Nexon Co., Ltd.3659.T
Nexon's gross margin fell from 77.3% in FY2018 to 59.4% in FY2025 while revenue rose 87%. The market reads the 18-point compression as unit-economic decay of the Dungeon&Fighter / Korea rent and has marked the stock down 41.5% year-to-date. The Korea segment carries the entire group operating profit — Korea OP was ¥158.9bn in FY2023 against group OP of ¥137.7bn. SOTP fair value reconstructs to ¥2,446 base ; spot is ¥2,307. The diagnostic is the FY2026 consolidated gross margin and whether the legacy rent margin holds ex-Embark.
The legacy Korea rent — Dungeon&Fighter, the China royalty through Tencent, MapleStory — carries ~¥128bn of FX-normalised AOP. At 8.6x EV/AOP, legacy EV is approximately ¥1,103bn.
The Embark horizontal leg carries ¥110bn of revenue at 1.0x and no profit credit, or ¥110bn.
Net cash is ¥834bn. Long-term securities are ¥146bn.
Consolidated market cap at spot is ¥1,829bn. Net cash alone is 46% of it.
The dossier rests on one question. Is the 18-point gross margin compression from FY2018 to FY2025 the decay of the Dungeon&Fighter / Korea rent unit economics, or a mix effect from the integration of a new-IP leg sold on third-party platforms at a ~30% commission against legacy proprietary channels at ~75% gross margin. The SOTP moves between ¥1,732 and ¥3,185 on which answer holds. The legacy multiple, the cash haircut and the horizontal optionality all hang off it.
The consensus reading takes the compression as unit-economic. The eroding engine prices a degrading rent, and at a 1.67x price-to-book — the decade floor — the stock encodes the China royalty in secular decline and the cash as permanently dormant.
The variant reading decomposes the compression by channel. Holding unit margins constant by pole, the consolidated gross margin still falls from 59.2% to 56.0% over five years on the rising horizontal weight alone, from 21.4% to 29.0% of revenue. The FCF/EBIT ratio at 132% in FY2025 confirms development is expensed rather than capitalised, so the consolidated margin already absorbs the full content cost and the cash is real. The residual unit component, if any exists, is not isolable without segment disclosure Nexon does not publish.
The FY2026 consolidated gross margin closes the debate empirically. A print at or above 59% with a rising horizontal share confirms the mix reading. A print below 57% with legacy gross margin ex-Embark falling sequentially confirms unit erosion and recompresses fair value toward ¥1,732.
Position framing is monitoring at current levels, gated by the FY2026 margin print and the DnF franchise revenue trajectory. The weighted asymmetry of +3% does not support a position at this level. The bias is long ; the queue of permanent-loss risk holds it on the watchlist. Conviction is moderate.
The eleven-year window from FY2015 to FY2025 reads as three sequential regimes. A high-margin PC-online apogee through FY2019, with gross margin near 74–77%, EBIT margin near 38%, and the DnF China royalty as the core. A diversification phase from FY2020 to FY2023 — COVID demand, the Embark acquisition in 2021, new mobile titles — with revenue scaling 1.5x and gross margin compressing from 76% to 67% as growth was bought at the cost of margin. A de-rating phase from FY2024, with the DnF Mobile China peak of FY2024 normalising 25% lower in FY2025, gross margin at 59.4%, and the stock down 41.5% year-to-date.
