1Q FY27: Margin Recovers, Migration Does Not

Thesis status: Affirmed on margin. Unresolved on migration. Priced for the peak.

Thesis Refresher

The primer framed Nintendo as a sequencing problem. The hardware is sold. FY26 closed with the Switch 2 base built and consolidated gross margin down 21.7 points to 39.3%, which is what a launch year costs. What follows depends on software monetisation restoring the economics of the prior cycle across FY27 to FY30, with the Pokémon equity layer and the revised capital return formula underneath as durability that does not require platform execution to work.

This is the first update. 1Q FY27, the quarter ended 30 June 2026 and reported 6 August 2026, is the first period in which any of the seven watch conditions could be read at all.

1Q FY27 in Brief

Mix contributed to the divergence between falling sales and rising profit, but it does not explain the headline increase. Hardware fell to 55.3% of dedicated platform sales from 78.8%, first-party rose to 82.6% of software sales from 64.8%, and IP related income more than doubled to ¥34,803m on the Super Mario Galaxy Movie. Fewer consoles shipped and more software sold, which is the flywheel the primer described. The June quarter is structurally the most software-weighted of the four, and 3Q carries the hardware-heavy holiday sell-in, so 54.3% will not hold across the year.

Nintendo left every full-year line unchanged. Operating profit reached 38.5% of the forecast in the seasonally lightest quarter, and profit attributable to owners reached 47.6%. Two readings fit. Management is holding a deliberately conservative forecast, or it expects the first quarter's help not to repeat. The next section sizes how much of that help was operating.

The Quarter Net of Tariffs and Currency

Operating profit rose ¥85,668m. Roughly ¥70.0bn of that came from two items that sit inside reported operating profit and outside the mix improvement the thesis tracks.

Nintendo recorded approximately US$300m, about ¥47.8bn at the quarter's average rate of 159.38, as a reduction of cost of sales on refunds of IEEPA tariffs. The refund reverses a cost the company had already carried, and management disclosed that the tariffs had been borne primarily by Nintendo rather than passed to consumers. Absorption is the behaviour Categorical Condition B assumes when it frames cost as contained, so the refund confirms the condition's mechanism at the same time as it flatters the margin. Nintendo disclosed no split of how much further refund falls into later quarters, which means the full-year effect cannot be apportioned from published disclosure.

Currency is the larger analytical problem. The yen averaged 159.38 to the dollar against the ¥150.00 the FY27 forecast assumes, and closed the quarter at 162.39. Translation added roughly ¥22.2bn at the operating line and ¥39.0bn to net sales. Procurement runs the other way, because Nintendo disclosed non-consolidated purchases of US$2.0bn in the quarter against consolidated dollar sales of US$1.3bn, which leaves the group net short dollars on operating flows. A weaker yen flatters the reported result now and raises the bill of materials that the September price revision was sized to cover.

Strip both and the quarter still improves. Underlying operating profit was about ¥72.6bn on net sales of ¥478.8bn excluding the currency benefit, which is a margin of roughly 15.2% against the prior year's 9.9%. The headline printed 27.5%.

The framework holds both variables, and holds them in the wrong form. Categorical Condition B names tariff exposure as a cost to contain and currency as pressure on the bill of materials, so each appears as a risk to be managed. Neither is scoreable as a driver of reported earnings, which is what both were this quarter. The two largest movements in the first testable period had nowhere to be recorded.

The Guidance Triangle Narrows

Nintendo guides operating profit and net sales. It discloses neither a gross margin nor an SG&A forecast, so the primer could fix only two of the three and flagged the gap at seeding. Holding SG&A at FY26's ¥548,839m required a gross margin of 44.8%. Holding the primer's 41.8% required SG&A to fall 11.3% to ¥486,900m. The two cases described different companies.

The 1Q FY27 Supplementary Information narrows it decisively toward the first case. Advertising falls 17.0% as the primer anticipated once FY26 launch spending normalised, and R&D rises 6.9%, which more than absorbs the saving. SG&A therefore falls 2.3%, and the implied gross margin lands at 44.2%.

The triangle is narrowed rather than closed. Nintendo disclosed two of the SG&A components it needs, and the remaining ¥226,439m of FY26 cost is held flat to reach ¥536,400m, so 44.2% is a model output resting on that hold rather than a published forecast.

