← back to picks

HQ Taichung · Taiwan Reporting TWD Credit synth Aaa/AAA · net cash Valuation 2026-06-01 Damodaran FCFF · 1000-iter MC FCFF · Dark v3

Cheap across every rate regime · +116% margin of safety

Cheap at base AND the 4.3% high bookend — a robust buy, not a rate bet

Intrinsic value NT$347/share (after a 5% governance haircut; pre-haircut NT$365) vs market NT$160.50 — +116% at the methodology-correct 2.65% TWD risk-free. The Monte Carlo p5 still sits at NT$299 (+86%): every one of 1,000 correlated draws — now stress-testing a sub-peak 15% operating margin — lands above today’s price, so P(intrinsic<market)=0%. Topkey trades at 10.8× P/E on 2025 EPS of NT$14.92 and a record 35.3% gross margin, struck at a volume trough.

SectorCarbon-fiber compositesCountry mixUS 40% · DE 20% · JP 17%β / MC σ0.90 levered · ±NT$26/sh (1000 runs)GovernanceFounder-controlled · 5% haircutQuality35.3% gross margin · net cash · synth AAAIncomeNT$8 dividend · ~5% yield · 54% payout

Intrinsic / share

NT$346.54

post 5% gov · pre NT$364.78

Market / share

NT$160.50

2026-06-01 close · TWSE

Margin of safety

+115.9%

vs intrinsic

Enterprise value

NT$29.52B

69.2% terminal

Cost of equity / debt

6.79% / 2.44%

β 0.90 · CRP 0.37%

Terminal ROIC / g

10.00% / 2.00%

spread ~360bp (ROIC 10.00% vs WACC 6.40%)

What it sells, where it sells

Operating segments

Bicycle frames~30% global share of composite frames — at a destocking trough, channel back to 2-3 months~47%

Aerospace / medicalBoeing/Airbus seat structures, multi-year backlog; Vietnam plant 2 ramps Q4 2026~25%

RacketsPickleball/padel-led, growing +30% YoY — fastest, highest-margin line~15%

Electronics / ToF + helmetsToF components into the largest US phone brand, +30% YoY; helmets +10%~13%

Rackets and ToF electronics (~28% combined) are already growing >30% and offsetting the bicycle (~47%) and aerospace (~25%) cyclical drag — so the 2025 record 35.3% gross margin is a mix-and-pricing story struck while the two largest lines sit at a volume trough, not a peak.

Country mix (revenue-weighted CRP input)

🇺🇸United States40%

🇩🇪Germany20%

🇯🇵Japan17%

🇳🇱Netherlands8%

🇹🇼Taiwan8%

🇨🇳China7%

US + Germany + Netherlands ≈ 68% of revenue, so the demand cycle is a Western-consumer (bike/racket) and aerospace-OEM story, not a China story — but earnings are highly TWD-sensitive: the 2025 revenue dip was entirely currency (flat in USD), and a strong TWD compresses the reported numbers the market anchors on.

Background — five things to know before the case

Context the share price doesn't carry on its face. Skim this once and the bull / bear bullets below stop reading as inside baseball.

Record earnings struck at a volume trough, not a peak. 2025 revenue NT$9.20bn (−2.9% YoY, but flat in USD — the dip is pure TWD appreciation), yet gross margin hit a record 35.31% and EPS a record NT$14.92, while the two largest end-markets (bicycles ~47%, aerospace ~25%) were both contracting (−10%+ and −20%+).
~30% global share of composite bicycle frames; the destocking is ending. Channel inventory has normalized to a healthy 2-3 months and management plus sell-side see a 2026 H2 "spring" recovery. April 2026 revenue of NT$861m returned to a normal run-rate; the 4-month YoY decline narrowed to −7.6%, with institutions forecasting Q3 turning to positive growth.
Founder family controls the board with zero pledged shares. The Shen family (chairman Shen Wen-chen 10.63%, president Shen Bei-ni 2.12%) controls the company; directors and supervisors carry no share pledges — a clean control structure underwriting the 5% governance haircut.
Three high-margin growth axes already compounding >30%. Rackets (pickleball/padel) +30%+, electronics +30%+ (ToF components into the largest US phone brand across phone/watch/earbuds), helmets +10%+ — collectively offsetting the bicycle and aerospace cyclical drag. Vietnam plant 2 begins aerospace-seat volume production from Q4 2026.
Net cash, ~54% payout, plus free optionality. NT$8 dividend at >5% yield, ~54% payout, with humanoid-robot (waist/knee/shoulder parts) and drone components in early talks, and a planned spin-off IPO of Xiamen subsidiary 新鴻洲 (precision molds/injection/electronics) as soon as 2027 — none of which is in the base case.