| Inflection | FY2015PC apogee | FY2018Peak margin | FY2022Diversification | FY2024China peak | FY2025Engine eroding |
|---|---|---|---|---|---|
| Revenue (¥bn) | 190.3 | 253.7 | 353.7 | 446.2 | 475.1 |
| Gross margin | 73.9% | 77.3% | 70.1% | 63.1% | 59.4% |
| EBIT (¥bn) | 62.3 | 98.4 | 103.7 | 124.2 | 124.0 |
| EBIT margin | 32.7% | 38.8% | 29.3% | 27.8% | 26.1% |
| FCF (¥bn) | 57.3 | 116.3 | 127.3 | 97.3 | 163.9 |
| FCF margin | 30.1% | 45.9% | 36.0% | 21.8% | 34.5% |
| Net cash (¥bn) | 286.8 | 479.4 | 560.7 | 610.3 | 834.0 |
| EPS basic (¥) | 63.9 | 121.0 | 114.7 | 161.8 | 114.5 |
| DPS (¥) | 5.0 | 0.0 | 10.0 | 22.5 | 45.0 |
Source: Data pack 31 May 2026, workbook nexon_3659.xlsm. December year-end. Per-share series adjusted for the 1-for-2 stock split of 28 March 2018. Net cash = total cash and short-term investments less total debt. FY2024 = year ended 31 December 2024.
Net cash rose from ¥479bn in FY2018 to ¥834bn in FY2025, an increase of ¥355bn across seven years through ¥98bn of buybacks in FY2025 alone — the machine generates more cash than it returns. The allocation ramp is recent. The dividend doubled to ¥45 in FY2025 and is guided to ¥60 in FY2026, with a ¥100bn buyback completed and ¥30bn in execution. Reported ROE at 8.9% sits far below the ex-cash operating economics, the gap being the dormant balance sheet inflating the denominator. The Shareholder alignment pillar at 2.5/5 captures this.
Korea is the company ; the rest is an option cost. Neople, the wholly-owned Korean developer of Dungeon&Fighter, books the China royalty Tencent pays, so the China geography carries almost no operating profit despite 23% of revenue. The DnF rent plus MapleStory cover the structural losses of every other geography. The relevant unit is the franchise by channel, never the consolidated line — the 59.4% consolidated gross margin is the average of legacy proprietary channels near 75% and the Embark horizontal leg structurally lower after the platform commission.
The switching costs of the legacy franchises are real and old. Dungeon&Fighter launched in 2005, MapleStory in 2003 ; progression investment and the social graph hold the base across two decades. The China royalty was extended ten years with Tencent, who now also takes the DnF Mobile development with Neople retaining creative control. The moat is concentrated on one or two IP, and the new-IP leg has no proven moat in a contested F2P shooter. ARC Raiders has passed 16m units ; its live-service durability past the launch spike is unproven.
The critical cost is the distribution structure ; the model carries no commodity input. The ~30% commission on Steam and console for the Embark titles and the Tencent share on the China royalty move the gross margin, and the compression coincides with the mix sliding toward third-party-distributed, higher-development-cost titles. Monetisation runs on live-ops engagement on captive bases plus a premium-and-pass hybrid on the new-IP, with no list-price power. Reported net income is unusable — the FY2025 figure fell 32% mainly on lower FX gains — so the earnings power reads off operating profit, not the bottom line.
FCF margin was 34.5% in FY2025 on capex below 2% of revenue, the best cash conversion in the sub-industry. The FCF/EBIT ratio of 132% reflects development expensed rather than capitalised, with a favourable working-capital float from deferred F2P revenue. The model's own clean NOPAT chain converts at roughly 63%, below the 132% historical that the deferred-revenue float will not perpetuate ; the normalised FCF yield sits at 4.3% against a 9.0% spot, and that normalised yield covers the cost of equity. The cash floor anchors the downside. Net cash at ¥834bn is ~¥1,052 per share, 46% of the market cap, and the Bear fair value of ¥1,732 rests on it plus a depreciated residual rent.
The ex-cash operating economics are first-rate — FCF margin 34.5%, FCF/EBIT 132%, ROIC ex-cash near 40% against a cost of capital near 6–7%, capex below 2% of revenue. The gross margin trajectory of −18 points and the reported ROIC crushed to 8% by the dormant balance sheet weigh against. This is where mix-versus-unitary is decided. If the legacy unit margin holds, the grade is conservative ; if the erosion is unitary, it is generous. The cardinal gap is that the disclosure does not let the segment margin by channel be read directly.
