The Same Heading, Thirty-Five Answers

The heading is six characters long: 指標及び目標, indicators and targets. It sits inside the sustainability section that Japan's Financial Services Agency added to the annual securities report by amending the Cabinet Office Ordinance on Disclosure of Corporate Affairs — promulgated and brought into force on 31 January 2023, applied to fiscal years ending on or after 31 March 2023. The amendment made two of the section's four parts, governance and risk management, compulsory for every filer, and left the other two, strategy and indicators and targets, to be written according to materiality — with one exception: on human capital, a policy for developing people and for the working environment, and indicators and targets for both, are required of everyone. So the heading is not guaranteed to appear under the sustainability section of every report, and it appears in all thirty-five read here. The amendment also brought three figures into the securities report — the ratio of women in management, the take-up of parental leave by men and the gender pay gap — for filers and their subsidiaries that already publish them under the laws on women's participation and on childcare leave. That is the frame. This piece reads what thirty-five of the largest listed companies put under the heading in the reports they filed in June 2026.

As this desk extracted them, the heading appears 195 times across the thirty-five reports — once in six of them (Nintendo, SoftBank Group, Keyence, HOYA, Takeda, FUJIFILM Holdings), nineteen times in Oriental Land's, sixteen in Sumitomo Mitsui Financial Group's, fourteen in Recruit Holdings'. It is not even spelled the same way. Twenty-eight companies write 指標及び目標 at least once; ten write 指標と目標; three write 指標および目標; ten write 指標・目標, FANUC alone inside square brackets. Twelve companies use more than one spelling within a single report. Recruit uses four; Honda and Oriental Land three.

Nintendo's answer is three short paragraphs. In the desk's translation of the first: because the products and content it handles are entertainment goods, and by their nature carry many uncertain factors in the running of the business, it does not set specific management indicators as targets, so that management can make flexible judgements in response to the results of its efforts and to changes in the business environment. The second and third add: sustainability issues are addressed by drawing up and carrying out annual plans, and the target is to carry out the planned activities without fail; on human capital, indicators are set for checking the state of efforts, and the details are in Part 4, the state of employees. The passage contains no percentage sign.

Keyence's answer is two sentences. It has set the rate of positive responses in its engagement survey as an indicator, with a target of 70% or more, and it is achieving it; this is a target for the parent company alone. That is the whole of it. The same report's employee table, read in an earlier piece on this desk, prints the highest average annual salary in this set, 21,783,259 yen across 3,306 employees. The company that prints the largest pay figure prints the shortest answer under this heading.

SoftBank Group's answer has a target and no percentage: carbon neutrality by fiscal 2030, a group target set in June 2022, covering scope 1 and scope 2 emissions of SoftBank Group Corp. and major subsidiaries including SoftBank Corp. and LY Corporation, with the emission figures on the group's website. Sony Group's report carries the heading twice, and both times it is a signpost — once as part of a cross-reference, once to say that the indicators for its workforce are in Part 4. Mizuho's report carries the heading four times; two of the four are cross-references, and the passage under the heading proper is cut short by the extraction at the first numbered sub-heading — a limit of the method, described below, not of the report. Recruit Holdings' fourteen occurrences include a climate section headed 指標・目標と実績 — indicators, targets and results — whose indicator is absolute scope 1, 2 and 3 emissions under the GHG Protocol.

A third kind of answer names another document. Daiichi Sankyo writes that the long-term targets, the KPI indicators and the fiscal 2025 results for each materiality are scheduled to be published on its corporate website in July 2026, and that the method of disclosing indicators for four of its strategic themes is under consideration, with publication scheduled for the second quarter of 2026. ITOCHU writes that the details of its sustainability action plan — the issues, the approach, the outcome indicators and the degree of progress — are in the ESG Report 2026, scheduled for September 2026; its greenhouse-gas targets, net zero by 2050 and a 75% reduction from 2018 by 2040, are in the securities report itself. East Japan Railway says its fiscal 2025 results will appear in the JR East Group Report 2026, marks with a star the indicators that KPMG AZSA Sustainability assured in the integrated report, and adds a sentence worth keeping: the securities report is not a disclosure medium subject to third-party assurance.

