One of the Places Is Not a Place

The segment note in Shiseido's FY2025 annual securities report lists six lines and a total. Japan, external sales ¥295,343 million (US$1,803 million), 30.4 per cent of the total. China & Travel Retail, ¥342,244 million (US$2,089 million), 35.3 per cent. Asia Pacific, ¥73,290 million (US$447 million), 7.6 per cent. Americas, ¥106,584 million (US$651 million), 11.0 per cent. Europe, ¥141,129 million (US$862 million), 14.5 per cent. Other, ¥11,399 million (US$69.6 million), 1.2 per cent. Consolidated, ¥969,992 million (US$5,922 million). Five of those names answer the question where. One of them answers where and also how, joined by an ampersand.

The company defines the second half of that ampersand in its own management discussion: travel retail is the sale of cosmetics and fragrances at airports and downtown duty-free stores. It is a channel — a counter type, a customs status, a receipt. The same passage locates the segment's weak year in China and Korea, where spending by Chinese travellers stayed subdued. So the line named after one country books sales made in a second country, and the thing the two have in common is not longitude but a shopper.

This is new as of FY2025, and the merger was arithmetically clean. In the prior year's report the two were separate: China ¥249,952 million (US$1,526 million) and Travel Retail ¥107,834 million (US$658 million). The restated FY2024 comparative for China & Travel Retail is ¥357,786 million (US$2,184 million). Add the first two: 249,952 + 107,834 = 357,786. Not approximately. Exactly. Whatever else changed in the restatement, the sales of the merged pair are the old sales, summed.

Something else moved between the two presentations, and it takes a second document to see it. Restated FY2024 Japan is ¥294,272 million (US$1,797 million) in the February 2026 segment note, against ¥283,776 million (US$1,732 million) as first published in last year's securities report, a difference of ¥10,496 million (US$64.1 million). Restated Other is ¥15,663 million (US$95.6 million) against ¥26,158 million (US$159.7 million), a difference of ¥10,495 million (US$64.1 million) the other way. Asia Pacific, Americas and Europe are identical to the yen in both presentations. Japan and Other together are unchanged to within one million yen, which is what a reclassification between those two lines would look like. The note says only that the prior year is presented on the changed basis, and neither document breaks out what moved.

Which base the growth rates use is checkable, and they use the restated one. The FY2025 sales table shows Japan at 0.4 per cent: (295,343 − 294,272) ÷ 294,272 = 0.36 per cent, which prints as 0.4. It shows China & Travel Retail at −4.3 per cent: (342,244 − 357,786) ÷ 357,786 = −4.34 per cent. Run Japan against ¥283,776 million (US$1,732 million), the figure printed in the prior year's report, and you get (295,343 − 283,776) ÷ 283,776 = 4.1 per cent, which is not the number in the table. The restated column is the base the printed percentages are consistent with.

The profit measure moved at the same time. The note to the restated comparatives says the prior year is presented under both the changed segmentation and the changed method of calculating segment profit; headquarters costs are now carried in the Adjustment line. As first published, the FY2024 segment lines totalled ¥30,364 million (US$185 million) with an Adjustment of positive ¥5,995 million (US$36.6 million). Restated, the six lines add to ¥95,042 million (US$580 million) with an Adjustment of negative ¥58,683 million (US$358 million). The swing is ¥64,678 million (US$395 million) on each side, and consolidated core operating profit is ¥36,359 million (US$222 million) both times.

For the merged line this produces a segment with two pasts. FY2024 China plus Travel Retail, as first printed, was ¥12,271 million plus ¥5,006 million, or ¥17,277 million (US$105 million) of core operating profit. FY2024 China & Travel Retail, restated, is ¥71,979 million (US$439 million). FY2025 is ¥64,525 million (US$394 million). Against the restated base the change is −¥7,454 million (−US$45.5 million), and the company's narrative says the segment fell ¥7.5 billion (US$45.8 million) year on year, which reconciles. Against the older pair the same segment would appear to have gained ¥47,248 million (US$288 million). The restated base is the one the note tells you to use.

