Nintendo Spent 288 Billion Yen of Cash and Securities to Build 815 Billion Yen of Inventory

Nintendo spent the first half of its financial year buying something. The company's own balance sheet says what, but not why: inventories rose from ¥539.8bn on 31 March 2026 to ¥814.6bn on 30 September, an increase of ¥274.8bn in six months and 65% more stock than it held a year earlier.

That is a larger sum than the ¥216.5bn of operating profit Nintendo earned in the same six months. It is 2.5 times the highest September inventory balance of the previous eight years — ¥326.1bn, in 2018, the Switch 1's second year. And it was not funded out of profit: cash and deposits fell by ¥237.9bn, marketable securities by ¥50.5bn, and notes and accounts payable rose by ¥134.6bn. Nintendo drew down ¥288bn of cash and securities and stretched its suppliers by another ¥135bn to pay for stock.

Nintendo does not hide any of this. Its Q2 results presentation of 5 November states that "our working capital needs for day-to-day operations have grown, driven in part by a substantial increase in inventories. We are also using our cash reserves to respond to these changes in the business environment." The tanshin filed the same day attributes the rise in total assets to "increases in inventories and investment securities, despite a decrease in cash and deposits."

What neither document does is say what the inventory is for. That is the question this article tries to answer, and the answer matters beyond Kyoto: Nintendo is the only console maker that discloses a quarterly balance sheet in this detail, so its stockpile is the clearest available measurement of how the industry is positioning itself against the component shortage that has dominated the last two years.

The company told investors it was buying memory

Eight months before this balance sheet, Nintendo was asked directly about component costs. At the nine-month briefing on 3 February 2026, the first question was how rising memory prices were affecting Switch 2 hardware profitability and whether a price revision was coming. President Shuntaro Furukawa's answer is the closest thing to an explanation on the record:

"We do not disclose details regarding individual components, but we can say that we are working to secure stable supplies of memory components by holding discussions from a long-term perspective with our business partners. As a result, the recent rise in memory prices did not have a significant impact on hardware profitability in the third quarter. In addition, we do not expect any significant impact in the fourth quarter."

A company that says it has secured memory "from a long-term perspective" and that price rises have not touched its margins has, by construction, bought forward. Inventory is where bought-forward components sit. The ¥275bn increase is consistent with exactly that: a forward purchase of components, paid for in cash and supplier credit, taken into stock before it is needed.

Consistent, but not proof. Nintendo discloses inventory as one line and does not break it into raw materials, work in progress and finished goods in its quarterly filings. A ¥275bn increase is equally consistent with consoles built in advance of a holiday quarter. Both readings are bets on the same thing, and the difference between them is the difference between a cost hedge and a demand forecast.

The demand side of the same filing is weaker

The second half of Nintendo's half-year report points the other way. Switch 2 hardware shipments were 6.49 million units in the six months to September, down 37.4% year on year — a comparison flattered by the launch quarter a year earlier, so it proves little on its own. The quarterly split is more informative, because the launch distortion falls away inside the year: 3.82 million in April–June, then 2.67 million in July–September, a fall of 30% between consecutive quarters.

Set that quarter against the same quarter of the previous console's life. The Switch 1 shipped 3.19 million units in July–September 2018, its second year, from an installed base of 19.67 million. The Switch 2 shipped 2.67 million in July–September 2026, from an installed base of 23.68 million — 20% larger. It sold 16% fewer consoles.

Nintendo has been signalling softness outside Japan for some time. In February it told investors that "sales in Japan exceeded our expectations while overseas sales were somewhat weaker than expected" — and that the resulting mix shift, toward Japan and toward other publishers' software, would offset the benefit of a weaker yen. It then raised the price of the console in Japan in May and in other regions in September, which is not what a company does with unmet demand.

