Square Enix Holdings9684.T
The Digital Entertainment core prints a 25.1% segment business-profit margin at FY March 2026, on 58% of revenue and 79% of group business profit. The consolidated 18.4% EBIT margin is the residual after ¥18.6bn of unallocated corporate cost, on a revenue base down to ¥297.7bn from the ¥365.3bn FY March 2022 peak. Net cash is ¥274bn — 29% of the ¥940bn market cap and 79% of book equity. SOTP fair value reconstructs to ¥2,400–2,500 ; spot is ¥2,558. The diagnostic is the durability of the Digital Entertainment margin across the FY March 2027 prints, and whether the ¥100bn three-year capital envelope is deployed into returns or into M&A.
Digital Entertainment at a ~24% normalised segment business-profit margin on ¥173bn of revenue produces ~¥40bn of segment business profit. At 12x the segment carries an EV of approximately ¥480bn.
Amusement, Publication and Merchandising print ¥30.0bn of combined segment business profit at FY March 2026. Net of ¥18.6bn of unallocated corporate cost, group business profit is ¥54.7bn.
Net cash is ¥274bn, 29% of the ¥940bn market cap and 79% of book equity.
At spot the market pays ~¥666bn of operating EV against an SOTP operating value near ¥610bn. The ¥100bn three-year capital-return envelope carries no value at this price.
The dossier rests on one question. Is the 18.4% consolidated EBIT margin at FY March 2026 — its decade high — the floor of a structurally cleaned earnings power, or the accounting peak of a low-cadence release year resting on a revenue base in secular erosion. The SOTP moves linearly with the multiple applied to the Digital Entertainment core, and the core multiple moves with whether its 25.1% segment margin is read as a floor or a peak. The fair-value dispersion is almost entirely the dispersion of that one number.
The consensus reading takes the print as a cyclical peak on a melting revenue base. Group revenue has fallen from ¥365.3bn (FY March 2022) to ¥297.7bn, and the forward P/E of ~26x on FY March 2027 earnings reads on that scepticism: a publisher whose earnings rebounded but whose top line is permanently lower.
The variant reading is that the revenue decline is bifurcated, not homogeneous. Digital Entertainment fell from ¥279.7bn (FY March 2022) to ¥172.9bn as the loss-making mobile portfolio was run off by choice and FFXIV sat between expansion cycles ; over the same window the satellites grew — Amusement from ¥43.9bn to ¥71.5bn, Merchandising from ¥12.8bn to ¥23.6bn. The HD sub-segment then carried operating profit from ~¥3.3bn to ¥14.1bn on roughly flat revenue, through low-marginal-cost remasters and catalogue. Mid-cycle Digital Entertainment margin reconstructs to ~22–25%, and the consolidated margin floor to ~15–16% after a recurring restructuring charge.
Two diagnostics close the debate. The Digital Entertainment margin across the FY March 2027 prints separates a structural floor from a low-cadence peak — a return below 20% on the arrival of a heavy AAA slate with fresh content-valuation losses confirms the peak reading. The capital decision — whether the ¥100bn three-year envelope funds returns or an acquisition — separates the protected anchor from a value-trap that erodes its own floor.
Position framing is patient monitoring. The downside is floored by the balance sheet, the upside is unproven, and the weighted asymmetry is roughly neutral at spot. Entry is justified on a sector purge toward ¥1,900–2,100 or on confirmation of either diagnostic. Conviction is moderate.
The eleven-year window from FY March 2016 to FY March 2026 reads as three regimes. The hit-publisher base through FY March 2020, with revenue scaling on the HD release calendar and a mobile portfolio in build-out, EBIT margin oscillating between 9% and 15%. The COVID demand peak and portfolio pivot, FY March 2021 to FY March 2023, with revenue cresting at ¥365.3bn (FY March 2022), the Western studios sold to Embracer (FY March 2023, a non-operating gain lifting net income to ¥49.3bn above EBIT of ¥44.3bn), and the contested blockchain bet under Matsuda. The multiplatform cleanup since FY March 2024 under Kiryu, with content write-offs collapsing the FY March 2024 EBIT margin to 9.1% and net income to ¥14.9bn, followed by margin restored to 18.4% on a smaller ¥297.7bn revenue base by FY March 2026.
| Inflection | FY Mar 2016Hit-publisher base | FY Mar 2022Post-COVID peak | FY Mar 2023Eidos gain | FY Mar 2024Write-off trough | FY Mar 2026Discipline recovery |
|---|---|---|---|---|---|
| Revenue (¥bn) | 215.4 | 365.3 | 343.3 | 356.3 | 297.7 |
| EBIT (¥bn) | 26.0 | 59.3 | 44.3 | 32.6 | 54.7 |
| EBIT margin | 12.1% | 16.2% | 12.9% | 9.1% | 18.4% |
| Net Income (¥bn) | 19.9 | 51.0 | 49.3 | 14.9 | 29.6 |
| Diluted EPS (¥) | ~54 | 142.0 | 137.0 | 41.4 | 82.1 |
| FCF (¥bn) | 16.1 | 22.1 | 6.6 | 43.0 | 45.3 |
| Net cash (¥bn) | ~108 | 158.8 | 191.3 | 223.7 | 274.5 |
| ROE (reported) | 12.3% | 19.3% | 16.4% | 4.7% | 7.5% |
Source: Data pack 31 May 2026 (Income Statement, Segments, Balance Sheet, Ratios) and Tanshin IR 14 May 2026. FY March 2026 = year ended 31 March 2026. EPS reflects the 1-for-3 split effective 1 October 2025 ; FY March 2016 EPS reconstructed from net income on split-adjusted share count and approximate. Net cash for FY March 2016 reconstructed from cash and equivalents net of total debt.