Where we diverge from the sector

Each row asks: if we'd used the sector median for this one assumption instead of our override, how much would the share price change? Negative = our override is more conservative (worth less); positive = more aggressive (worth more). Skim down the rightmost column to see which bets are doing the work.

Assumption

Our input

Sector median

Value impact

01Revenue growth10-year revenue CAGR vs Recreation median

~4.5% CAGR6% Y1-5 (destocking recovery) fading to 2.0% terminal — no robot/drone/IPO upside

~4.0% CAGRRecreation, global cross-sector median

+NT$12

02Operating marginYear-5 target vs sector median EBIT margin

18.0%Ex-peak through-cycle level (FY24 19.8% / FY25 18.7%); re-centered from 20% to 18% per the 2026-06-01 council

~12%Recreation sector median operating margin

+NT$55

03Sales-to-capitalReinvestment efficiency vs sector median

1.21×Capital-intensive composites manufacturing (autoclaves, Vietnam plant build-out)

~1.5×Recreation industry standard

−NT$14

04Terminal growthYear 10+ steady state vs TWD risk-free ceiling

2.0%TWD long-run inflation; held ≤ the 2.65% risk-free (Damodaran ceiling)

2.65%risk-free ceiling (65/35 blend); terminal g held below it

NT$0

05Cost of capital10y WACC vs WACC implied by sector-median β

6.40%β 0.90 (sector-aware, anti-suppression) · rf 2.65% (65/35 blend) · ERP+CRP · D/V low single digits

~5.3%Recreation sector β re-levered → ~1.05 → higher WACC

+NT$18

Net effect of overrides

Overrides net positively vs all-defaults — the work is done by holding a through-cycle 18% operating margin (re-centered from 20% per the council) rather than reverting to the ~12% sector median, partly offset by a conservative, capital-intensive sales-to-capital that bakes in the Vietnam plant build.

+NT$72

Our override Sector median Override adds value Override subtracts value

The story & the five claims

The 10-year story this DCF is built on, plus the five anchor claims that translate the story into engine inputs.

The 10-year story

Topkey is the carbon-fiber composites leader the cycle is hiding — roughly 30% of every composite bicycle frame on earth, plus aerospace seat structures, pickleball rackets, and electronic components, made at a record 35.3% gross margin even while its two biggest markets sit at a volume trough. The central debate is whether 2025’s record earnings are a mid-cycle peak the cheap multiple correctly prices, or a trough the market is mistaking for a peak. The base case is simply that bicycle destocking ends (channel already back to 2-3 months) and the fast-growing racket and electronics lines keep compounding. Revenue grows about 6% a year through year five and then fades, compounding from NT$9.2bn today to roughly NT$14.3bn by year ten. Operating margin holds at 18% — the ex-peak through-cycle level (FY24 19.8% / FY25 18.7%, already ~17% in Q1-26), re-centered down from an earlier 20% after an independent council flagged that 20% anchored on the cyclical high. Steady-state growth past year ten is pinned at Taiwan’s ~2.0% long-run inflation, held below the 2.65% risk-free, and the whole valuation is taken after a 5% haircut for founder control. The result survives every rate regime: +178% at the 1.13% local-bond floor, +116% at the 2.65% base blend, and still +68% at a normalized 4.3% rate.

Two debates worth pressure-testing

Is 2025’s record margin a mid-cycle peak the cheap P/E correctly prices?

Our view: No — the record 35.3% gross margin and NT$14.92 EPS were struck while bicycles (−10%+) and aerospace (−20%+) were contracting, the opposite of a peak. We still re-centered the terminal operating margin from 20% to the ex-peak through-cycle 18% (FY24 19.8% / FY25 18.7%); even at a bear 16% the base is +94%, and the normalized 4.3%-rate stress still lands NT$269 — +68% above today’s NT$160.50.

Does the low-rate TWD discount flatter the whole valuation?