Denominators need the same care. The margins printed beside each segment are struck on sales including intersegment transfers, not on the external sales in the composition table. China & Travel Retail: 64,525 ÷ 345,662 = 18.67 per cent, printed as 18.7. Japan: 38,972 ÷ 296,450 = 13.15 per cent, printed as 13.1. A reader who divides the same profit by external sales of ¥342,244 million (US$2,089 million) gets 18.85 per cent instead — a small gap, and one that widens or narrows with how much a segment sells to its siblings.

The clearest evidence that the axis is not geography comes from the company's own forecasting. Sizing the effect of strained Japan–China relations on the first quarter of 2026, it gives a single figure of about ¥10.0 billion (US$61.1 million) — for China & Travel Retail and Japan's inbound sales combined. One exposure, one customer group, two segment lines, added together because that is the only way to state it.

Run that backwards through the composition table and the shares stop being cartography. Japan's 30.4 per cent includes inbound sales, which the company adds to the China & Travel Retail line when it sizes that single exposure. The China & Travel Retail 35.3 per cent covers the counters the company describes in China and Korea. Asia Pacific's 7.6 per cent is what remains of Asia after those two lines have taken what they take. The figures reproduced here do not separate inbound from domestic inside Japan, and no arithmetic on them will.

The Europe line has its own gap between name and contents, and the filing says so directly: the segment note states that the Europe business includes the Middle East and Africa. The English supplemental data uses the label EMEA and the Japanese report uses 欧州事業, which translates as Europe. Same ¥141,129 million (US$862 million), two names, one of which describes the contents and one of which describes a subset of them.

That matters more here than at most issuers, because of who holds the shares. On the register at 31 December 2025, foreign corporate holders account for 1,871,481 units and foreign individuals 5,860 units, out of 3,994,323 units in total: 1,877,341 ÷ 3,994,323 = 47.00 per cent. The form prints each row's percentage — 46.85 for the corporate line, 0.14 for the individual one — and has no total row, so the 47.00 is the reader's own addition, and it is not the sum of the printed rows, which is 46.99. Two readings of that individual row differ, 0.14 in one and 0.15 in another, and 5,860 ÷ 3,994,323 = 0.1467 per cent, which is what both of them are a version of. Across the 3,167 companies we have measured the same way, the median is 7.31 per cent. Nearly half the register, by units, is held in foreign names, and the English filings those holders would turn to are where the label reads EMEA.

Other is not a geography either. The segment note describes it as including the restaurant business. It is also where most of the group's production is booked: in the production table of the same securities report, stated at manufacturing cost, Other shows ¥125,146 million (US$764 million) of a ¥201,046 million (US$1,227 million) total, or 62.2 per cent, while the Japan line shows a dash. A reader moving through that document meets Other on page 36 as 1.2 per cent of sales — 11,399 ÷ 969,992 = 1.18 per cent, printed as 1.2 — and meets it again on page 46 as most of what the group makes.

Adjustment is the other line that behaves like a place. In FY2025 the six segments total ¥99,700 million (US$609 million) of core operating profit and Adjustment is negative ¥55,179 million (US$337 million), against a printed consolidated ¥44,520 million (US$272 million). On capital expenditure, the headquarters function carries ¥20,442 million (US$125 million) of a ¥42,974 million (US$262 million) total: 20,442 ÷ 42,974 = 47.6 per cent. Group research and development is ¥27,100 million (US$165 million), or 2.8 per cent of sales, and the report states that research activity is not tied to a particular segment and is therefore not broken out by one.