The condition asks Nintendo to recover toward 41.8%. Management's own guidance implies 44.2%, so the threshold sits 2.4 points below the forecast, and meeting guidance passes the condition automatically. The quarter printed 54.3%, and 45.1% stripping the tariff refund from cost of sales entirely. The currency benefit cannot be stripped at the gross line, because Nintendo disclosed its effect on operating profit and on net sales but not on gross profit.

Gross profit in yen barely moves under the guidance. The margin recovery arrives arithmetically, out of net sales falling 11.4% against a cost base that falls with them. Incremental gross profit from software mix is still ahead of the company, and 3Q is the hardest read on whether it arrives, because holiday hardware sell-in carries the heaviest mix against it.

Watch Condition Assessment

*C2 is met but non-discriminating. Its threshold now sits below management's own implied forecast, so it passes whenever guidance is met. Read the green as met rather than as informative until the restatement re-bases it.

Seven conditions, five of them load-bearing. Three statuses differ from the ledger seeded at the primer. Two of those changes correct a seed that graded a mechanism the quarterly disclosure does not carry, and the third follows the guidance triangle above.

C1: Switch 2 standalone software attach carries unit growth

🟡 Developing → The bundle test is met decisively. The migration test moved backwards.

Nintendo began disclosing quarterly bundled units with this filing. Approximately 0.35m of the 9.46m Switch 2 software units were bundled, putting the bundled share at 3.7% against 25.9% in FY26. The condition requires that share to fall. The reading is ambiguous on mechanism, because a bundle programme ending produces the same collapse as consumers choosing to buy software separately, and Nintendo confirmed at FY26 that Mario Kart World hardware bundle shipments concluded in December. Voluntary attach and programme timing are not separable from what is disclosed.

The second strengthening trigger asks Switch 2 to rise materially above 26.2% of combined platform software. It fell to 21.9%, the weakest of the five quarters on record. Switch 2 software grew 9.2%. Legacy Switch software grew 38.6%, carried by Tomodachi Life: Living the Dream at 7.94m units.

Japan is converting. The Americas and Europe are moving the other way, and both now sit more than five points below Japan on a measure where all three were within 1.6 points of each other a year ago. Tomodachi Life is the obvious candidate for the whole effect, and the primer's noise filter names exactly this kind of release-window distortion. Removing it from the Western totals narrows the divergence without closing it, though the filing splits the title's 5.39m units outside Japan no further, so that adjustment carries direction and not magnitude.

The volume test remains open. Ex-bundle Switch 2 software of 9.11m is 15.2% of the 60.00m the guidance requires, delivered in the seasonally lightest quarter, and the year still asks for 3.64x ex-bundle attach against the 2.38x this quarter produced.

C2: Gross margin recovers toward ~41.8% through FY27

🟢 Affirmed* ↑ Every leg the condition names has a read, and each one improved. Met, and no longer discriminating.

The condition names gross margin recovery, rising software share and moderating hardware mix. All three are readable this quarter and all three moved the right way. Hardware fell to 55.3% of platform sales, first-party software rose to 82.6%, and digital reached ¥132,700m at 61.5% of software sales, up 90.0% on the prior year.

The green says more about the calibration than about the business. Every mechanism the condition names is observable and improving, which is what the status records. Whether that tells a reader anything is the subject of the framework modification below.

C3: Western hardware share holds at or above FY26's 56.1%

🟡 Developing ↓ The hardware leg holds. The conversion leg is diverging, and the price test has not happened.

The Americas and Europe took 218 of the 382 ten-thousand Switch 2 hardware units in the quarter, or 57.1%, against the 56.1% FY26 baseline the condition is measured from. On sales rather than units, the two regions were 65.8% of net sales against 64.2% in FY26. The condition is met on the printed numbers, and it is met in a quarter that tests nothing, because the US and Europe do not reprice until 1 September 2026.

The prior-year quarter is the less flattering comparison. Western share of Switch 2 hardware was 58.8% in 1Q FY26, so the quarter sits above the annual baseline while falling 1.7 points year on year.

One early directional datapoint exists, and it points away from the worry. Japan was the only region carrying a price increase in the quarter, from 25 May, and Japanese Switch 2 hardware fell 18.1% against the Americas down 41.3%, Europe down 28.4% and other regions down 47.8%. The priced region declined least. The comparative is a launch quarter in which Japanese supply was tight against a very large Americas spike, so the base periods are not equivalent and the reading is suggestive only.

