← back to picks

HQ Kaohsiung · Taiwan Reporting TWD Credit strong · int. cover ~16× Listed TWSE 1988 Valuation 2026-06-04 Damodaran FCFF · 1000-iter MC FCFF · Dark v3

A P/E 6.3 mirage · the “cheap” earnings are a 金居 mark-to-market windfall, not the cable business

Headline P/E 6.3 looks like deep value; strip the investment windfall and it is ~2× its intrinsic value

The screen flagged Hua Eng cheap on FY2025 EPS NT$5.50 (P/E 6.3) — but that EPS is dominated by unrealized mark-to-market gains on the company’s ~3% stake in 金居/Co-Tech (8358.TW), an AI copper-foil momentum stock. Core operating EPS is ~NT$1.5–2.0, and Q1 2026 confirmed it: record revenue +22% YoY but EPS reverted to NT$0.37. On core operations the stock trades at ~18.6× EV/EBIT — fully valued. Normalize the peak margin to the company’s own 5-yr average and value the trapped non-operating assets honestly and intrinsic is NT$14.21 (post-governance; NT$17.77 pre-governance). Three independent methods — FCFF DCF, Monte Carlo (P(intrinsic<market)=100%), and a cash-to-minority Gordon DDM (~NT$18) — all land NT$14–19 against the NT$34.50 price. The hidden NAV (金居 stake + land) is real but trapped behind a controlling reciprocal cross-holder with a 15-year record of not monetizing it. SELL / PASS.

SectorElectrical Equipment (wire & cable)Country mixTaiwan 86% · China 12%β / MC σ1.0 levered · ±NT$1.8/sh (1000 runs)GovernanceReciprocal cross-hold · 0.20 haircutQualitycore ROIC ~7% ≈ WACC · headline ROE 24.5% = windfallIncomeNT$1.0 div · ~2.9% yield · windfall-funded

Intrinsic / share

NT$14.21

core value NT$14–19 · three methods agree · vs NT$34.50 price

Market / share

NT$34.50

4 Jun 2026 · TWSE

Margin of safety

-58.8%

vs intrinsic

Enterprise value

NT$8.92B

75.7% terminal

Cost of equity / debt

7.40% / 2.17%

β 1.00 · CRP 0.86%

Terminal ROIC / g

7.20% / 2.00%

spread ~84bp (ROIC 7.20% vs WACC 6.36%)

What it sells, where it sells

Operating segments

Wire & cable (operating)Power cables, comm/optical cables, oxygen-free copper. 1 of ~5 TW firms certified for 345kV ultra-high-voltage cable; single dominant customer = Taipower. Core EBIT margin ~8% (cyclical peak) vs ~6% 5-yr average100%

Trapped non-operating NAV~NT$3.9bn book: 金居/Co-Tech stake (FV ~1.6bn, at a momentum peak) + 39.4% of First Copper + Far EasTone + ~1.0bn investment property + a 13.2-ha Kaohsiung land plot un-developed for 15 yrs~27% of mcap

Hua Eng is a single-franchise wire & cable maker — the donut is whole-company revenue. The number that distorts everything is off the income statement: a ~NT$3.9bn stack of non-operating assets, dominated by a ~3% stake in 金居/Co-Tech whose AI-driven quadrupling was marked through FY2025 net income. That windfall, not the cable business, is what made the screen’s P/E 6.3 look cheap. The cable franchise itself earns roughly its cost of capital.

Country mix (revenue-weighted CRP input)

🇹🇼Taiwan86%

🇨🇳China12%

🌏SE Asia / other2%

Overwhelmingly a Taiwan story — domestic grid (Taipower) and fab/construction demand, with a ~12% China export tail. The thesis rides the Taiwan grid-capex cycle; copper is a commodity-price input (pass-through, margin-diluting when it spikes), not a country-risk exposure.

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.

