% rev / 58% EPS actuals
Market under-pricing trajectory
FCF yield (anchor)
~1.6% (heavy capex supercycle)
Expensive on cash
Price / Book
10.6×
Rich
Dividend yield
~0.9% (payout ~26%)
Growth, not income
Reverse-DCF read. At $425.83 the ADR trades on ~21.7× forward EPS (~$19.66). For a near-monopoly compounding revenue 35% with expanding margins, that multiple implies the market is discounting only low-teens long-run growth — i.e. it is not extrapolating the current AI-led trajectory. That is the bull's valuation argument: you are paying a market-average growth multiple for a clearly-above-average grower.
Analyst consensus & grades. 18-analyst mean target ~$468, median ~$465–472 (Yahoo) → price is ~9% below median = meaningful upside. FMP shows a wider, more bullish spread (median $465, consensus $520, high $700) skewed by one fresh $700 print; the high-vs-low gap ($354–$700) is wide, so conviction is trimmed. Grades consensus: 18 Buy / 7 Hold / 0 Sell (Buy, ~72% bullish); Yahoo's recommendation reads "strong buy" (1.47). Recent grade actions are all "maintain" — no fresh upgrades or downgrades.
Embedded Optionality / Free Upside. The reverse-DCF prices the in-production business; these are the call options you largely get for free at $425.83:
N2 / N1.4 leadership ramp — pricing and mix uplift from the next nodes is only partly in 2026 numbers and barely in out-years.
Advanced packaging (CoWoS / SoIC) — a structurally supply-short, high-margin layer expanding faster than wafers; under-modelled.
Geographic de-risking — Arizona / Japan / Germany fabs lower the single-island concentration discount over time (a re-rating lever, not a base case).
Treat as a tilt (+a few points, already in the 57), not a reason the core is cheap — the in-production business is fair-to-full.
5
Pillar Detail: Underlying Drivers
The dominant external force the stock is tethered to, scored 0–100 across historical / current / forward. A context pillar: a tailwind ≥65 can lift a base BUY to STRONG BUY (with a supportive economy); it does not change the fundamental pillar scores.
Primary Driver
AI / HPC + advanced-node foundry demand cycle
88
STRONG TAILWIND
Horizon
Weight
Score
Evidence
Historical (12–24m)
25%
85
AI accelerator demand drove revenue from 673B → 1,134B TWD/qtr; margins expanded through the period
Current state
50%
92
Advanced-node utilisation >90%; CoWoS packaging sold out; hyperscaler capex at records; N3/N2 ramping
Forward (6–12m)
25%
82
Consensus sees continued AI capex through 2026–27 + N2 ramp; some digestion risk; smartphone/auto recovering modestly
Amplification role: at 88 (Strong Tailwind, ≥65) the driver is eligible to lift a base BUY to STRONG BUY on any horizon where Economic Alignment's pressure is also a Tailwind — which is the case on the long horizon only. It does not alter the base BUY/HOLD/SELL or the three fundamental pillar scores.
Thesis-invalidation floor: a sustained cut in hyperscaler AI-capex guidance combined with advanced-node utilisation falling below ~80% would break the tailwind. The offsetting tail is geopolitical (Taiwan), handled in §2 as a caution.
6
Pillar Detail: Economic Alignment
Whether the latest DonatienInvestment macro report is a tailwind or headwind for this name, as a Tailwind/Neutral/Headwind pressure plus a Trend-Following or Contrarian stance. Its pressure is the second amplifier — a Tailwind enables STRONG BUY, a Headwind enables STRONG SELL.
Input (macro report 2026-06-13)
Short
Medium
Long
Technology sector (XLK)
Underperform
Underperform
Outperform
Emerging-Market equities (EEM — Taiwan)
Underperform
Neutral
Outperform
Resulting pressure
Headwind
Headwind (mild)
Tailwind
Stance: Contrarian near-term · Trend-Following long-term · Conviction 55. The macro regime is "Stagflation (oil shock + hawkish Fed)" — risk-off, with capital rotating to materials/gold/energy and away from high-multiple tech and EM in the near term (QQQ short "Strong Underperform"). Buying TSM today therefore fights the near-term macro (contrarian), but aligns with its long-horizon call that tech, EM and the AI compute-demand story recover (XLK & EEM long "Outperform").
Staleness caveat: the macro report (2026-06-13) predates the 2026-06-15 Iran-US peace deal, on which Asian tech surged — so the near-term risk-off framing may already be easing. The long-horizon signals (which drive the long-horizon Tailwind and the STRONG BUY) are unaffected. The Short = HOLD call below rests on technical extension, not on this macro headwind, so it holds either way.
7
Pillar Detail: Entry/Exit Timing
Why Timing scored 55: multi-timeframe trend, risk-reward framing, relative strength, macro overlay (at semis' medium sensitivity), sentiment from grade actions, and the 0–12 month catalyst cluster.