The condition's second leg asks Western software mix to converge toward Japan's. It diverged, and the weakening trigger naming Western markets still weighted to legacy Switch is the one currently firing. The trend moves to deteriorated on that leg alone.

Switch 2 hardware components sum to 381 against a printed total of 382, a rounding difference at the ten-thousand unit Nintendo discloses.

C4: Pokémon Company equity contribution grows in FY27 and FY28

🟡 Developing → Directionally supportive, and not separately measurable.

Non-operating income rose 62.3% to ¥63.7bn, which Nintendo attributes to foreign exchange gains and the share of profit of equity-method entities together. The equity-method line does not appear in the quarterly Explanatory Material at all. Stripping the disclosed ¥16.8bn of FX gains leaves ¥46.9bn of interest income, equity-method profit and other against ¥37.0bn a year earlier, up 26.8%, and part of that residual is interest income on a cash pile earning more as Japanese rates rise.

The release evidence is intact. Pokémon Pokopia sold 1.27m units in the quarter to reach 3.68m life to date, Pokémon Champions released on 8 April 2026, and Pokémon Winds and Pokémon Waves remain scheduled for 2027 on Switch 2.

The ledger seeded this condition affirmed on the FY26 step-up. A condition naming a contribution that no quarterly disclosure isolates cannot be graded affirmed on a quarter, so the status moves to developing. The trend holds unchanged, because the correction is to the grading rather than to the business.

C5: Dividend at or above ¥162 with buybacks continuing at scale

⏸ Dormant → A maintained forecast is not a test, and the buyback leg cannot be read in a first quarter.

The FY27 dividend forecast is unchanged at ¥162. No repurchase was executed, and the balance sheet confirms it. Net assets fell ¥36,229m despite ¥147,423m of profit, and profit plus the ¥20,344m move in accumulated other comprehensive income less that fall implies distributions of roughly ¥204,000m, which is the ¥177 FY26 year-end dividend on 1,152,828,705 shares. The share count is unchanged and needs no re-mark.

Absence of a buyback carries no signal here. FY26's entire ¥99,935m repurchase was a single transaction on 3 March 2026, at the year end. The weakening trigger requires repurchases to decline materially without explanation, which the year cannot answer until it is closer to done.

A tension is building underneath. Profit attributable to owners reached 47.6% of the full-year guide in one quarter, and the formula keys off operating profit and the payout ratio, so an unrevised forecast implies an unrevised dividend against earnings running well ahead.

CAT-A: Software pipeline emergence

🟡 Developing ↑ The closest this condition has come to satisfaction, and it is not satisfied.

The 1Q FY27 launch schedule adds The Legend of Zelda: Ocarina of Time under Nintendo Switch 2 software for 2026, with no Switch counterpart disclosed. A mainline Zelda property on a Switch 2-only line is the shape the condition asks for. It fails on two counts. The date is "2026" rather than firm, and a remake of a 1998 title leaves "confirmed system-seller scale" unsettled.

Pokémon Winds and Pokémon Waves remain listed for 2027 on Switch 2 with no Switch path, undated and unchanged from the FY26 disclosure. Nintendo Switch Sports Resort is dated 22 October 2026, Fire Emblem Fortune's Weave firmed to 17 September 2026, and Xenoblade Genesis is listed for 2027.

What shipped in the quarter was mid-tier. Yoshi and the Mysterious Book arrived on 21 May and Star Fox on 25 June, alongside Switch 2 Editions of Xenoblade Chronicles X and Xenoblade Chronicles. The quarter's biggest seller was a Switch title, and Tomodachi Life is the cross-generational strategy working commercially while moderating the conversion pull this condition exists to measure.

CAT-B: Cost structure containment

🟡 Developing → Every disclosed test passes. The assumption the condition named as its own risk is now adverse.

No further price revisions were announced. Gross margin recovered to 54.3% and holds at 45.1% stripping the tariff refund. The ¥100,000m component and tariff cost assumption inside guidance is unchanged. The IEEPA refund resolves the tariff leg positively, reversing a cost and confirming that Nintendo absorbed it rather than pricing it through.

Currency is where the condition is exposed. The yen averaged 159.38 against the ¥150.00 the price revision was sized on and closed the quarter at 162.39.

Two of the condition's named tests cannot be read at all. Nintendo discloses consolidated gross margin only and never a hardware gross margin, and no memory pricing commentary appeared this quarter. Both are proposed for refinement below. The status holds at developing because the readable tests pass in a quarter that a non-recurring refund flattered.