There is a China-labelled number elsewhere in the same document, on a different basis. The related-companies section names two subsidiaries whose sales exceed 10 per cent of consolidated sales: Shiseido Japan at ¥272,740 million (US$1,665 million) and Shiseido (China) Investment at ¥172,833 million (US$1,055 million), both stated net of intra-group sales. Set beside the segments, the first is 92.3 per cent of the Japan line and the second is 50.5 per cent of China & Travel Retail. The note flags that this is a different disclosure section from customer or geographic reporting, and it is: an entity is not a segment, and subtracting one from the other produces nothing.

Headcount follows the segment names too, which is where a channel starts to look like a workplace. At 31 December 2025 China & Travel Retail employs 5,982 people, with an average of 110 temporary staff on top. A year earlier, under the old labels, China employed 5,946 and Travel Retail 550, or 6,496 combined. Those two figures are a year and a segmentation apart, so the difference is not a measurement of anything; it is two counts taken on two different maps.

The axis is not only presentational. For impairment testing the company sets the cash-generating unit at segment level rather than brand level, stating that this follows how resources are allocated and performance evaluated, and the Drunk Elephant and Dr. Dennis Gross Skincare trademarks are tested inside the Americas unit. Goodwill in that unit was written down by ¥46,818 million (US$286 million), reported as a third-quarter 2025 event; the Americas goodwill balance is shown at ¥58.4 billion (US$357 million) at end-2024 and ¥9.7 billion (US$59.2 million) at end-2025. Where the line is drawn determines what is tested against what.

One housekeeping note for anyone re-adding the tables. The six FY2025 external sales components as printed come to ¥969,989 million against a printed total of ¥969,992 million (US$5,922 million either way), a difference of ¥3 million (about US$18,000); the restated FY2024 components fall three million short of their printed total the same way, and the FY2024 components as first published fall four short. The profit side goes the other way by one: the six segment lines add exactly to the printed subtotal of ¥99,700 million (US$609 million), and 99,700 − 55,179 = 44,521 against a printed 44,520. Every figure is printed in millions of yen, and the components come up short of the printed sales totals in all three presentations — the direction truncation would produce rather than symmetric rounding, though none of the documents quoted here states which convention is used.

The 2026 guidance arrives on the same axis and mostly above it: consolidated net sales of ¥990,000 million (US$6,044 million) and core operating profit of ¥69,000 million (US$421 million), with segment-level guidance given as growth ranges — mid-single-digit growth for Japan, a low-single-digit decline for China & Travel Retail, high-single-digit growth for Asia Pacific, Americas and Europe, and more than a 30 per cent decline for Other. The consolidated forecast is a number. The segments are directions.

So the table adds up, in the sense that matters to an auditor: components agree with printed totals to within a few million yen, the restated year reconciles to the reported year, and segment sum less adjustment lands on consolidated within one. What it will not do is the thing readers use it for. Add the five geography-shaped shares and you get 98.8 per cent of sales, which is a share of revenue and not a distribution across the world, because one of the five is a way of buying, one of them contains the Middle East and Africa as well as Europe, and the sixth carries ¥125.1 billion (US$764 million) of the group's ¥201.0 billion (US$1,227 million) of production while being described as restaurants and other businesses. The company adds two of the lines together itself — China & Travel Retail plus Japan's inbound sales, about ¥10.0 billion (US$61.1 million) in the first quarter of 2026 — when it wants to state a single risk. That is the instruction, and it is in the document.

On the arithmetic: yen amounts are converted at ¥163.8 = US$1, the yen–dollar rate on 24 July 2026 as published by TradingEconomics, applied uniformly to every period cited here including the FY2024 comparatives and the 2026 forecasts. The rate is an editorial convenience of this publication, not a company disclosure and not a measure of anything that happened in currency markets; the company's own 2026 planning assumptions are stated separately in its results release as ¥150 to the dollar, ¥170 to the euro and ¥20.5 to the yuan, and none of the dollar figures above are its numbers. Significant figures follow the original yen disclosure — amounts printed in millions convert to millions, amounts printed in billions convert accordingly. Percentages, ratios, multiples and headcounts are not converted.