Nintendo7974.T
Nintendo reports a 15.6% operating margin for FY March 2026, against a 26–28% mid-cycle and a 36.4% COVID peak. The launch of Switch 2 was a record — 19.86m consoles in ten months — on a diluted margin. The stock has given back the entire June-2025 hype spike, −48% from ¥13,880 to ¥7,148. The operational entity is priced near 10x normalised EBIT with no value to the ~¥2tn net cash. The diagnostic is the FY March 2027 operating margin print and the cumulative Switch 2 tie-ratio.
Nintendo at a 20.5% mid-cycle margin on ~¥2,150bn normalised revenue, restated to USD/JPY 130, produces operational EBIT of ~¥441bn.
At 17.8x EV/EBIT, justified by 36% ex-cash ROIC and the first-party IP moat, operational EV is approximately ¥7,845bn.
The Pokémon Company equity-method pole adds ~¥880bn; excess cash adds ~¥2,047bn.
Float market cap net of treasury at spot is ¥8,240bn. The operational entity is priced near 10x normalised EBIT, with no value to the ~¥2tn cash.
The dossier rests on one question. Is the FY March 2026 margin compression to 15.6% a timing air-pocket — a launch-year mix dilution plus a transitory memory cost that reverses — or the first reading of a generation whose unit economics are structurally degraded, where a higher DRAM floor and a softer software attach at the new price tier hold the margin below the Switch-generation mid-cycle. The valuation moves linearly with the normalised generation margin. Everything in the fair-value dispersion tracks that single number.
The consensus reading takes the memory shock as cyclical. Mid-cycle reverts mechanically toward ~24% by FY March 2029 (consensus estimates), in line with the 26–28% historical mid-cycle.
The variant reading is that the cost regime has changed in nature. DRAM prices doubled in Q1 2026, Nintendo pays ~+41% on its memory modules, memory is 21–23% of the hardware BOM, and the combined component-plus-tariff hit is ~¥100bn on FY March 2027 operating expense. AI datacenter demand keeps supply tight to ~2028 (TrendForce, IDC, SK Hynix). The historical console playbook — component costs fall with process maturity, prices come down, the base expands, margin rebuilds — no longer holds. The Switch 2 generation may cap below the historical mid-cycle. Nintendo raising the hardware price by $50 while easing software pricing through bundles is the inversion of its own model and the tell to watch.
The print resolves it. An FY March 2027 operating margin below ~17%, with the cumulative Switch 2 tie-ratio stalled below ~5.5x at the close of year two, reclassifies the compression as a structural ceiling and compresses fair value toward ~¥6,100. Above those thresholds, the quality compounder is confirmed. The first reading is the 30 July 2026 result.
The position framing is patient ownership at current levels, gated by the FY March 2027 prints on a published calendar. Sizing at entry is contained by the net-cash floor. The prints govern subsequent adds or trims. Conviction is moderate-to-strong.
The window from FY March 2015 reads as three sequential regimes. The exit from the Wii U failure through the Switch launch, FY March 2015 to FY March 2021, with operating margin moving from ~5% to a COVID-amplified 36.4% peak and ROE reaching 28.1%. The mature plateau and end-of-cycle, FY March 2022 to FY March 2025, with yen-translated profit elevated by the weak yen of 2022–2024 and the pre-Switch-2 trough at FY March 2025 — revenue −30.3%, operating cash flow near zero on pre-launch inventory build. The Switch 2 launch and purge, FY March 2026 onward, with revenue +98.6% on the best hardware launch in Nintendo history at a margin diluted to 15.6%, and the stock peaking at ¥13,880 in June 2025 before purging to ¥7,148.