What the Price Implies

Nintendo trades at ¥8,900 on the Tokyo exchange, and two thirds of that is the console platform. The remainder is a balance sheet holding ¥2,216,856m of cash and securities against no interest-bearing debt, and an equity-method stake valued on its earnings.

The Pokémon layer is struck at 15x FY26 equity-method income of ¥82,792m rather than deducted at book, because Nintendo does not disclose the stake's carrying value. It sits inside ¥420,875m of non-current investment securities. Deducting book and capitalising the income would count the same asset twice.

That leaves an operating enterprise value of ¥6,801,439m for the platform. The solve runs a six-year explicit period to FY32, one full console generation, at a 30% normalised tax rate with capex at 1.2% of sales and working capital at 19.5%. Terminal revenue is normalised on the mean of FY26 to FY32, a full generation with the launch year in the basis, because Nintendo oscillates by cycle instead of compounding. At a 7.79% cost of equity, built from a 2.86% JGB yield and a 0.85 beta on a 5.8% Japan equity risk premium, the terminal operating margin the price requires is:

Terminal growth is the reason for the band. The 1.0% assumption was set against a 1.7% risk-free rate. Japanese ten-year yields now stand at 2.86%, and holding nominal growth fixed while the nominal discount rate rises widens the capitalisation spread by 21%. With three quarters of the value sitting in the terminal, that spread drives most of the result. Move terminal growth up by the same 116 basis points and the requirement falls to 34.1%.

Which end applies depends on whether the yield move reflects inflation or real rates, and this memo does not resolve it. The conclusion does not turn on the answer.

Nintendo has cleared these margins before. FY21 reached 36.4% on Animal Crossing and a housebound world, and FY09 reached 30.2% at the Wii peak, so 34.1% sits 2.3 points below the company's own best year rather than beyond anything it has done.

The solved figure is the steady state the platform reaches in FY32 and then holds, with the explicit years gliding to it from FY27's guided 18.1%. Those six years average 26.1% at the lower bound and 30.2% at the upper. FY26 enters the normalisation through its revenue alone, so the 15.6% it printed while the Switch 2 base was being built is already absorbed.

With the launch year behind it, the price requires roughly pandemic-peak economics to arrive by FY32 and then hold. Peak economics happen. Peak economics as the terminal state are a different claim, and at the upper bound of 42.4% there is no precedent on either an annual or cross-cycle basis.

Two limits on reading this. The enterprise value bridge uses the 31 March 2026 balance sheet against a 15 August 2026 price, because the quarterly disclosure carries no cash and securities split. And this states what the traded price embeds, not a target price.

Thesis Standing

The business thesis survives its first test. The stock setup is the harder question, and the two should not be collapsed into one answer.

On the business, the primer asked for software monetisation to restore cycle economics, and the mix evidence supports it. Hardware share fell, first-party and digital share both rose, and the margin recovery holds once the tariff refund comes out. The flywheel is turning.

The framework measured the business correctly and missed the quarter. Both variables that decided it sit in the condition set as risks to manage rather than as scoreable drivers of reported earnings, which is a defect in the framework and not in the business.

The migration leg is the genuine complication. A cross-generational strategy that sells a legacy title in the millions is working commercially and slowing the conversion the thesis depends on. One quarter carrying one large legacy release cannot settle whether that is a release window or a pattern. The 2H FY27 read, after the Western repricing, is the one that will.

On the stock, the price embeds peak-like economics as the terminal state. An affirmed thesis and a forgiving entry price are separate findings, and only the first is established here.

Framework Modification

C2's threshold has stopped functioning. The primer set it at 41.8% when Nintendo had disclosed neither an SG&A nor a gross margin forecast. The implied FY27 figure is now 44.2%, so the condition sits below management's own forecast and passes automatically whenever guidance is met. Re-base it to 44.2% at the FY27 restatement.

Standing Overlay

The overlay exists because the variables that moved this quarter sit in no condition as scoreable earnings drivers and the restatement is eight months away. It carries no status and no thresholds, which keeps it a reporting layer and holds the line between what a memo may change and what only a restatement may add.

What to Watch

Source: Nintendo 1Q FY27 Financial Report and Supplementary Information, quarter ended 30 June 2026, reported 6 August 2026. Nintendo historical annual results, including FY09, FY21, FY25 and FY26. Valuation figures from in-house modelling, price marked 15 August 2026.