The decade produced the lowest total return in the bucket alongside Nexon, roughly half of Capcom. The IP equity stayed intact and the ex-cash economics stayed excellent, but neither was converted — not into a catalogue annuity, as at Capcom, nor into a disciplined return, as at Konami. The prospective thesis is precisely the conversion that never happened.
Demand runs on three Digital Entertainment pockets and two satellites. HD console and PC is transactional plus an emerging catalogue annuity — new titles plus digital back-catalogue resold at near-zero marginal cost, the model proven at Capcom and beginning to form here through remasters and remakes (FF Tactics Ivalice Chronicles, DQ I&II HD-2D, DQ VII Reimagined). FFXIV is the only contractual pocket: ~30m registered accounts, a 12.99 USD monthly subscription, a ~30%-plus implied margin, indexed to the expansion calendar rather than the console cycle. Mobile is in voluntary run-off. Amusement is domestic arcade footfall ; Publication and Merchandising are manga and IP royalties.
The critical cost is content development — capitalised development amortisation plus content-valuation losses — and it explains the decade's entire margin volatility. The FY March 2024 trough at 9.1% was a write-off artifact ; the FY March 2026 print at 18.4% is discipline plus a remaster-weighted slate. Square Enix manages it through greenlighting discipline, the multiplatform pivot and a remaster focus, without Capcom's proprietary-engine cost moat — which is why the margin is structurally less stable and lower than the compounder's. Platform fees of ~30% are a stable second-order cost. FCF is structurally robust: capex runs at ~2% of revenue, FY March 2026 FCF of ¥45.3bn at a 15.2% margin, FCF/EBIT of ~0.83.
The certified segment economics reset the consolidation reading. The dilution that depresses the consolidated multiple is the ¥18.6bn corporate cost center, not a break-even Amusement leg. Each operating segment carries a healthy segment margin ; the group line nets a corporate overhead that has compounded from ¥9.2bn over the decade and a cash balance earning near nothing. Reconstructed ROIC ex-cash exceeds 40% against a certified cost of equity of 6.9%, while reported ROE sits at 7.5% because net cash is 79% of book equity. The reported return understates the operating economics by the width of the corporate line and the dormant balance sheet.
The ex-cash model is excellent and the reported model is poor, and the gap between them is the dossier. Reconstructed ROIC ex-cash exceeds 40% against a certified cost of equity of 6.9% ; gross margin is 53%, FCF margin 15.2%, capex ~2% of revenue. Reported ROE is 7.5% because ¥274bn of net cash sits at 79% of book equity earning near nothing, and ¥18.6bn of unallocated corporate cost is netted at the group line. The engine generates the cash. The allocation strands it.
No abusive control structure, widely held, TSE Prime — a favourable contrast with NXC-controlled Nexon and the founder families at Koei Tecmo. The policy is a 30% payout, a ¥100bn three-year strategic envelope and a ¥20bn buyback recently activated, against the TSE capital-efficiency reform. The record is ten years of accumulation: net cash up ~¥166bn over the decade, ¥18.6bn of dividends in FY March 2026, buybacks historically negligible, no P/B re-rating. The rerating optionality sits entirely in this pillar, unconsumed.
FFXIV is the only contractual pocket and it is cyclical on expansions ; the catalogue annuity is emerging but unproven. Concentration on two franchises. Digital Entertainment revenue −16.3% with mobile in run-off.
Final Fantasy and Dragon Quest IP equity is deep and the FFXIV switching cost is real, but the moat is narrative-strong and execution-weak. The 9–18% margin range proves the IP depth has not produced annuity stability.
Kiryu's cost discipline is measurable — HD operating profit +¥10.8bn, headcount 4,770→4,290. The track record is short, the Matsuda legacy heavy, and the recurring ¥12.1bn restructuring line flags an honesty question on what counts as exceptional.
A medium-quality hybrid, floored by the balance sheet — above a value-trap, below a compounder. Comparable to the consolidated grade of Toridoll at 16.5/25, below Konami at 18/25 where the balance sheet was converted, below Capcom at 19/25 where the catalogue annuity is proven. The grade does not justify a premium multiple on the consolidated line. The swing factor is the capital decision ; the operation is already sound ex-cash.
Is the restored Digital Entertainment margin a structural floor or a low-cadence peak ?