Entry/Exit Timing — Pillar Score
Constructive primary trend (monthly/weekly/daily all up) but extended and overbought — monthly RSI 79, ~90% of the 52-week range, and rolling over intraday (hourly & 15-min downtrends; −3.5% on the day). For the short horizon this classifies as Neutral timing, which is why the short-term read is HOLD rather than BUY.
55
Confidence 68%
Component
Weight
Score
Read
MTF trend
30%
70
Higher TFs bullish, intraday bearish — mostly bullish but pulling back
Risk-reward
20%
47
Near 52-wk high; support far below (SMA50 $400, swing $384); wide stop
Macro overlay (medium sens.)
15%
58
Hawkish Fed risk-off near-term; AI rotation supportive longer
Sentiment
18%
55
Grades all "maintain"; recommendation "strong buy"; estimates rising
Catalysts
17%
65
One clear catalyst (Q2 earnings ~mid-July); FOMC Jun 17
Relative strength: +106% over 52 weeks, outperforming both the S&P 500 and the semis group — leadership, but also a marker of how extended the move is. Why Short = HOLD: at Q88 / V57 / T55 the raw Decision Matrix (High + Fair + Improving) reads BUY on all three horizons. For the short horizon, Timing is 55% of the weight and the technical picture (overbought monthly RSI, 90% of range, intraday breakdown, only 1 of 5 entry criteria met) classifies as Neutral rather than Improving — pulling the Timing-dominated short read down into HOLD. The medium and long horizons, where Quality and the AI driver carry more weight, stay BUY / STRONG BUY.
8
Economic Event Risk
High-impact US macro releases in the next ~2 weeks plus recent surprises. Semiconductors carry medium macro sensitivity (PMI/manufacturing, trade/export-controls; growth-stock rate sensitivity), so the WAIT-override does not apply, but a hawkish FOMC into a risk-off tape is the near-term swing factor.
Date
Event
Impact
Forecast / Prev
Relevant?
Jun 17
FOMC rate decision + projections (Warsh's first)
High
Hold 3.75% (hawkish)
✅ Growth-stock multiples
Jun 17
US Retail Sales (May)
Med
+0.5% / +0.5%
⚠️ Demand read
Jun 25
Core PCE (May)
High
+0.2% / +0.2%
✅ Rate path
Jul 1
ISM Manufacturing PMI (Jun)
High
52.5 / 54.0
✅ Semis demand proxy
Jul 2
Non-Farm Payrolls (Jun)
High
70K / 172K
⚠️ Macro regime
Recent surprises (last 7 days): May core CPI MoM came in cool at +0.2% (vs +0.3%), but headline CPI YoY rose to 4.2% and PPI was hot (+1.1%) — sticky inflation keeps the Fed hawkish. Consumer sentiment ticked up (48.9). Net: a stagflationary mix that argues for near-term caution on high-multiple names — consistent with the Short = HOLD. The Jun 17 FOMC tomorrow is the immediate swing event; a hawkish surprise pressures growth/EM, a dovish tilt (the Warsh "go-ahead" thesis in the press) would be a near-term tailwind.
9
Multi-Timeframe Technical Analysis
Trend, RSI and breakout status from monthly down to 15-min, with the confluence verdict. Higher-timeframe levels are the strongest.
Classic higher-TF-bullish + lower-TF-pullback pattern: monthly/weekly/daily uptrends with hourly and 15-min turning down — a short-term pullback inside a larger uptrend. Both monthly and weekly RSI are overbought (79 / 74), so the pullback has room. The high-probability accumulation zone is the daily SMA20–SMA50 band (~$400–$420); the line in the sand is the $384 swing, then the $332 200-day.
10
Price Chart (6-Month Daily)
Daily structure with key moving averages and levels. Visual companion to §9 — the pullback, the support band, and the primary uptrend at a glance.
Level type
Price
Note
Stop / 200-day
$384 / $332
Swing-low support; daily SMA200 $332 is the deeper floor
Support band
$400–$420
Daily SMA50 $400, SMA20 $420 — accumulation zone
Current
$425.83
Just above the support band, −3.5% on the day
Resistance / 52-wk high
$438 / $450.16
Prior highs — the cap to break for continuation
Analyst median target
~$465–472
~9% above current
11
Scenario Summary
Bull / Base / Bear 12-month paths with triggers and probability weights. Stress-tests the AI-driver thesis against the cyclical and geopolitical tails.
Bull · 30% · ~$570 (+34%)
Q2/Q3 guidance raised; N2 + CoWoS ramp ahead of plan; gross margin holds >62%; AI capex re-accelerates. Multiple re-rates toward the AI-peer band. Target approaches the $600 analyst high.
Base · 50% · ~$480 (+13%)
Continued AI-led growth; ~22–24× forward multiple holds; earnings compound into the price. Lands around the analyst median ($465–472) with modest upside.
