376300 — Dear U Co., Ltd.

Dear U Co., Ltd. (KOSDAQ: 376300) — deep-value analysis

Analysis date: 28 August 2026 · Reference price: KRW … (close 27.08.2026, Yahoo 376300.KQ) · Shares: 23,738,406 · Market cap: KRW 446.0 bn · Net cash: KRW 176.7 bn · Enterprise value: KRW 269.4 bn · Single currency: KRW for both reporting and price (no conversion risk in the metrics).

Symbol note. The bare symbol 376300 is empty on Yahoo, and 376300.KS returns a ghost quote (no issuer name, no reporting currency, price 25,100 vs …). The data pack generated at 02:02 came out completely empty for this reason. Dear U is listed exclusively on KOSDAQ; primary filings are at DART/KIND, not SEC EDGAR — step 2 (EDGAR) doesn’t apply and wasn’t run.

Executive summary

The thesis, briefly. Dear U operates “bubble”, the app through which K-pop fans pay ~KRW 5,000/month to receive private messages from their favorite artists. It’s a remarkable business on paper: an operating margin of 40.4% over the trailing four quarters, a 2025 ROIC of 65.8% (Samsung Securities estimate), zero financial debt, KRW 176.7 bn of net cash — i.e. 39.6% of market cap is cash. The stock fell from KRW 80,836 (December 2021) to KRW …, … over four years and eight months, and … just from the June 2025 peak (61,024). At 7.2x EV/EBIT, the market says this business is in structural decline.

The analysis’s central discovery. The research brief, the GBL tracker, and the mechanical data pack all converge on the same flow figure: TTM FCF ≈ KRW 37 bn, an 8.3… yield. The figure is arithmetically correct and economically misleading. Dear U capitalizes advance payments to talent agencies (“MG” — minimum guarantee, the minimum guarantee paid to obtain the right to monetize an artist) as advance assets, then impairs them below the operating line, as a non-operating item. The Samsung Securities report of 5 August 2026 explicitly states, in note 2 to the estimates table: “in Q2 2025, large-scale impairments of MG costs prepaid to artists were reflected as non-operating expenses, per conservative asset-valuation criteria (US subsidiary KRW 7.1 bn, headquarters KRW 4.8 bn)” — KRW 11.9 bn. And it repeated: the same report attributes the Q2 2026 consensus miss to the same cause — “impairment of MG costs prepaid to artists, reflected as a non-operating expense”.

The consequence is that operating profit never contains the real cost of acquiring artist IP, and 2025’s CFO is inflated twice over: once by adding back an 11.9 bn non-cash impairment, and again because in 2025 the company barely paid any new advances (the change in advances in CFO: +0.20 bn in 2025, versus −21.52 bn in 2023 and −14.40 bn in 2024). In Q1 2026 the cycle restarted: −2.46 bn in a single quarter.

The FCF bridge. Interest is reported in operating (K-IFRS, “Interest Received CFO” 5.29 bn, “Interest Paid CFO” −0.20 bn) — so there’s no AFYA trap here; the divergence between simple FCF and FCFE is 2.2%, below the 10% threshold. The problem is elsewhere and the mechanical bridge doesn’t see it: interest from the cash pile and the MG cycle. I normalized both (details in the valuation chapter) and arrived at owner earnings of … mn KRW, ex-cash-interest (cash is added separately), versus 33,935 mn of mechanical FCFE — a 34% gap.

Estimated value. Five models, a wide range: from … (bear DCF) to … (bull DCF), median …. EPV Greenwald, the harshest test, gives …. Monte Carlo over 20,000 scenarios: median …, P10 …, P90 …, undervaluation probability …%, but probability of a margin of safety above 30% only …%.

Verdict: DO NOT BUY at KRW …. Not because the business is bad — it isn’t — but because, under assumptions I can defend with documents, the price is roughly fair, not cheap. A stock at 7.2x EV/EBIT with a 40% margin and 40% cash in market cap looks like a market mistake; once you normalize the IP-acquisition cost and accept that domestic subscribers are falling (2.01 mn in Q1 2026 → 1.97 mn in Q2 2026), the margin of safety evaporates. The zone where the thesis turns asymmetric is below KRW 14,500 (where even EPV gives a positive value). To watch: the Q3 2026 report — if a third consecutive MG impairment appears, the earnings-quality thesis breaks for good; if the S1 TME/JOOX royalties, deferred to Q3, come in without a new impairment, it’s worth re-evaluating.