| Inflection | FY Mar 2017Wii U trough | FY Mar 2018Switch launch | FY Mar 2021COVID peak | FY Mar 2025EOL trough | FY Mar 2026Switch 2 launch |
|---|---|---|---|---|---|
| Revenue (¥bn) | 489.1 | 1,055.7 | 1,758.9 | 1,164.9 | 2,313.1 |
| EBIT (¥bn) | 29.4 | 177.6 | 640.6 | 282.6 | 360.1 |
| EBIT margin | 6.0% | 16.8% | 36.4% | 24.3% | 15.6% |
| Net income (¥bn) | 102.6 | 139.6 | 480.4 | 278.8 | 424.1 |
| ROE reported | 8.5% | 10.9% | 28.1% | 10.5% | 14.9% |
| FCF (¥bn) | 8.6 | 142.6 | 605.1 | −6.9 | 262.6 |
| Net cash (¥bn) | 946.1 | 988.0 | 1,736.6 | 2,051.0 | 2,216.9 |
| EPS (¥) | 85.4 | 116.2 | 403.3 | 239.5 | 364.5 |
Source: Data pack and issuer xlsm, pull 31 May 2026. FY March 2026 = year ended 31 March 2026. Net cash = cash and short-term investments net of financial debt (zero at FY March 2026). EBIT is the J-GAAP operating-profit line.
The value creation is operational and real, not manufactured by share-count compression. EPS rose from ¥85.4 to ¥364.5 across the window while the share count fell only modestly; buybacks are not the EPS lever. The one financial distortion runs the other way — ordinary income of ¥542.1bn against operating profit of ¥360.1bn in FY March 2026, ~¥182bn of non-operating income from interest on the cash, FX gains, and the equity-method stakes including The Pokémon Company. The reported P/E reads on an inflated earnings line. The Shareholder-alignment pillar at 3.0/5 captures the dormant capital directly.
The engine runs on the convertible installed base, not raw hardware volume. The 155.92m Switch gen1 units plus 19.86m Switch 2 form an annuity reservoir; gen1 software still sold 136.91m units in FY March 2026. The economically meaningful demand is the stock of users buying first-party software at near-pure margin, on a base whose beta is 0.54. Hardware is the subsidised entry point at low or negative launch margin; the value is the software lifetime per installed user. Group gross margin fell from 61.0% to 39.3% in the launch year on the hardware mix alone, while software gross margin runs above 60% in regime.
Mario, Zelda and Pokémon constitute the deepest first-party IP in the bucket — more durable than Capcom's catalogue annuity, incomparable to Sony's portfolio where games are ~32% of profit. The closed ecosystem creates real library switching costs. The constraint is recurrent: the hardware can fail at a generation transition, and the Wii U is the live memory of that risk. The IP rent off-console — mobile, films, parks, licensing — remains marginal at ¥73.5bn and fell 9.7% in FY March 2026, which keeps the dossier exposed to the platform cycle rather than to a decyclicalised IP annuity.
The critical cost has changed nature. Memory DRAM/NAND, not development cost, now explains the majority of margin volatility, and the shift is plausibly structural rather than a cyclical spike. Nintendo passes the cost partly through the hardware price hike, +$50 to $499.99 in the United States from 1 September 2026, in line with the Sony PS5 +$100 move but without precedent in Nintendo's own history, which used to cut prices mid-cycle. The combined cost runs at ~¥100bn of FY March 2027 operating expense, ~25–28% of FY March 2026 EBIT.
Cash conversion is strong but lumpy. The model is asset-light at capex ~1.2% of sales, zero financial debt, FCF of ¥262.6bn in FY March 2026 after the −¥6.9bn launch-inventory year. The cash conversion cycle ran to 95 days on the launch build, with reversion modelled. Conversion is threatened by working-capital timing across the cycle, not by capital intensity. The yen-translated profit of FY March 2022–2026 was lifted by a weak yen against the ~130 normative; the Base case neutralises that translation before any multiple.
First-party IP — Mario, Zelda, Pokémon — is the deepest and least replicable asset in the bucket, paired with a closed hardware-software-OS ecosystem that creates library switching costs. The combination of proprietary world-class IP and proprietary platform is unique in 06a: Sony holds the platform but weaker IP, Capcom holds IP but no platform, Konami neither at scale. The rating is held below 5 by the recurrent generation-transition risk — the hardware can fail, as the Wii U did — and by an off-console IP rent that is regressing, IP income −9.7%.