Bear · 20% · ~$345 (−19%)
AI-capex digestion / cyclical pause; risk-off + strong USD compresses EM/tech multiples; or a Taiwan geopolitical scare. Price retraces toward the $332 200-day. The tail (invasion/blockade) is larger but low-probability.
Probability-weighted 12-month expectation ≈ $467 (0.30·570 + 0.50·480 + 0.20·345), ~+10% from $425.83 — squarely the analyst-median zone, with the asymmetry skewed to the upside if AI capex holds and to a sharp (low-odds) drawdown on the Taiwan tail.
12
Entry / Exit Rules
Mechanical conditions for entry and exit. Entry criteria met: 1 of 5. Exit conditions live: 0 of 3.
Entry Rules — 1 / 5 met
METRule 1 (Value): Price < fair value $470 AND no earnings within 7 days AND driver ≥50 → all true at $425.83.
WAITRule 2 (Buy-the-dip): pullback into the $400–$420 SMA20/50 band with monthly/weekly uptrend intact and daily RSI 35–65 → price not yet in the zone.
WAITRule 3 (Momentum): daily close > SMA20 ($420) on volume >1.5× AND MACD histogram positive ≥2 days → volume only 0.9×, daily MACD histogram negative.
WAITRule 5 (Event clarity): after FOMC (Jun 17) and Q2 earnings (~mid-July) with guidance maintained/raised → events still ahead.
Exit Rules — 0 / 3 live
CLEARStop-loss: close below $384 support for 2 consecutive days → not triggered ($425.83).
CLEARThesis invalidation: FY guidance cut AND revenue decelerates below sector median AND a hard gate fires → guidance strong, none triggered.
CLEARProfit-take / trim: price reaches ~$465–472 median AND RSI > 70 AND quality hasn't improved to justify it → not reached.
Imagine you act at the current price $425.83 · as of Jun 16, 2026
What if you bought now?
You'd be risking ~$42 (−9.8%) to the hard stop to gain ~$54 (+12.7%) to the base target.
Risking: downside to stop $384 (−9.8%); bear case $345 (−19%); plus — entry rules NOT yet met: buying overbought, above the $400–420 entry zone, ahead of FOMC and Q2 earnings.
Gaining: base $480 (+12.7%) · bull $570 (+33.9%); plus the small dividend and the N2/CoWoS embedded optionality you'd own for free.
Net: risk-reward ≈ 1.3:1 to base, ~3.5:1 to bull — assessment only: a great business at a full near-term price; waiting for the $400–420 dip or a post-earnings confirmation improves the entry.
What if you sold now?
You'd be giving up base-case upside to ~$480 (+13%) to protect against the ~$345 bear drawdown (−19%).
Giving up: upside to $480 (+13%) and the long-run AI-compounding; selling slightly below fair value $470.
Protecting: capital if the bear case ($345) or a Taiwan scare plays out. Exit rules currently triggered? None.
Net: no mechanical reason to sell — this is a hold/accumulate zone, not a distribution one.
13
Position Sizing Context
Illustrative framework only — no allocation or portfolio role was specified for this batch run, so no position size is prescribed.
Risk per share to the $384 stop is ~$42 (~9.8%), or ~2.4× the daily ATR ($17.35) — a relatively wide stop, consistent with buying near highs. As a single-catalyst name with one clear event (Q2 earnings) inside 30 days, the catalyst calendar is calm enough for a normal position, but the overbought extension argues for scaling in (e.g. into the $400–420 zone) rather than a full entry at the high. The Taiwan tail is the reason this is typically held as a diversified semis/EM exposure rather than an outsized single-name bet. None of this is advice — size to your own plan and risk budget.
14
Calibration Snapshot
Machine-readable snapshot of every score, confidence, key level and signal override — saved alongside this report as calibration-TSM-20260616-1703.json for the next run.
Full audit trail of every data source used, with OK / partial indicators and any confidence haircuts. Per the design system, this lives at the bottom of the report.
Data Source Status
✓
get_company_profile / get_stock_snapshot— price $425.83, mkt cap $2.21T, ADR confirmed
Impact on scores: Data coverage is strong. Two haircuts: (1) Timing/catalyst confidence reflects an estimated Q2 earnings date — robust to any mid-July date, so the signal is safe; (2) Economic Alignment is built on a macro report 3 days stale on a material geopolitical event, so its near-term Headwind framing is treated as a basis and caveated. Neither moves the per-horizon ladder. Data-basis note: financials report in TWD on an ADR-equivalent share base (~5.19B units × 5 = ~25.93B ordinary shares); market cap $2.21T verified internally consistent — no stale-share-count trap.
DISCLAIMER: This is a quantitative framework for educational purposes only. It is not financial advice. Always do your own research and consult a licensed financial advisor before making investment decisions.