The channel agrees. Nintendo reports 26.35 million Switch 2 units sold in (sell-in) at 30 September and says cumulative sell-through passed 25 million by the end of October. Since sell-through at the end of September must therefore have been below 25 million, at least 1.35 million consoles — more than six weeks of shipments at the July–September rate — were sitting between Nintendo and consumers at the quarter end. That is a lower bound, and it is on the high side of normal for a console in its second year.

What the forecast now requires

None of this has changed Nintendo's guidance. On 5 November it reaffirmed the forecast it set on 8 May: 16.5 million Switch 2 units for the year to March 2027, alongside 60 million software units. The company's own slide puts the hardware forecast 16.9% below last year's 19.86 million.

The arithmetic inside that number is the tightest constraint in the filing. With 6.49 million shipped in the first half, Nintendo needs 10.01 million in the six months to March — 1.54 times its first-half shipments, and 5.4% more than the 9.50 million it shipped in the same half a year earlier, which contained the console's first Christmas.

Is that unreasonable? No, and it is worth saying so plainly, because the obvious version of this story — "the guidance is impossible" — does not survive contact with the company's own history. In its second year the Switch 1 shipped 11.88 million units in the October–March half against 5.07 million in the first half, a ratio of 2.34. Nintendo is asking the Switch 2 for a second-half ramp of 1.54, well short of what the previous console actually delivered. The holiday bet is large; it is not unprecedented.

Which returns the question to the inventory. Nintendo is carrying 1.7 times the cost of goods it booked in the entire first half — 309 days of cost of sales at the first-half rate, against 0.7 times and 128 days a year earlier. If the October–March half arrives as forecast, that stock converts into revenue, the forward purchase of components turns into a cost advantage over competitors buying at spot, and the record balance sheet looks like foresight. If it does not, Nintendo is holding ten months of cost of sales into a year in which it has already raised prices once and its overseas demand is below plan.

The gross-margin jump in that table is not, mostly, a margin story. Nintendo attributes part of the profit increase to refunds of US tariffs imposed under IEEPA that had previously been recorded in cost of sales. It does not disclose the size of the refund in the tanshin, so the underlying hardware and software margin cannot be separated from it, and the 14-point move should not be read as a hardware-margin achievement.

Two readings, and what would separate them

Reading one: a hedge. Nintendo bought memory and other components forward, at pre-spike prices, in the quantities a 16.5-million-unit year requires. The cash left the company in the first half; the goods arrive across the second. On this reading the inventory is an asset that has already appreciated, and Nintendo's hardware margin in the holiday quarter should hold up better than Sony's: Sony said in its annual report filed in June that it would flexibly adjust unit sales to protect margin, the opposite response to the same cost pressure.

Reading two: a build. Nintendo produced consoles ahead of the quarter in which it expects to sell them, as it did in 2018, when September inventory reached ¥326bn and the following half shipped 11.88 million units. On this reading the inventory is a finished-goods pipeline, and its size is a statement of how large Nintendo thinks the holiday will be.

The two readings have different fingerprints, and the next two filings will show which is right. A components hedge shows up as a falling cost of sales per unit in the third quarter even as prices elsewhere rise, and as inventory converting to revenue without a write-down. A finished-goods build shows up as a jump in sell-in during the December quarter and, if demand disappoints, as a valuation loss on inventories — a line Japanese issuers take when the carrying value exceeds net realisable value. Nintendo's first-half income statement carries no loss on valuation of inventories.

What this does not prove

Inventory is one line. Nintendo does not publish its composition quarterly. The claim that most of the increase is memory is an inference from Furukawa's February answer and from the timing; it is not disclosed. The claim that the balance is a record is limited to the September interim balance sheets compared here: 2018 through 2026, plus 2014 (¥126.8bn) and 2016 (¥49.1bn) from the same series. No interim balance sheet from the last twelve years is close to the current figure, but decades before that are not covered.