Takeda's answer is a table with an expiry date. Since fiscal 2022 it has disclosed what it calls corporate philosophy metrics, drawn up by employees in the relevant departments and visible to staff on an internal portal. The table gives, for fiscal 2024 and fiscal 2025, a combined count of regulatory approvals and pivotal clinical trial starts of 29 and 33 — the fiscal 2025 figure being 23 approvals and 10 trial starts; the share of completed trials whose result summaries were posted on public registries in time, 100% and 100%; and orders shipped in the quantity and by the date specified, 99.5% and 99.6%. Then the report says that from fiscal 2026 disclosure under this framework will end, and that key ESG indicators will be disclosed mainly in the ESG data book. The reader of next June's report will find the heading; whether the table is under it is a question the company has already answered.

Where the answer is a climate target, the grammar is shared: a base year, a target year, a percentage. JR East gives 2.65 million tonnes of CO₂ in fiscal 2013 for the group, a target of 1.33 million by fiscal 2030 (a 50% reduction), and an actual of 2.18 million in fiscal 2024 (17.7% down), then 60% by 2035, 73% by 2040 and net zero in 2050. NTT gives an 80% reduction in scope 1 and 2 emissions by fiscal 2030 against fiscal 2013, with a preliminary fiscal 2025 figure of 2.041 million tonnes, 56% down; and net zero across scopes 1 to 3 by fiscal 2040, with 20.32 million tonnes, 28% down from fiscal 2018. FUJIFILM Holdings gives a 50% cut by the end of fiscal 2030 against fiscal 2019 on both the product life-cycle measure and its own scope 1 and 2 emissions, with progress of 8% and 18% at the end of fiscal 2024, and a table in which scope 1 and 2 emissions fall from 1,124 thousand tonnes in fiscal 2019 to an estimated 862 in fiscal 2025 — a 23% reduction, which the table labels as 47% of the way to the target. FANUC gives zero scope 1 and 2 emissions by 2050, 42% down by 2030 against 2020, and a 12.3% cut in the use-phase emissions of sold products over the same period. Shin-Etsu gives net zero in scope 1 and 2 by 2050, and an interim target that reads not as a cut but as a level: greenhouse-gas emissions per unit of output brought to 45% of the fiscal 1990 figure by fiscal 2025 — a 55% reduction — with the index printed beside it at 57.0, 56.9 and 54.7 for the three latest years; on 28 May 2026 it published a new interim target, a 30% cut in emissions per unit of output by 2035 against 2025. Mitsubishi Heavy Industries gives a 50% cut in scope 1 and 2 by 2030 against 2014 and net zero by 2040.

The base years do not match. Shin-Etsu measures from 1990; JR East and NTT's scope 1 and 2 from 2013; Mitsubishi Heavy from 2014; NTT's scope 3 and ITOCHU from 2018; FUJIFILM from 2019; FANUC and Honda from 2020. Kyocera measures its renewable-energy target from the fiscal year ended March 2014 and its greenhouse-gas target from the fiscal year ended March 2020. NTT explains its two choices in a note: 2013 matches the base year of the Japanese government's plan for global warming countermeasures; 2018 is the first year for which its current consolidation scope, including overseas group companies, was calculated. A 50% reduction and a 46% reduction are not on the same ruler until the base year is carried across, and the base year is in the filing.

Honda's answer is a table that changed during the year. The report says that in the fiscal year and in the period up to the filing date, the company revised its targets for the fiscal year ending March 2031. The product CO₂ intensity reduction target for motorcycles went from 34.0% to 15.0%; for automobiles from 27.2% to 13.6%; for power products from 28.2% to 13.4%, each against the fiscal year ended March 2020. The report gives its reason in its own words: the changes follow a review of the powertrain portfolio and the product launch plan, owing to changes in the market environment and in the direction of trade policy. It also says the group has until now used the sales ratio of electrified products as an indicator, and that it will instead widen its activity to the reduction of greenhouse-gas emissions across society as a whole, shifting its management indicator to the reduction rate of total life-cycle emissions, with the fiscal year ending March 2036 as the target year. The revised table also carries a 46% cut in total corporate CO₂ against fiscal March 2020; a 12% cut in industrial water intake; a 20% cut in industrial waste sent to incineration or landfill, an indicator the report says was changed from a total-waste measure against business-as-usual; and a new KPI, a 30% share of recycled and biomass materials in Japan-built motorcycles for Europe and in Japan-built electric cars. Beside the line for total product CO₂ the table prints: not disclosed.

