What Gets Its Own Line

Two Kubota plants melt iron. The Onkajima Business Center, in Osaka's Taisho ward, casts engine crankcases and other iron parts. The Hanshin plant, in Amagasaki, casts water pipe. Both are moving from coke-fired cupolas to electric furnaces — Hanshin already has one installed — a change the company expects to remove roughly 26,000 tonnes of CO2 a year once the switch is complete. The integrated report mentions the two foundries in that context only, decarbonisation. What happens to the metal after it cools, machining included, does not appear at all.

None of that appears in the English record. Kubota filed twelve annual reports on Form 20-F with the U.S. Securities and Exchange Commission, covering the years to March 2002 through March 2013, and the history section of the last one runs to four facts: founded in 1890 by Gonshiro Kubota, incorporated in 1930 under the Commercial Code of Japan, listed in Tokyo and Osaka in 1949, listed in New York in 1976. The waterworks pipe that started the company appears nowhere in that account. Ductile iron pipe turns up only in the same section's description of what Kubota makes today, undated. Four dated lines are the whole chronology; how the company got from one business to the next is not narrated.

The Japanese filings are less reticent about the early years. Gonshiro Kubota set up Kubota Ironworks in Osaka in February 1890 and made and sold castings of all kinds; in July 1893, three years in, the shop began producing cast iron pipe for water supply. The company's separate online chronology adds that in 1897 it developed a split-mould inclined-pouring method and moved iron pipe into volume production — and dates the adoption of the name Kubota Ironworks to that same year, where the securities filing has been using the name since 1890. The chronology puts the casting method and the name in one entry, does not say that one led to the other, and gives no background for either.

The demand behind the 1893 move cannot be sized from the record. Japan's official water-supply penetration series begins in fiscal 1950, at 26.2%, and the ministry states outright that Meiji, Taisho and prewar coverage is not included; the market a pipe shop was building for in July 1893 is simply not measured. What the series does capture is post-war catch-up: 53.4% by fiscal 1960, 80.8% by fiscal 1970, 91.5% by fiscal 1980, with the steepest recorded five-year gain in fiscal 1955-60, up 17.4 points, roughly six decades after the shop started pouring pipe. Then it flattens — 98.1% in the year to March 2021, 98.3% to March 2023, 98.2% to March 2024, 98.3% to March 2025 — and the publisher flags a separate methodology annex for disaster-affected areas, so the tail is a plateau with seams in it. The chronology separately records successive iron pipe orders from the Netherlands in 1932: an export record, evidence of capability, not of demand at home.

The next line of business arrived in February 1922: small kerosene engines for farm and industrial use. Kubota dates its engine business specifically to production of the Type A unit and marked the centenary in 2022. In 1930 the Ministry of Commerce and Industry designated the Kubota kerosene engine a superior domestic product. In November 1937 the company opened the Sakai plant to build them in volume. When Kubota first put one of its own engines inside one of its own machines is not in the official chronology; the securities filing marks only much later plants — Utsunomiya, for rice transplanters and binders, in May 1969, and Tsukuba, for tractors, in August 1975.

The line non-Japanese readers actually recognise came third. Kubota developed a tiller and began making and selling it in 1947; the chronology gives no model or type. A riding tractor for upland fields was commercialised in 1960; paddy-field tractor production started in 1962; rice transplanters in 1968; combines in 1969, at a newly built plant in Utsunomiya. The verbs differ across those entries — developed, commercialised, began production — so it is a sequence rather than a release calendar. On the chronology's own dates, all of them production or commercialisation entries, the elements of a full rice-mechanisation line-up were in place within twenty-two years.