The seasonal build is not the largest in percentage terms. Nintendo builds stock every year between March and September. In 2018 the balance rose 130% over the same six months and in 2021 by 211%; this year's 51% is mid-range. What is unprecedented is the level, not the rate of increase.

Shipments are not sales. A 2.67 million-unit quarter measures what Nintendo invoiced to retailers and its own direct-sales channel, not what consumers bought. The sell-through disclosure — "exceeded 25 million" — is a threshold, not a figure, so the channel gap of 1.35 million is a floor, not a measurement.

The comparison quarter is not exact. The Switch 1's July–September 2018 quarter came 16 to 19 months after its launch; the Switch 2's came 13 to 16 months after. The Switch 2's quarter is therefore earlier in its life, which if anything makes a shortfall against the older console more notable, not less — but the two are not perfectly aligned.

The narrow claim this article makes is arithmetic on Nintendo's own filings: the company converted ¥288bn of cash and securities and ¥135bn of supplier credit into a ¥275bn inventory build, the largest balance in the nine years of comparable quarterly filings checked here, in the same half-year that its newest console's quarterly shipments fell below its predecessor's at a comparable age, and it reaffirmed a full-year forecast that requires the strongest October–March half the Switch 2 has had. Whether that is a hedge that pays or a bet that does not is a judgement. That it is the largest inventory position Nintendo has reported in at least nine years is not.

Method. All balance-sheet and income-statement figures are from Nintendo Co., Ltd.'s consolidated financial statements in the quarterly tanshin for the six months ended 30 September 2026 (file 261105e, published 5 November 2026), with comparatives from the same document and from the corresponding Q2 tanshin for each year 2018–2025 (181030e, 191031e, 201105e, 211104e, 221108e, 231107e, 241105e, 251104e). Unit figures are from the same tanshin series and from the Q1 FY2026 tanshin (260806e); the July–September figures are derived by subtracting first-quarter shipments from first-half shipments, as noted in the table. The 16.5 million unit forecast and the 60 million software forecast are read from the forecast slide of the 5 November presentation (261106e) and cross-checked against the 8 May 2026 full-year tanshin (260508e). The February 2026 quotations are from Nintendo's English translation of the nine-month Q&A (260204e). Days of cost of sales are inventory divided by half-year cost of sales, multiplied by 182. Nintendo's fiscal year ends 31 March; "FY2025" and "FY2026" in the tables are the years ending 31 March 2026 and 2027, matching Nintendo's own labels.

Sources

• Nintendo Co., Ltd., Consolidated Financial Results for the Six Months Ended September 30, 2026, 5 November 2026 — balance sheets at 31 March and 30 September 2026, income statement, segment and regional commentary.
• Nintendo Co., Ltd., Financial Results Explanatory Material, six months ended 30 September 2026, 5 November 2026 — working-capital and inventory statement, hardware and software unit forecasts, sell-through disclosure, price-change statement.
• Nintendo Co., Ltd., Nine Months Financial Results Briefing Q&A Summary, 3 February 2026 — memory procurement, component-cost impact on Q3 and Q4, regional demand commentary.
• Nintendo Co., Ltd., Consolidated Financial Results for the Six Months Ended September 30, 2025, 4 November 2025 — prior-year comparatives, H1 FY2025 shipments of 10.36 million.
• Nintendo Co., Ltd., Consolidated Financial Results for the Fiscal Year Ended March 31, 2026, 8 May 2026 — FY2025 shipments of 19.86 million, FY2026 forecast.
• Nintendo Co., Ltd., Consolidated Financial Results for the Six Months Ended September 30, 2018, 30 October 2018 — Switch 1 H1 FY2018 shipments of 5.07 million, September 2018 inventory of ¥326.1bn.
• Nintendo Co., Ltd., Q2 tanshin for the years ended 31 March 2020, 2022, 2023, 2024 and 2025 (files 191031e, 201105e, 211104e, 221108e, 231107e, 241105e) — the September inventory and first-half cost-of-sales series used in the ratio table.