Net cash of ¥834bn, 78% of equity, has never re-rated the stock. The allocation ramp is recent — the dividend doubled to ¥45 then guided to ¥60, buybacks of ¥100bn plus ¥30bn, one-share-one-vote, TSE Prime. The structural overhang sits above the listed entity: NXC and NXMH control ~46–49%, and the Korean government holds ~30.65% of the unlisted NXC holding through an inheritance-tax payment after the founder's death in 2022, with repeated disposal attempts failing. The cash dormancy haircut moves fair value ±¥124 per share — the balance sheet weighs nearly as much as the operating multiple.
Bi-modal. MTX recurrence on 20-year franchises ; horizontal +188% YoY. Offset by Korea-concentrated profit, the DnF reset at −26% franchise revenue in Q1, and unproven new-IP life.
Switching costs on DnF and MapleStory ; the China royalty secured ten years with Tencent. Concentrated on one or two IP ; the new-IP leg has no proven moat in F2P shooters.
The Embark bet produced a hit at scale. The 2026 cost discipline — headcount frozen, three projects cancelled — and candid DnF communication are positive. The DnF reset came late and segment disclosure is thin.
Above Koei Tecmo (14/25, hidden value gated by 51.92% family control), below Capcom (19/25, proven catalogue annuity) and Konami (18/25, re-rated core). Comparable to Square Enix (16.5/25, net-cash floored on declining revenue) — Nexon shares the cash floor but carries a real revenue trajectory. The grade does not justify a premium on a consolidated basis. It justifies separating the legacy rent, the horizontal optionality and the cash into three economies. The cellular SOTP reading.
Is the 18-point gross margin erosion of channel mix or of unit economics ?
Is ARC Raiders a launch spike or a third rent pillar ?
The horizontal leg grew 188% YoY to ¥59.3bn, about 39% of Q1 2026 revenue, with ARC Raiders past 16m units. Embark is wholly owned and globally consolidated, not equity-method, so the diversification is real and on the P&L. Whether it is structural rests on retention past the launch front-load. The base case credits it as a drag in optionality at 1.0x revenue ; the bull case converts it to a rent and swings the SOTP by ~¥335bn — the single largest source of upside, larger than the legacy re-rating.
Is the ¥834bn cash dormant or in re-rating by allocation ?
Net cash is 78% of equity and 46% of the market cap and has never re-rated. The allocation ramp accelerates into the drawdown — the dividend up 33%, buybacks of ¥100bn plus ¥30bn — while the Korean government's unsold 30.65% of the NXC holding leaves the ultimate control structure undetermined. The dormancy haircut runs 12–35% across scenarios, a ±¥124 swing per share, making the cash as much a driver of fair value as the operating multiple.
At ¥2,307 spot, 1.67x price-to-book at the decade floor and ~7.1x EV on normalised AOP, the market prices a concentrated Korea rent in erosion with the cash dormant. Two anchoring choices do the work. Reported net income — down 32% in FY2025 mainly on lower FX gains — is read through, which makes a low P/E a false value signal until the FX is neutralised. The 18-point gross margin compression is treated as unit-economic rather than channel mix. Neither anchor survives the cellular SOTP that values the legacy on FX-normalised AOP and credits the cash and horizontal separately.
Legacy gross margin ex-Embark falls — unit erosion confirmed — with DnF in secular decline despite Tencent and ARC Raiders failing retention. AOP normalised ¥112bn at a 6.4x de-rated legacy multiple, horizontal at 0.3x, cash haircut 35% on capture. The downside is timing bounded by the cash ; permanent loss requires the rent broken and the cash mis-allocated together.
Korea roughly flat with MapleStory offsetting DnF, China in managed decline, the horizontal ramping without yet carrying profit, the erosion of mix not unit, the return sustained. AOP normalised ¥128bn at 8.6x legacy, horizontal 1.0x on revenue, cash haircut 25%. The cellular SOTP delivers ¥2,446 ; a consolidated re-rating is not required for it.