The human-capital numbers are where the ordinance was most specific, and the answers are specific in return. Toyota's table for the parent company gives 455 women in management as of 31 March 2026, against a target of five times the 2014 level by 2030 — the actual is 4.5 times — and a target of an average male parental-leave take-up of 85% or more over the five years from April 2025 to March 2030, against 79.0% in fiscal 2025. Toyota adds that it is now collecting the human-capital materiality indicators from consolidated subsidiaries to establish the actual state. JR East gives an engagement-survey positive-response target of 70.0% or more for the parent company by the fiscal year ending March 2032, against 62.5% in fiscal 2025; a male parental-leave target of 85% by the end of fiscal 2027, against 80.3% for the parent and 81.1% consolidated; 252 mid-career hires for the parent and 1,005 consolidated; a target of 10% or more women in management, against 9.0% at 31 March 2026; a target of 35% or more women among new hires, against 30.1%; and a retention target of 85% for women hired ten years earlier, against 80.2%.

Oriental Land's answer is a framework. The group has identified seven ESG materialities to prioritise by 2035, selected on a double-materiality principle — an environmental and social impact view and a financial impact view — and for each it has drawn up risks, opportunities, a strategy, and indicators and targets, monitored through its sustainability governance. Employee happiness, children's happiness, diversity, equity and inclusion, supply-chain management and climate change are among them. The heading appears nineteen times in the report because the framework repeats it for each materiality and again in the tables that follow.

A crude ruler for the thirty-five: the number of percentages — a figure followed by a percent sign — in the passages under the heading, as extracted. A percent sign in a column label is not counted. Each printed percentage is counted once, even where the heading repeats and the extracted passages overlap. Three reports have none — Nintendo, SoftBank Group, Mizuho. Three have one — Keyence, SMC, Nippon Steel. Mitsui & Co.'s passages have 103, JR East's 34, Oriental Land's 33, ITOCHU's 24, Mitsubishi Corporation's 23, Honda's 22. Kyocera's twenty include a renewable-energy target it says it reached early — a twentyfold increase in renewable-energy adoption against fiscal March 2014, planned for fiscal March 2031 and achieved in fiscal March 2025, replaced by a new target of 60% or more renewable electricity by fiscal March 2031 — alongside 46% reductions in scope 1 and 2 and in scope 1 to 3 emissions by fiscal March 2031 against fiscal March 2020, and carbon neutrality in fiscal March 2051. A percentage is not a target, and a count of them is not a measure of ambition. It measures how much of the answer was written in numbers.

What a reader can take from the heading is narrower than its name suggests and more useful. It is the place in a document filed under law where a company states the yardstick it has chosen to be measured by — an engagement survey, a renewable-electricity share, a headcount of women managers, a count of regulatory approvals — and the yardsticks are not standardised. Three things travel with each number and are printed beside it: the base year; the scope, which is the parent company for Keyence and for Toyota's table, and a named list of subsidiaries for SoftBank Group; and the year the number was last measured. Carry those three and the thirty-five answers can be compared. Drop them and they cannot.

What this reading cannot do. The extraction takes the text after each occurrence of the heading, in any of its four spellings, up to the next numbered heading, from page text with the whitespace removed, keeping at most 3,500 characters per occurrence — and where no numbered heading ends the passage on its own page, continuing into the first 1,500 characters of the next page and stopping there — and it treats a table as a run of characters; where a heading repeats, the passages overlap, so each printed percentage is counted once by its position on the page. The 195 occurrences and the percentage counts describe that extraction, not the documents' intent. A first pass that matched only three spellings missed the 指標・目標 headings of ten companies, Recruit's climate section among them, and would have reported them as absent; the counts here are from the second pass. A target may sit under another heading — the ordinance requires the human-capital figures in Part 4, and Nintendo and Sony point there — so the absence of a percentage sign here is not the absence of a target in the report. And a heading that appears eleven times may be one table repeating its column label.

Nintendo's three paragraphs and Honda's revised table are both answers the form permits. The ordinance asked for indicators and targets according to materiality; thirty-five companies decided what was material to them and wrote it under the same six characters. The form did not produce uniformity. It produced a record of thirty-five decisions, dated June 2026, that can be read next June against what they have become.