The market already pays ≈ Ennoconn's sum-of-parts. Raw look-through NAV (~NT$345–396) sits at the price; after a 0.12 governance haircut it falls below it. The growing, well-covered dividend (NT$14, ~4% yield, ~64% payout) underpins the downside but doesn't make it cheap here. No margin of safety after the +25% YTD rally to the analyst target. A genuine entry reopens in the high-NT$200s / low-NT$300s.
Ennoconn consolidates NT$142B revenue but the parent owns only ~38% of consolidated profit. The two biggest subs are separately listed — Kontron AG (Frankfurt, 27.9%) and Marketech (Taipei, 38.2%). Valuing consolidated FCFF then subtracting the minority at book (NT$36.5B) while the listed minorities are worth ~NT$93B at market flatters the parent. Deconsolidation reveals the punchline: of NT$3.21B parent earnings, ~NT$2.7B is the Marketech + Kontron stakes — the core IPC "stub" earns only ~NT$0.5–0.9B. Ennoconn is largely a holding vehicle for two listed stakes.
| Component | NT$M |
|---|---|
| Marketech stake @ market (38.2% × NT$92B) | 35,144 |
| Kontron stake @ market (27.9% × €1.44B × 36.4) | 14,624 |
| Deconsolidated stub (parent NI 544 × 14×) | 7,614 |
| Equity-method cross-holdings (book) | 5,356 |
| Less: parent-level net debt (est.) | −5,000 |
| Look-through CGT treatment | Raw NAV/sh | Gov-adj ×0.88 | vs NT$360 |
|---|---|---|---|
| Gross stakes (no CGT) | NT$396 | NT$348 | +10% / −3% |
| Net 20% CGT on gain (base) | NT$347 | NT$305 | −4% / −15% |
| Net 20% on full value (conservative) | NT$328 | NT$288 | −9% / −20% |
Gordon value DPS×(1+g)/(r−g). DPS NT$14, sustainable g ~4.2% (ROE 11.6% × 36% retention), CoE ~6.9% at β=0.85.
| r \ g | 3.0% | 3.5% | 4.0% | 4.5% |
|---|---|---|---|---|
| 6.5% | 412 | 483 | 582 | 731 |
| 6.9% | 370 | 426 | 502 | 610 |
| 7.5% | 320 | 362 | 416 | 488 |
| 8.0% | 288 | 322 | 364 | 418 |
Supportive — the dividend the minority actually receives is reasonably-to-modestly cheap, not overvalued. This is the real downside protection.
Risks: parent net-debt estimate ±NT$5B ≈ ±NT$34/sh; stub multiple ±2× ≈ ±NT$10/sh; EUR/TWD ±5% ≈ ±NT$5/sh; marks inherit the cycle that suppressed the regression beta.