Operational ROIC ex-cash at 33–36% against a cost of capital of 7–9%, capex ~1.2% of sales, software gross margin above 60% scalable on the installed base. The reported ROC is crushed to 14.9% by the dormant cash, which is a Shareholder-alignment matter, not a model defect. The model carries one unresolved swing: whether the generation margin is resilient to the memory regime, which is the dominant debate. The launch-year compression to 15.6% is a mechanical mix point, not the earnings power.
Record launch and multigenerational IP loyalty, beta 0.54 confirming a defensive entertainment profile. Cyclical of generation, with EBIT amplitude ~22x; recurrence is behavioural, not contractual. Launch attach soft at 2.45x.
Post-Wii U recovery from ¥29bn to ¥641bn EBIT, record Switch 2 execution, guidance prudent and honest — the FY27 16.5m guide sits below the +1.2% consensus. Drag from slow IP diversification and a historically timid allocation ramp.
~¥2.2tn dormant cash, 75% of equity, crushing reported ROC. A ¥100bn buyback with 14m shares cancelled and a ~60% payout mark a nascent TSE responsiveness against a hoard. Dispersed shareholding, no controlling family — an advantage over Koei and Nexon.
The highest quality in 06a alongside Capcom (19/25), above Konami (18/25) and Sony (17.5/25). The grade is not a hierarchy of entry: the two best qualities offer the least favourable asymmetry once the FX is normalised, while Nintendo's +27% comes from the cycle trough, not from quality superiority. The grade justifies underwriting the operational entity at a quality multiple and treating the cash and the equity-method pole as separate SOTP blocks.
Is the launch-year margin compression a timing air-pocket or a structural memory-regime ceiling ?
Is the software attach a durable cross-generational annuity or a launch-bundle artifact ?
Launch attach of 2.45x is inflated by bundling — Mario Kart World 81% in-bundle — the digital ratio slipped to 54.5%, and Nintendo is easing software pricing. The consensus expects the attach to build in year two as the library thickens, toward the ~9x Switch-generation life. The risk is that buyers take the hardware but not the games at the higher price tier. The interaction with Debate 1 is the real point: a durable attach absorbs the memory shock through software margin, a soft attach amplifies it.
Is the ~¥2tn cash a TSE allocation ramp or a structural hoard ?
The ¥100bn buyback against ~¥2.2tn of cash is ~5%, the IP income is regressing −9.7%, and the P/B has never re-rated on this name — the contrast with Konami, re-rated from 1.6x to 4.55x on capital conversion, is the reference. Dispersed shareholding and no controlling family leave the optionality available, unlike Koei or Nexon. The cash is the floor under the downside, but it has never bought a premium.
At ¥7,148 the market prices the operational entity at ~10x normalised EBIT against a 17.8x base multiple, with zero value to the ~¥2tn cash and a scenario near the bear margin. Two anchoring choices do the work. The reported P/E of 20.5x reads on a net income inflated by ~¥182bn of non-operating income, rather than on the operating line. The 15.6% launch margin is treated as the earnings power, rather than as a mechanical mix trough below a 20.5% normalised base. The valuation rejects the P/E and the EV/EBITDA — the launch EBITDA is depressed — and runs on EV/EBIT over normalised operating profit, a three-block SOTP, and the FCF yield as floor.
The AI-DRAM cost proves structural to 2028, pricing does not absorb it, the attach stays soft at the $500 tier, and the mass-market transition stalls. Generation margin caps at ~18% on ~¥2,050bn revenue, EBIT ~¥369bn, and the market de-rates the operational multiple to 12.0x. Fair value ¥6,124 central. The downside is timing while the net-cash floor holds; it becomes permanent loss only if the structural ceiling and the attach failure confirm together, taking fair value toward ¥5,000–5,500.
