Backlog-floored cyclical, net-cash, deeply mispriced · +61.2% margin of safety
Market far below MC p5The post-governance DCF lands at ₩58,318/sh against a ₩22,600 market price; even the Monte Carlo 5th percentile (₩46,872) sits ~100% above today. A thin-margin but #1-in-niche, net-cash cleanroom contractor riding ₩70tn of 2026 Samsung/SK Hynix fab capex, still priced as plain construction.
What it sells, where it sells
Operating segments
~85% of revenue is the percentage-of-completion derivative of Korean memory-fab construction starts — so the margin and growth debate is really a Samsung/SK Hynix capex-cycle debate, with HBM specialty-gas piping the mix upside and structural EPC thinness the floor.
Country mix (revenue-weighted CRP input)
Near-pure Korea: assets, labour, financing and end-customer demand are all won-based, so the weighted CRP lands at 0.74% (essentially Korea's Aa2 0.64% + a 3% Vietnam-proxy nudge). True economic exposure to the Korean chip cycle is arguably closer to 100% than 97%.
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.
- Priced as construction, behaves as semiconductor infrastructure. Classified KOSDAQ "construction" with 0-1 sell-side analysts, Sebo trades ~7x P/E vs 19-48x semicap peers despite a 2021-26 revenue CAGR ~11%. A Korea IR Council note (Feb-2026) argued for re-rating the whole cleanroom-infra cohort.
- Record backlog is a hard forward floor. Order book hit an all-time-high ₩757.1bn at Q3-2025 — above a full year of FY2025 revenue (₩715.8bn) — with >₩200bn signed in Sept-2025 alone, including ₩136.8bn of SK EcoPlant Yongin contracts running to 2027-05-31.
- Demand is two Korean chaebol building fabs. ~85% of revenue is tied to Samsung and SK Hynix, who plan ~₩70tn combined capex in 2026; SK Hynix Yongin phase-2 breaks ground Aug-2026 and Samsung P5 cleanrooms expand 4→6, lifting cleanroom area ~50%.
- Net-cash fortress with rare pro-minority capital allocation. Borrowings cut ₩62bn (2020) → ₩2bn; cash & ST investments ₩148.8bn. A 10-year rising dividend (₩200→₩600), repeated buybacks and a permanent treasury-share cancellation (May-2026) — unusually clean for Korea.
- The bear flag is an earnings-quality question, not distress. FY2025 EBIT rose +41% to ₩40.4bn yet 2H operating cash flow went negative on a ~₩41bn accounts-receivable build; one analyst flagged "value trap — dividend paid from the cash pile." Read here as POC-ramp working capital given the net-cash balance sheet.
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.
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
Sebo is a thin-margin, asset-light, net-cash cleanroom contractor — Korea's #1 in first-stage fab piping and ~60% of the specialty fluororesin-duct market — whose revenue is the one-to-two-year echo of Samsung and SK Hynix memory-fab construction starts. This is not a growth compounder; it is a cyclical riding, not expanding, its incumbent share. Off a three-year-trough ₩715.8bn base, a record ₩757bn backlog and ₩70tn of 2026 chip capex re-accelerate revenue back toward the prior-cycle ₩1tn peak by year five, before a normal downcycle gives some of it back — netting to ~7% annual growth that fades to a mature 3% by year ten. Operating margin holds the now-demonstrated ~6%, and because reinvestment per won of new revenue is deliberately set conservatively, the cash the model credits is defensible rather than balance-sheet-flattered.
Two debates worth pressure-testing
- We charge for the AR build the bulls ignore. Sales-to-capital cut 2.5→1.5 (ramp) after the FY2025 ~₩41bn receivables swing — a deliberate FCFF haircut, not a cash-flattered optical.
- Terminal ROC pinned to 10%, not the optical 28%. The net-cash pile flatters reported ROIC to ~28%; we let only a modest ~150-200bp spread over the ~7.5% Korean WACC into perpetuity.
- Governance haircut sized to the track record, not the stereotype. 10% — above a clean 5% for the ~48-52% family control, below 15-20% given treasury cancellation and a decade of rising dividends.
- We model a ramp-then-give-back, not a new permanent base. The FY2022 +138% spike faded three years; revenue peaks near ₩1tn at Y5 then mean-reverts, vs the bull "₩910bn compounds."
- A 10-15% chip-capex cut flows straight to revenue. ~85% of demand is two buyers; a Samsung/SK Hynix pullback the next year would push revenue back toward the FY2025 trough and compress the ramp.
- The AR build is a value-trap tell, not WC timing. If 2H-2025 negative operating cash flow persists, the dividend really is funded from the cash pile and the FCFF case erodes.
- Margin re-acceleration dilutes mix. A volume up-leg historically pulls in lower-margin first-stage piping; if the 6% slips back toward ~5%, terminal value falls materially.