Our view: This is the key test, and Topkey passes it. The risk-free was re-rated from the suppressed 1.13% local bond to the methodology-correct 2.65% blend (65% inflation-parity / 35% bond), lifting WACC to 6.40%. The BUY does not hinge on the low-rate regime: at a normalized 8.0% stress WACC it is still +68%, and the MC p5 (NT$299) sits +86% above price.

CLAIM 01Revenue compounds ~6% Y1-5 as destocking ends, fading to 2.0%.growth_high: 6.0% · terminal: 2.0% · NT$9.2B → NT$14.3BBicycle channel already back to a healthy 2-3 months; rackets and ToF electronics already +30%. Excludes robot/drone and the 新鴻洲 spin-off optionality.

CLAIM 02Operating margin holds at 18% — the ex-peak through-cycle level.target_op_margin: 18% from Y5FY24/25 op margin ran 19.8%/18.7%; the ex-2022-peak through-cycle is ~18%. Re-centered from 20% to 18% per the council — prices durable mix/pricing power without anchoring on the cyclical high.

CLAIM 03Reinvestment stays capital-intensive through the Vietnam build.S2C: 1.21× Y1-5 and Y6-10Composites need autoclaves and tooling; Vietnam plant 2 ramps aerospace seats from Q4 2026. A conservative 1.21× — below the sector — is a drag we accept, not a stretch.

CLAIM 04Margin and growth converge by Year 5 (2030).year_of_convergence: 5Recovery and margin maturation complete by year 5; no further expansion in years 6-10, with growth fading linearly to the terminal rate.

CLAIM 05No failure risk — net cash, founder-controlled, no distress tail.terminal_g: 2.0% · failure: 0% · gov haircut: 5%Net cash, ~54% payout, synth Aaa/AAA, zero pledged founder shares. 5% governance haircut for founder control; terminal ROIC faded to 10% from the engine's ~19%.

Where we diverge from sell-side

Operating margin held at 18% (through-cycle), not reverted to the ~12% sector median. The market's 10.8× P/E implicitly treats record 35.3% gross margins as a peak; we treat them as durable mix/pricing power struck at a trough — but re-centered the terminal to the ex-peak ~18% (not 20%) after the council. This call still carries most of the gap to NT$347.
Earnings read as trough, not peak. 2025 EPS NT$14.92 was a record despite bicycles −10%+ and aerospace −20%+; the cheap multiple anchors on the optical −2.9% revenue dip that was entirely TWD appreciation (flat in USD).
β set to 0.90 by sector-anti-suppression, not the 0.10 regression. The raw regression β is implausibly suppressed (thin-trade small-cap artifact); using 0.90 lifts WACC honestly to 6.40% (at the 2.65% rf) rather than flattering the valuation with a near-zero β.
Governance haircut 5%, terminal ROIC faded to 10%. Founder control with zero pledged shares earns a light 5% haircut; terminal excess returns faded from the engine's ~19% to 10% to avoid an unfaded perpetual-moat assumption.
Base case excludes all optionality. Humanoid-robot/drone parts and the 新鴻洲 spin-off IPO are real but unquantified — none is in the NT$347; they are free upside, not load-bearing.

Two-sided case — bear anchors

This is mid-cycle, not trough. If 2025's record margins reflect a cyclical mid-point and bicycles/aerospace don't recover, the cheap multiple is fair. The 8.0% stress WACC + the 18% margin still lands NT$269 — a drawdown from base but +68% above price.
TWD strength keeps compressing reported earnings. 2025's decline was entirely currency; a structurally strong TWD shrinks the reported NT$ numbers the market anchors on, even with flat USD volumes. Roughly a 1-for-1 hit to reported revenue and margin.
Low-rate dependence on the terminal block. ~69% of value is terminal; the risk-free is already re-rated to the 2.65% blend (WACC 6.40%) and a further move toward 4.3% only compresses the base to +68% — it does not break the thesis.
Raw-material cost inflation. Higher oil lifts carbon-fiber and chemical feedstock costs; without matching price increases, the record gross margin compresses. A 2-3pp margin hit ≈ NT$50-60/share.
Optionality stays optionality. Robot/drone parts are early-stage talks with zero revenue; the 新鴻洲 IPO is a 2027+ event. If none converts, the thesis must stand on the core composites cycle alone — which it does.

