264450 — Ubiquoss Inc.

Ubiquoss Inc. (유비쿼스, KOSDAQ: 264450) — deep-value analysis, REFRESH

Analysis date: September 17, 2026 · Reference price: … (tracker, 09.17.2026) · Shares: 14,665,493 · Market cap: … (~…) · Reporting currency = price currency = KRW (FX = 1)

Regime: REFRESH. The reference analysis is rapoarte/deep/2026-08-24-deep-264450.md. The mechanical deep_delta.py triage requested a refresh for three unverifiable falsifiers (pb, pe, crestere_venit_yoy — metrics yfinance doesn’t serve for a .KQ-suffixed symbol), not for a price move. The price went … → … over 24 days (…); that’s context, not the reason for the refresh, and the thesis below isn’t built around it.

What I explicitly inherit, since unchanged: the business description and moat structure; the control genealogy (the 03.01.2017 spinoff, the loss of legal holding-company status in April 2021, the Nexite / Ubiquoss Investment structure); the 2017–2025 dividend series adjusted for splits; the 2022–2025 annual statements; the FCF-bridge methodology and the 2025 fiscal-year bridge; the owner-earnings base of …; the g1 = 5%, r = 11%, gt = 1% assumptions.

What gets re-derived, since the new filing moved the facts: everything touching Q2 2026 — balance sheet line by line, margins, inventory, receivables, earnings quality, net debt, access to cash (λ), the Monte Carlo simulation, the five valuation models, red flags, the verdict.

The new primary source, which didn’t exist at the prior analysis: the DART semi-annual report 반기보고서 (2026.06), filing number 20260814001072, filed August 14, 2026 — full consolidated statements, the inventory and financial-instrument notes, the ownership table, the shareholder-return policy, and the product-sales table. The August analysis declared this report “not found publicly”; it had been public for ten days. I correct the error and record it: the document wasn’t missing, the DART search was.

Executive summary

The thesis, in one sentence. Ubiquoss remains a profitable Graham net-net — … against an NCAV of …/share — but the refresh shifts the thesis’s center of gravity: the operating business, valued alone, with zero credit for the … of net cash, is worth …/share, i.e. … versus price. The thesis no longer depends on the treasury unlocking; unlocking has become optionality.

The three new facts justifying the re-evaluation.

First: the cycle has turned in figures, not promises. Q2 2026 delivered revenue of …, … year over year (DART H1), the best quarter since Q4 2023, with a gross margin of 49.8% — the maximum in the entire available series — and an operating margin of 25.2%. For the half-year: revenue 58,618 mil (…), operating profit 12,667 mil (…), net profit 17,262 mil (…). The test the August report explicitly set — “if Q2 2026 revenue exceeds …” — was passed.

Second: there’s now a declared, dated, and quantified return-of-capital policy. On February 3, 2026, via a fair-disclosure communication repeated in the semi-annual report, the company announced that for 2026–2028 the shareholder-return reserve is 40% of adjusted consolidated net profit, through cash dividends and buybacks. The August report wrote that “the payout ratio remains under 30%” and treated the pivot as an inference from a reserve transfer. It’s no longer inference. It’s this refresh’s most important piece of information, and it’s the only documented anchor I use for calibrating λ.

Third, and it’s negative: profit has stopped coming from the business. Of the … of pretax profit for the half-year, 5,550 mil (26.5%) is net gains from fair-value revaluation of a securities portfolio (fair-value gains 5,587.8 mil minus losses 37.6 mil, DART H1), versus 428 mil in H1 2025. The fair-value note lists … of level-2 financial assets — 46,729 fund units, 13,420 government and municipal bonds, 7,939 debt securities, 2,524 hybrid capital securities, 1,456 unlisted equity. More than half of what the tracker calls “cash” is a market portfolio, not bank deposits.

The updated balance-sheet figure. As of 06.30.2026: cash and equivalents 71,658 + other current financial assets 58,096 = 129,754 mil; total debt 2,825 (of which 1,825 lease obligations and 1,000 bank loan); provisions 5,636. Adjusted net cash = … mil = …/share = 80.2% of price. Adding the 36,407 mil of non-current financial assets, total net liquid assets reach 157,700 mil = …/share = 104.3% of price, and enterprise value becomes negative, −12,136 mil. On the conservative basis, EV = 24,272 mil = 0.89x TTM EBIT of 27,338 and 0.80x EBITDA.

Earning power. The FCF bridge over the last four quarters: CFO … − capex … = simple FCF 20,786; minus lease principal … = 19,191; minus interest received … and dividends received 545 = operating FCFE … mil. Divergence versus CFO−capex: …. The TTM figure is below the normalized base for an identifiable reason — the 19,215 mil inventory build in H1. The three normalization routes (4-year flow 17.42 / 4-year-window NOPAT 18.71 / 3-year NOPAT 17.07) still bracket …, so the base stays unchanged.

Estimated value — a range, not a point. Five triangulated models give margins of safety between … and …, median …. Monte Carlo across 20,000 scenarios, with λ = 0.40 applied to net cash: P10 …, median …, P90 …, undervaluation probability …%, probability of margin above 30% at …%. The prior simulation, without λ, gave a median of … and …% undervaluation probability — a degenerate distribution that hid exactly the risk the August report called “the modal scenario.”

What worsened. Net inventory rose to 44,932 mil, … in six months on revenue …; DIO jumped from 141 days (2025) to 184 (Q1 26) to 249 days (Q2 26). The inventory provisioning rate fell from 14.35% to 9.68% of gross value. Half-year operating cash flow was negative, −1,155 mil, and cumulative 4.5-year profit-to-cash conversion fell from 1.03x to 0.86x — the prior report’s best defense has eroded.

Verdict: MAINTAINED — SPECULATIVE BUY, SMALL POSITION. The reason has changed, though. In August, the thesis was “cash locked up with governance optionality.” Today, the thesis is “operating business undervalued 43% even if the cash never reaches you, plus a 40% distribution policy that’s just been put on paper.” The position stays small for three reasons that don’t show up in percentiles: turnover of ~…/day, a quarter of profit coming from mark-to-market on a securities portfolio, and 249 days of inventory that will be tested in Q4.