- Korea-discount overhang resolves badly. A founder-family controls ~48-52%; if minority treatment reverts to the regional norm, more than the 10% haircut leaks out of the per-share number.
Risks to thesis (tail, not bear case)
~85% of revenue rides Samsung + SK Hynix fab capex; a synchronized memory-capex deferral hits revenue the following year with little offset.
FY2025 2H operating cash flow turned negative on a ~₩41bn receivables build; if it does not unwind, the earnings-quality flag chips at the FCFF thesis.
A strong volume up-leg historically pulls in lower-mix first-stage piping; ~100bp of margin give-back meaningfully lowers terminal value.
The 3.78% risk-free is cyclically elevated by an oil shock; a fall to ~3.0-3.3% would lower WACC and raise value — a tail to the upside, modeled separately in the stress case.
0-1 sell-side analysts and KOSDAQ small-cap liquidity mean the re-rating, even if right, can stay un-arbitraged for an extended period.
10-year forecast
Revenue ramps ₩765.9B → ₩1.00T by Y5 (back near the prior-cycle ₩1tn peak on a record ₩757bn backlog), then mean-reverts to ₩1.26T by Y10. Operating margin holds the demonstrated mid-cycle ~6.0%; FCFF inflects sharply from Y6 as reinvestment intensity eases.
Monte Carlo distribution
Even at the 5th-percentile outcome (₩46,872), intrinsic value exceeds today's ₩22,600 price by 52% — across 1,000 correlated draws not a single path lands below the market. The debate is the size of the gap, not its sign.
1000 iterations (0 failed) drawing revenue growth, operating margin, terminal growth, ramp-year sales-to-capital and the governance discount. Mean ₩55,434 ± ₩5,062/sh. P(intrinsic < market ₩22,600) = 0.0%.
Cost of capital build
| Risk-free rate | 3.78% |
| Mature-market ERP | 4.00% |
| Levered β | 0.76 |
| Weighted CRP | 0.74% |
| Cost of equity | 7.58% |
| Pre-tax cost of debt (synth Aaa/AAA) | 3.26% |
| D / V | ~0% |
| WACC | 7.54% |
Full year-by-year DCF
| Year | Revenue | Op mgn | EBIT | EBIT(1−t) | Reinvest | FCFF | PV |
|---|---|---|---|---|---|---|---|
| 1 | ₩765.9B | 5.71% | ₩43.7B | ₩33.6B | ₩33.4B | ₩237M | ₩220M |
| 2 | ₩819.5B | 5.78% | ₩47.4B | ₩36.4B | ₩35.7B | ₩708M | ₩612M |
| 3 | ₩876.9B | 5.86% | ₩51.3B | ₩39.5B | ₩38.2B | ₩1.2B | ₩1,000M |
| 4 | ₩938.3B | 5.93% | ₩55.6B | ₩42.8B | ₩40.9B | ₩1.9B | ₩1.4B |
| 5 | ₩1.00T | 6.00% | ₩60.2B | ₩46.3B | ₩43.8B | ₩2.5B | ₩1.8B |
| 6 | ₩1.07T | 6.00% | ₩64.0B | ₩49.3B | ₩34.6B | ₩14.8B | ₩9.5B |
| 7 | ₩1.12T | 6.00% | ₩67.4B | ₩52.1B | ₩32.0B | ₩20.2B | ₩12.1B |
| 8 | ₩1.18T | 6.00% | ₩70.5B | ₩54.7B | ₩28.7B | ₩26.0B | ₩14.5B |
| 9 | ₩1.22T | 6.00% | ₩73.2B | ₩56.9B | ₩24.8B | ₩32.1B | ₩16.6B |
| 10 | ₩1.26T | 6.00% | ₩75.4B | ₩58.8B | ₩20.3B | ₩38.5B | ₩18.4B |
Methodology & flags
Damodaran FCFF DCF, 10-year explicit forecast plus perpetuity, in
KRW (₩ millions). No R&D or lease capitalisation (asset-light
EPC/MEP contractor); net-cash balance sheet drives a synthetic
Aaa/AAA credit and near-zero debt weight. Country risk premium is
revenue-weighted across South Korea (97%) and a Vietnam overseas
proxy (3%). Monte Carlo: 1000 iterations over five correlated
inputs. Engine v1.0.0 · result:
/Users/valentin/Documents/notes_jiliac_labs/Finance/Damodaran/valuations/sebo/output/2026-05-31-result.json
- R&D cap: OFF · Lease cap: OFF · Failure: OFF · ESO: OFF
- Governance haircut: 10% applied post-DCF (₩64,798 > ₩58,318)
- Sensitivity tornado: not run · uncertainty captured via 5-axis Monte Carlo