Nintendo executes the cycle without surprise, the attach builds in year two on the 175.8m base, the memory cost is partly passed through the hardware price, and margin rebuilds toward the low mid-cycle. Revenue ~¥2,150bn at USD/JPY 130, margin 20.5%, EBIT ~¥441bn at 17.8x EV/EBIT, plus the ¥880bn equity-method pole and ¥2,047bn excess cash. Fair value ¥9,344 central. Consolidated re-rating happens or does not; the fair value does not require it.
The invisible catalysts materialise together: the memory cost normalises post-2027, the attach builds toward the Switch-generation path, the catalogue annuity and the films-and-parks IP optionality rise, and the TSE allocation ramp accelerates into a P/B re-rating. Revenue ~¥2,250bn, margin approaching the 26% mid-cycle, EBIT ~¥585bn at a 20.0x compounder multiple. Fair value ¥13,022 central.
| KPI | Latest value | Status | What it tells us |
|---|---|---|---|
| Operating margin + tie-ratio (FY March 2027) | 15.6% / 2.45x | Cardinal | The joint number governing the dossier. First reading 30 July 2026. Margin below 17% with cumulative tie-ratio below 5.5x triggers the bear and a ~¥6,124 reframe; above both confirms the cyclical compounder. |
| Margin <17% + tie-ratio <5.5x on two prints | not yet | Trigger | Bear-confirmation invalidation. Reframes the dossier toward ¥5,964–6,284 fair value. |
| DRAM/NAND memory cost | +41% / ~22% BOM | Asymmetric | ~¥100bn FY27 opex hit. A persistent non-decreasing trajectory confirmed structural is Thesis Breaker T3. Vigilance +10% YoY, alert +25%. |
| Operational ROIC ex-cash | 36.4% | Asymmetric | Against a 7–9% cost of capital. The capital machine the reported 14.9% ROE masks. Anchors the quality multiple on the operational pole. |
| Net cash / equity | 75.0% | Holding | ~¥2,217bn, zero financial debt. The downside floor. A ratio falling over 12–18 months would mark the allocation ramp; flat confirms the hoard. |
| Capital return (buyback) | ¥99.9bn FY26 | Watch | 14m shares cancelled, ~60% payout. ~5% of the cash. Below ¥80bn/year would revise the share-count compression. |
| Equity-method share of profit | ~¥40bn | Watch | The Pokémon Company ~32%, DeNA, Niantic. Valued at 22x for ~¥880bn. Yuho certification pending; run-rate below ¥35bn revises the pole. |
| Performance 1Y / from June-2025 peak | −39.6% / −48% | Reference | Worst decadal drawdown. Spike fully retraced from ¥13,880. Consistent with a quality-premium purge, not a value-destruction print. |
| Beta brut vs TOPIX | 0.54 | Reference | Defensive entertainment-demand profile. Confirms decorrelation from the consumer cycle. |
An FY March 2027 operating margin below ~17% with the cumulative Switch 2 tie-ratio below ~5.5x across two prints reframes the dossier toward bear and compresses fair value toward ¥5,964–6,284. The DRAM cost confirmed structural beyond 2028, invalidating the 26–28% mid-cycle, is the integral Thesis Breaker that returns the dossier to a re-underwrite this initiation does not attempt.
An acceleration of the TSE allocation ramp — buybacks and dividends taking the net-cash-to-equity ratio materially lower over 12–18 months — would confirm the dormant optionality and support the bull P/B re-rating. The cumulative tie-ratio building toward the Switch-generation path, paired with a memory cost normalising post-2027, confirms the structural compounder at the upper end of the prudent range.
A large value-destructive acquisition financed out of the cash, against a conservative track record that makes it currently unsignalled, would force a complete re-underwrite. The dominant downside is the hoard — an opportunity cost, not a destruction — bounded by the asset floor at ~¥2,200 per share from cash and the equity-method pole.
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.