DnF stabilises and re-grows under Tencent, ARC Raiders proves durable live-service retention and becomes a third rent pillar, transformation opex plateaus, the cash is returned. AOP normalised ¥139bn at 10.0x legacy, horizontal 2.0x, cash haircut 12%. The horizontal conversion supplies most of the ¥335bn swing over the legacy re-rating.
| KPI | Latest value | Status | What it tells us |
|---|---|---|---|
| Consolidated gross margin | 59.4% (FY2025) | Cardinal | The mix-versus-unitary pivot. A FY2026 print at or above 59% with rising horizontal share confirms the base case ; below 57% with legacy margin ex-Embark falling sequentially triggers the bear case. |
| DnF China franchise revenue | −26% (Q1 2026) | Trigger | The cardinal rent. Revenue below −30% over two consecutive quarters despite Tencent signals secular decline and switches the dossier to bear at ¥1,732. |
| ARC Raiders MAU / retention | >16m units | Trigger | The bull optionality. Stable retention past "Frozen Trail" (October 2026) confirms the third pillar ; MAU below −40% after T+2 marks the horizontal to 0.3x. |
| OP by geographic segment | Korea ¥158.9bn FY23 | Watch | The cardinal gap. Last published FY2023. Re-disclosure would let the legacy rent margin be read directly rather than reconstructed by proxy. |
| Capital return executed | ¥60 DPS · ¥30bn BB | Asymmetric | Total return sustained above 60% of FCF compresses the dormancy haircut toward 12% ; suspension re-imposes the 35% capture haircut. |
| FCF / EBIT (hist. / normative) | 132% / 63% | Asymmetric | Development expensed makes EBIT understate the cash machine. Normalised FCF yield 4.3% covers cost of equity and floors the valuation by cash-flow. |
| FX KRW / JPY | ~0.107 | Reference | The operational translation runs through KRW/JPY, with USD/JPY secondary. The legacy AOP carries a 5% normalisation haircut in base ; drift above 10% revises it. |
A FY2026 consolidated gross margin below 57% with legacy gross margin ex-Embark falling sequentially confirms unit erosion and reclassifies the dossier toward Trap, compressing fair value toward ¥1,732. DnF franchise revenue below −30% over two consecutive quarters despite the renewed Tencent partnership signals secular decline of the cardinal rent and acts as the second-order bear confirmation.
ARC Raiders sustaining retention past the "Frozen Trail" update converts the launch spike into a third rent pillar and supplies the ¥335bn SOTP swing. Total shareholder return sustained above 60% of FCF with a rising payout reprices the cash from dormant to optionality and compresses the dormancy haircut from 25% toward 12%. Either confirms the upside leg the watchlist is waiting on.
A value-destructive acquisition above ¥200bn at ROIC below cost of capital, or a structural Chinese regulatory reset on Dungeon&Fighter, burns the only tangible floor and forces complete re-underwriting that this initiation does not attempt. Currently not signalled.
The information provided on this website is for informational and educational purposes only and should not be construed as financial, investment, legal, or tax advice. All content reflects the personal opinions, interpretations, and analyses of the author at the time of writing and is subject to change without notice. Nothing contained herein constitutes, or should be interpreted as, a recommendation, solicitation, or offer to buy or sell any securities, financial instruments, or other investment products. The author is not a licensed financial advisor, broker, or investment professional. Any references to specific assets, markets, or strategies are illustrative in nature and do not constitute personalized investment advice. Investing in financial markets involves risk, including the potential loss of capital. Past performance is not indicative of future results. Readers are solely responsible for their own investment decisions and should conduct their own independent research and due diligence before making any financial commitments. You are strongly encouraged to consult with a qualified financial advisor, legal professional, or other relevant specialist before making any investment or financial decisions. By accessing and using this blog, you agree that the author shall not be held liable for any direct or indirect losses, damages, or consequences arising from the use of, or reliance on, the information presented herein. All content is provided "as is" without any warranties of completeness, accuracy, or reliability.