FY2025 EPS NT$5.50 is a mark-to-market mirage. Reported net income NT$2,335M / ROE 24.5% is dominated by unrealized gains on Hua Eng’s ~3% stake in 金居/Co-Tech (8358.TW), an AI copper-foil momentum stock that quadrupled. Strip it and core operating EPS is ~NT$1.5–2.0. Q1 2026 confirms: record revenue +22% YoY, EPS reverted to NT$0.37. The screen’s P/E 6.3 was computed on windfall earnings.
On core operations the stock is fully valued, not cheap. EV/EBIT ~18.6×, EV/EBITDA ~15× on the ~NT$14.6B market cap. Operating EV in the DCF is only NT$8.9B; the market implies ~NT$14.8B for the cable ops alone.
The hidden NAV is real but trapped. ~NT$3.9bn book of non-operating assets (金居 stake + 39.4% of First Copper + Far EasTone + ~NT$1.0bn investment property + a 13.2-ha Kaohsiung land plot rezoned 2011 with a paper NT$10bn plan). The land has been un-monetized for 15 years (FY2024 disposals NT$5.9m); buybacks are held-not-cancelled; the dividend is windfall-funded and volatile.
Cash-to-minority cross-check (the mandated holdco test) confirms the DCF. A Gordon DDM on the actual NT$1.00 dividend at the 7.4% TWD cost of equity gives NT$13.5 (g=0%) / NT$18.9 (g=2%) / NT$23.4 (g=3%). The cash a minority actually collects is worth ~NT$18, not NT$34.50 — the NAV is a ceiling, not a target.
Controlling reciprocal cross-holder + no catalyst. First Copper owns 49.2% of Hua Eng (a 208.56M-share block, treated as treasury-equivalent → economic share count 424.21M vs 632.77M issued); Hua Eng owns 39.4% of First Copper back. No Value-Up plan, no activist, no cross-holding unwind. Fails the controlled-company two-condition viability test → trap, not target.

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 / company ref

Note

01Revenue growthY1-5 vs Electrical Equipment median

5% Y1-5Taipower grid capex + copper pass-through; no AI-hype credit, fading to 2.0% terminal

~4.5%cross-sector median

generous

02Operating marginYear-7 target vs company’s own history

6.0%Faded DOWN from the 8.16% cyclical peak to the company’s own 5-yr average (8.11/6.42/5.10/3.17/7.49%)

~4% thru-cyclecompany’s own wire&cable through-cycle norm

already generous

03Sales-to-capitalReinvestment efficiency vs industry

1.3×Capital-intensive cable mfg; current operating S2C ~1.1

1.76×Electrical Equipment industry

conservative

04Terminal growthYear 10+ vs TWD risk-free ceiling

2.0%TWD long-run inflation; ≤ the 2.31% risk-free (Damodaran ceiling)

2.31%rf ceiling (50/50 blend)

at ceiling

05Cost of capital10y WACC · β triangulated

6.36%β 1.0 (lifted off a suppressed 0.78 regression) · rf 2.31% · low net debt

β_u 1.34Damodaran-global re-lever → 1.60 (overstates)

middle

Why the ladder still says SELL

Every row here is set generous-to-neutral — 5% growth, margin faded only to the 5-yr average (not the ~4% through-cycle), terminal growth at the rf ceiling, a low 6.36% WACC. Even so, intrinsic is NT$17.77 pre-governance, ~48% below the NT$34.50 price. The maximally-generous bull bookend — hold the 8% peak margin forever, zero governance discount — still only reaches ~NT$25. There is no defensible combination of these inputs that reaches the market price.

NT$14–18

Our input Sector / company ref Generous to the bull Conservative

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

Hua Eng is a Kaohsiung wire & cable maker that screened cheap at a P/E of 6.3 — but the cheapness is an accounting artifact. FY2025 EPS of NT$5.50 is dominated by an unrealized mark-to-market gain on the company’s ~3% stake in 金居/Co-Tech, an AI copper-foil stock that quadrupled; core operating EPS is ~NT$1.5–2.0, and the very next quarter EPS reverted to NT$0.37. On its core operations the cable business trades at ~18.6× EV/EBIT — fully valued — and earns roughly its cost of capital, so it creates little value. Normalize the peak operating margin to the company’s own 5-year average, lift the suppressed beta to 1.0, and credit the trapped non-operating assets (haircut), and intrinsic is NT$17.77 pre-governance, NT$14.21 after a 0.20 governance discount for the controlling reciprocal cross-holder. Three independent methods agree: the FCFF DCF (NT$14–18), the Monte Carlo (every one of 1,000 draws below market), and a cash-to-minority Gordon DDM on the actual NT$1.00 dividend (~NT$18). All sit at roughly half the NT$34.50 price. The hidden NAV — the 金居 stake and a 13.2-hectare Kaohsiung land plot — is real, but it is trapped behind a reciprocal ownership loop with a 15-year record of not monetizing it, so it is a ceiling a minority never collects, not a catalyst. An LLM council pressure-test confirmed the SELL, noting it survives the two best bull arguments. This is the value-trap inverse of a bargain: a headline that looks like deep value precisely because a momentum-stock windfall is flowing through reported earnings.

Two debates worth pressure-testing

Isn’t fading margin from 8% to 6% too harsh during a real grid supercycle?

Our view: It doesn’t change the verdict. The company’s own 5-yr average op margin is 6.05% (history 7.49/3.17/5.10/6.42/8.11%), and Q1 2026 showed the +22% revenue did not hold margin — net margin compressed to 4.6% as copper inflated revenue. Even granting the bull fully — hold the 8% peak forever and zero governance discount — intrinsic is only ~NT$25, still −28% vs NT$34.50.