Risks to thesis (tail, not bear case)

Bicycle/aerospace recovery stallsMed

If the 2026 H2 "spring" recovery doesn't materialize and inventory re-builds, the trough extends. ~15-20% equity drag; cyclical, not structural — destocking is already largely complete.

TWD appreciationMed

Earnings are highly TWD-sensitive — the 2025 revenue decline was purely currency. A persistently strong TWD compresses reported revenue and margin 1-for-1 even at flat USD volume.

Raw-material cost spikeMed

Higher oil prices push carbon-fiber and chemical feedstock costs up. Without matching price increases, the record 35.3% gross margin erodes — roughly NT$50-60/share on a 2-3pp hit.

Rate normalizationLow

~69% of value is the terminal block. The risk-free is already at the methodology 2.65% blend (WACC 6.40%); the explicit 4.3%/8.0%-WACC stress still clears +68%.

Optionality fails to convertLow

Humanoid-robot/drone parts (early talks, no revenue) and the 新鴻洲 spin-off (2027+) may never land. Pure upside omission — the base case credits none of it.

Founder-succession / controlLow

Shen family controls the board (chairman 10.63%, president 2.12%, zero pledged shares). Clean today; key-person concentration is a long-tail governance risk, already priced via the 5% haircut.

10-year forecast

Revenue NT$9.76B → NT$14.69B over 10y; operating margin to 18.00% from Y5 onward.

Monte Carlo distribution

Even at the 5th-percentile outcome (NT$299/sh), intrinsic value exceeds today's NT$160.5 price by 87% — across 1,000 correlated stress draws not one lands below the market.

Mean NT$342.96 ± NT$26.50/sh, 1000 iterations (0 failed). P(intrinsic < market NT$160.50) = 0.0%.

Cost of capital build

Risk-free rate	2.65%
Mature-market ERP	4.19%
Levered β	0.90
Weighted CRP	0.37%
Cost of equity	6.79%
Pre-tax cost of debt (synth Aaa/AAA)	3.05%
D / V	~9%
WACC	6.40%

Full year-by-year DCF

Year	Revenue	Op mgn	EBIT	EBIT(1−t)	Reinvest	FCFF	PV
1	NT$9.76B	18.57%	NT$1.81B	NT$1.40B	NT$456M	NT$948M	NT$891M
2	NT$10.34B	18.43%	NT$1.91B	NT$1.48B	NT$484M	NT$993M	NT$877M
3	NT$10.96B	18.28%	NT$2.00B	NT$1.55B	NT$513M	NT$1.04B	NT$864M
4	NT$11.62B	18.14%	NT$2.11B	NT$1.63B	NT$544M	NT$1.09B	NT$851M
5	NT$12.32B	18.00%	NT$2.22B	NT$1.72B	NT$576M	NT$1.14B	NT$838M
6	NT$12.96B	18.00%	NT$2.33B	NT$1.82B	NT$529M	NT$1.29B	NT$889M
7	NT$13.53B	18.00%	NT$2.43B	NT$1.91B	NT$471M	NT$1.44B	NT$932M
8	NT$14.01B	18.00%	NT$2.52B	NT$1.99B	NT$402M	NT$1.59B	NT$967M
9	NT$14.41B	18.00%	NT$2.59B	NT$2.06B	NT$324M	NT$1.74B	NT$991M
10	NT$14.69B	18.00%	NT$2.64B	NT$2.12B	NT$238M	NT$1.88B	NT$1.01B

Methodology & flags

Damodaran FCFF DCF, 10y explicit + perpetuity, in TWD. Risk-free 2.65% (local-currency government bond). Synthetic credit Aaa/AAA. CRP 0.37% from revenue-weighted country mix × Damodaran 2026 CRPs. Terminal ROIC faded to 10.00%; 5% governance haircut applied post-DCF. Monte Carlo: 1000 iterations. Engine v1.0.0 · result: /Users/valentin/Documents/notes_jiliac_labs/Finance/Damodaran/valuations/topkey/output/2026-06-01-result.json

R&D cap: OFF · Lease cap: OFF · Failure: OFF · ESO: OFF
Governance haircut: 5% applied post-DCF (NT$364.78 > NT$346.54)
Sensitivity tornado: not run

FCFF DCF · dark theme v3. Generated 2026-06-01T10:08:57 from /Users/valentin/Documents/notes_jiliac_labs/Finance/Damodaran/valuations/topkey/output/2026-06-01-result.json.