Will ¥274bn be activated or accumulated ?
The market prices the cash as permanently dormant — the P/B has not expanded despite net cash at 29% of the cap and 79% of book equity. The ¥100bn three-year envelope, the ¥20bn buyback and the TSE reform make activation more probable than at any prior point, against ten years of accumulation. This is the pillar that carries the rerating, and the same envelope is the principal vector of permanent loss if it funds a poor acquisition rather than returns.
Is FFXIV a durable annuity or an MMO in secular decline ?
The market treats FFXIV as a stable annuity underwriting the downside. The MMO declined at FY March 2026 and active characters near ~880k are falling, but the decline coincides with the absence of an expansion after Dawntrail (July 2024). Cyclical trough or structural erosion against free-to-play. It is the only contractual pocket, so the answer governs whether the recurring base holds.
At ¥2,558 spot, ~26x forward P/E and 2.6x book, the market prices an ex-growth publisher whose earnings rebounded cyclically on a permanently lower revenue base, with the balance sheet treated as dormant. Two implicit choices do the work. The revenue decline is read as a homogeneous melt rather than the chosen run-off of loss-making mobile alongside an FFXIV cycle trough. The cash is assigned no activation value — the P/B has not expanded across ten years despite ¥274bn of net cash. Neither choice survives the certified segment reconstruction, but at spot the operation is already paid at ~¥666bn of EV against an SOTP operating value near ¥610bn — the conglomerate discount has been absorbed.
FFXIV enters secular decline with active characters below ~700k. HD margin reverts below ~12% on a heavy AAA slate with fresh write-offs, and the restructuring charge persists. Digital Entertainment operating profit compresses from ¥43bn toward ¥28bn. The downside is timing, floored at ~¥950–1,000 by book and net cash unless an acquisition consumes the trove.
Discipline holds, Digital Entertainment margin stabilises in the low-twenties, FFXIV oscillates around its trough and recovers at the next expansion, mobile completes its run-off. Cash stays largely accumulated at a 30% payout with a modest buyback. The certified SOTP — core at 12x, satellites on segment margin, net of capitalised corporate overhead — delivers ¥2,400–2,500. Re-rating happens or does not ; the fair value does not require it.
HD margin confirms structurally above 16% across the year on the catalogue annuity and multiplatform, FFXIV rebounds at the next expansion, and the capital ramp materialises — buyback sustained, payout lifted — driving a simultaneous P/B re-rating. The invisible catalysts of the expectations analysis materialise together over 18–24 months.
| KPI | Latest value | Status | What it tells us |
|---|---|---|---|
| Digital Entertainment segment margin | 25.1% (FY26) | Cardinal | The number governing the dossier. The FY March 2027 prints close Debate 1. Sustained above 20% validates the base ; below 20% with fresh write-offs on a heavy slate triggers bear. |
| Capital return — payout / buyback | 30% / ¥20bn | Trigger | Payout above 30% or buyback executed above ¥20bn per year re-rates the P/B and confirms the shareholder-alignment pillar. |
| M&A announcement >¥50bn | none | Trigger | An acquisition at an unjustified multiple consumes the floor and forces re-underwriting toward bear. The principal vector of permanent loss. |
| FFXIV active characters | ~880k | Watch | Recovery at the next expansion confirms the cyclical annuity ; below ~700k reveals secular erosion of the only contractual pocket. |
| Restructuring charge | ¥12.1bn (FY26) | Asymmetric | Third consecutive year of cleanup. A fourth year above ¥10bn confirms it is endemic and the normalised earnings power must carry it perpetually. |
| FCF yield (gross / ex-cash) | 4.8% / 6.8% | Asymmetric | FY March 2026 FCF ¥45.3bn. Ex-cash yield on ¥666bn EV sits above the 6.9% cost of equity — the defensive return floor. |
| Net cash / book equity | ¥274bn / 79% | Holding | 29% of market cap. The dormant trove that floors the downside and carries the unconsumed rerating optionality. |
| P/B vs decade corridor | 2.6x | Reference | Inside the 2.2–2.6x corridor held for ten years. Above 3.0x signals rerating ; below 2.0x signals de-rating toward book. |
| Beta brut vs TOPIX | 0.24 | Reference | Lowest in the bucket. Confirms the balance-sheet-anchor defensive profile empirically. |
A Digital Entertainment segment margin below 20% across the FY March 2027 prints, or a fourth consecutive restructuring charge above ¥10bn, reframes the dossier from protected false negative toward slow-destruction value-trap and compresses fair value toward ¥1,800. FFXIV active characters below ~700k at the next expansion would confirm secular erosion of the only contractual pocket as the second-order bear confirmation.
A payout lifted above 30% or a buyback executed above ¥20bn per year re-rates the P/B and consumes part of the discount on its own. HD margin sustained above 16% across the year without fresh write-offs confirms the catalogue annuity and the structural floor at the upper end of the reconstruction.
An M&A announcement above ¥50bn at an unjustified multiple inverts the protection — it consumes the net cash that floors the thesis 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.