If the earnings are a mirage, why is the market paying NT$34.50?

Our view: Reflexivity on a thin, un-covered small-cap. As 金居 rises, Hua Eng’s reported EPS and book value rise, attracting momentum flows into a stock with ~102M real float and 0 analyst coverage. The market is paying for the windfall and the paper NAV — neither of which a minority can extract through the reciprocal structure. “Only I see it” is a fair caution, but a mark-to-market windfall flowing through EPS is a well-understood way small-caps misprice.

CLAIM 01The P/E 6.3 is a windfall mirage — core EPS is ~NT$1.5–2.0.FY25 EPS 5.50 → core ~1.5-2.0 · Q1’26 EPS 0.37FY2025 net income is dominated by unrealized gains on the ~3% 金居 stake. Q1 2026 reverting to NT$0.37 (annualizes ~NT$1.5) ≈ confirms the core. The screen ranked windfall earnings.

CLAIM 02Core operations are fully valued at ~18.6× EV/EBIT.operating EV NT$8.9B · core op margin ~8% (peak)A cyclical, copper-input, single-customer cable maker at a full multiple on peak margins. Normalized DCF intrinsic on the ops alone is well below the implied price.

CLAIM 03Three independent methods all land NT$14–19.DCF 14.2/17.8 · MC p50 14.1 · DDM ~18FCFF DCF, Monte Carlo (P(below market)=100%), and a cash-to-minority Gordon DDM on the real NT$1.00 dividend independently triangulate to roughly half the NT$34.50 price.

CLAIM 04The hidden NAV is a ceiling, not a target.~NT$3.9bn book non-op · land un-monetized 15 yrs金居 stake + land + stakes are real but trapped behind a reciprocal cross-holder. No Value-Up plan, no activist; FY2024 land disposals were NT$5.9m. Fails the controlled-company viability test.

CLAIM 05632.77M issued, but 424.21M is the economic share count.First Copper 208.56M reciprocal block = treasury-equivEPS, BVPS and every ownership % compute on 424.21M. Using 632.77M would make it more overvalued — the SELL is robust to the share-count choice.

Where we diverge from sell-side

An LLM council confirmed the SELL. A 5-advisor pressure-test (2026-06-04) affirmed the call and strengthened it: the verdict survives the two best bull arguments (margin fade → bull bookend still −28%; governance discount → pre-gov still −48%), and the peer-review surfaced the cash-to-minority cross-check that independently confirms ~NT$18.
The cable business creates ≈ zero value. Core ROIC ~7% ≈ WACC, so the operating franchise is worth roughly book. The entire investment case is the trapped non-operating assets — which a minority cannot extract.
The NAV is real but un-collectible. 金居 stake (at a momentum peak), 39.4% of First Copper, Far EasTone, investment property, and a 13.2-ha land plot — all credited at a haircut in the DCF, yet intrinsic is still NT$17.77 pre-governance. Crediting them at full paper value to a minority who can’t force monetization is the trap.
Margin normalized to the 6.05% 5-yr average, not held at the 8.16% peak. Even this generous fade (the through-cycle norm is ~4%) doesn’t close the gap.
Governance haircut 0.20 for the reciprocal structure — but the SELL does not depend on it (pre-governance is already −48%).

Two-sided case — bear anchors

The 金居 stake is a double-edged windfall. It inflated FY2025 EPS, book value, and the cross-holdings mark simultaneously — and a revert from its quadrupled peak would deflate all three at once. The market is paying for a peak it treats as permanent.
Single customer, single geography. ~86% Taiwan revenue anchored on Taipower; the grid order book is real but operational, not a value-release catalyst, and copper-price spikes pass through revenue while compressing margin (Q1 2026: +22% revenue, net margin to 4.6%).
Capital allocation is minority-unfriendly. Buybacks held-not-cancelled (~NT$969m), a volatile windfall-funded dividend (NT$1.5/0.4/1.0/1.0 over 2022-25), and a 15-year refusal to develop or sell the Kaohsiung land. Cash does not reliably reach outside holders.
The reciprocal cross-holding loop concentrates control. First Copper 49.2% ↔ Hua Eng 39.4% of First Copper — a circular structure that insulates the board and routes value internally, a textbook related-party leakage channel.
No catalyst with teeth. Taiwan has no Korea-style Value-Up mandate; there is no activist, no holdco/opco merger, no disclosed cross-holding unwind. Absent a catalyst, the trapped value stays trapped and the stock de-rates as the 金居 windfall fades.

Risks to thesis (tail, not bear case)

Asset monetization catalystWatch

A genuine sale/JV of the 13.2-ha Kaohsiung land or a cross-holding unwind would release trapped NAV and could flip the thesis. But 15 years of non-monetization (FY2024 disposals NT$5.9m) and no announced plan make this a low-probability watch-item, not a base case.

Sustained UHV margin step-upWatch

If the 345kV ultra-high-voltage mix structurally lifts margins to 9-10% and holds through-cycle, the operating value rises — but even an 8% perpetual peak only reaches ~NT$25, below the price. Would need 9-10% margins and higher growth and asset release simultaneously.

金居 stake re-rates higherMed

A further leg up in 金居/Co-Tech would re-inflate reported EPS and book value, sustaining momentum flows and the mispricing for longer. This is a reason the overvaluation can persist, not a reason it is justified — and it reverses violently if 金居 falls.

Copper / grid order surpriseMed

Taipower’s NT$564.5bn resilient-grid plan could push volumes above the conservative 5% growth. But copper spikes dilute margin on pass-through, and the order book is operational momentum, not capitalizable economic profit above WACC.

Long-only: SELL = PASSLow

In a long-only pipeline the actionable output is simply to remove it from the buy queue. The thin float, reciprocal control block, and zero coverage make it impractical to short — its value is as a published value-trap counter-example, not a short position.

10-year forecast

Revenue NT$11.33B growing 5% Y1-5 fading to 2.0% terminal; operating margin faded from the FY25 8.16% cyclical peak to the company’s own 5-yr-average 6.0%. This is the generous normalized path — and it still produces intrinsic ~NT$14–18, roughly half the NT$34.50 price.

Monte Carlo distribution

The Monte Carlo spans NT$11.4 (p5) to NT$17.2 (p95) and every one of 1,000 draws sits below the NT$34.50 price — P(intrinsic < market) = 100%. The overvaluation is robust to the entire joint distribution of growth, margin, terminal-growth and governance inputs. Independently, a cash-to-minority Gordon DDM on the actual NT$1.00 dividend lands ~NT$18 — the same neighbourhood from a completely different method.

Mean NT$14.18 ± NT$1.79/sh, 1000 iterations (0 failed). P(intrinsic < market NT$34.50) = 100.0%.

⚠ Active diagnostic: stable_state.override_roc (0.0624) < WACC (0.0636); every dollar reinvested in stable state destroys value

Cost of capital build

Risk-free rate	2.31%
Mature-market ERP	4.23%
Levered β	1.00
Weighted CRP	0.86%
Cost of equity	7.40%
Pre-tax cost of debt (synth Aaa/AAA)	2.71%
D / V	~20%
WACC	6.36%

Full year-by-year DCF

Year	Revenue	Op mgn	EBIT	EBIT(1−t)	Reinvest	FCFF	PV
1	NT$11.90B	7.85%	NT$934M	NT$747M	NT$436M	NT$311M	NT$292M
2	NT$12.49B	7.54%	NT$942M	NT$753M	NT$458M	NT$296M	NT$262M
3	NT$13.12B	7.23%	NT$949M	NT$759M	NT$480M	NT$278M	NT$231M
4	NT$13.77B	6.92%	NT$954M	NT$763M	NT$505M	NT$258M	NT$202M
5	NT$14.46B	6.62%	NT$957M	NT$765M	NT$530M	NT$236M	NT$173M
6	NT$15.18B	6.31%	NT$958M	NT$766M	NT$556M	NT$210M	NT$145M
7	NT$15.94B	6.00%	NT$957M	NT$765M	NT$584M	NT$181M	NT$118M
8	NT$16.58B	6.00%	NT$995M	NT$796M	NT$491M	NT$305M	NT$186M
9	NT$17.08B	6.00%	NT$1.02B	NT$820M	NT$383M	NT$437M	NT$250M
10	NT$17.42B	6.00%	NT$1.05B	NT$836M	NT$263M	NT$573M	NT$308M

Methodology & flags

Damodaran FCFF DCF, 10y explicit + perpetuity, in TWD. Risk-free 2.31% (local-currency government bond). Synthetic credit Aaa/AAA. CRP 0.86% from revenue-weighted country mix × Damodaran 2026 CRPs. Terminal ROIC faded to 7.20%; 20% governance haircut applied post-DCF. Monte Carlo: 1000 iterations. Engine v1.0.0 · result: /Users/valentin/Documents/notes_jiliac_labs/Finance/Damodaran/valuations/huaeng/output/2026-06-04-result.json

R&D cap: OFF · Lease cap: OFF · Failure: OFF · ESO: OFF
Governance haircut: 20% applied post-DCF (NT$17.77 > NT$14.21)
Sensitivity tornado: not run

FCFF DCF · dark theme v3. Generated 2026-06-04T14:34:12 from /Users/valentin/Documents/notes_jiliac_labs/Finance/Damodaran/valuations/huaeng/output/2026-06-04-result.json.