Read in order, the company's own chronology runs pipe first: cast iron water pipe from July 1893, the small farm-and-industrial oil engine from February 1922, the tiller developed and put into production in 1947. The engine entry precedes the machinery line rather than following it, which is the reverse of the order the 20-F uses when it introduces the businesses — farm equipment, then engines, then construction machinery. The timeline records only the order. It asserts no causal link, gives no first-tractor date, and notes that engine mass production began at a newly built plant in Sakai in 1937, so whether one melt shop and one set of moulders carried through each step is not something it establishes. What does run through all of it is casting: assorted castings in 1890, water pipe in 1893, heat-resistant cast iron in 1922.

Construction equipment followed a slower path. The company records entering construction machinery in 1953 — an entry that names no product — and starting mini-backhoe production in 1974, roughly two decades later; the archive treats the two as separate events and claims no industry first for the backhoe. Small construction machinery production began in Germany in 1989, the same year implement manufacturing began in the United States. Kubota bought the American implement maker Great Plains Manufacturing in July 2016, per the securities filing; the corporate archive gives the same purchase as a year with no month. Compact track loader production in North America started in 2022, which is the first construction machinery Kubota is recorded as building on that continent.

Overseas expansion began in 1957, and not with water: a farm machinery production base in Brazil, the company's first overseas manufacturing site. Water came three years later, a Phnom Penh waterworks contract in 1960, which the company describes as the first overseas water project by a Japanese firm. Then Kubota Tractor Corp. in the United States in September 1972 — a sales operation, not a plant — a French sales company in March 1974, the New York Stock Exchange in November 1976. Nothing pauses after that: small construction machinery in Germany and tractor implements in North America in 1989, China from 1998, Thailand's Siam Kubota taken as a subsidiary in 2004, tractor and ductile iron pipe plants in Thailand and India in 2009. The three purchases usually cited — Kverneland in March 2012, Great Plains in July 2016, Escorts in April 2022 — are the late, visible end of a longer list.

The New York listing ended in 2013. In the twelve months to May 2013, U.S. trading in the ordinary shares averaged 267,164 shares a day against 6,230,176 worldwide — 4.29%, under the 5% threshold that permits deregistration. The Form 25 went in on 5 July 2013, the Form 15F on 16 July. Neither document states a reason; the forms have boxes for rules, not for motives. What that closure means practically is that the SEC filing record stops a decade before Kubota's 2025 accounts put destination-based revenue from North America at ¥1,218,454 million (US$7,439 million) against Japan's ¥685,184 million (US$4,183 million). The direction in that single year runs against the easy story: North America down 4.2% from ¥1,272,503 million (US$7,769 million), Japan up 8.3%. The English record of the intervening decade lives in Kubota's own annual and integrated reports, not in Washington.

For the year to December 2025, group revenue was ¥3,018,891 million (US$18,430 million) and operating profit ¥265,470 million (US$1,621 million). Machinery accounted for 87.1% of external sales at ¥2,628,618 million (US$16,048 million); Water & Environment 12.4% at ¥374,352 million (US$2,285 million); Other 0.5%. Inside Machinery, the filing gives exactly two product lines: farm equipment and engines together at ¥2,003,307 million (US$12,230 million), and construction machinery at ¥625,311 million (US$3,818 million).

Note what that means. In neither filing do engines get a line of their own. The 2025 Japanese report splits the Machinery segment into agricultural machinery and engines on one side, construction machinery on the other, so an engines-only figure cannot be extracted from it. The 20-F did the same thing in English, placing them inside Farm & Industrial Machinery, formerly Internal Combustion Engine and Machinery, and showing farm equipment and engines combined at ¥399.4 billion (US$2,438 million) for the year to March 2003; the qualitative remark that year was that overseas growth came from sales to original equipment manufacturers in the EU and the United States. Research and development spending for 2025 was ¥110.3 billion (US$673 million) in total — ¥103.5 billion (US$632 million) in Machinery, ¥6.8 billion (US$41.5 million) in Water & Environment — with no engine figure inside either.

Kubota's own account of the engine franchise, given in a July 2022 release, is that it holds the world's top share in industrial diesels of 100 horsepower and below; that it has shipped more than 30 million industrial engines cumulatively, with no start or cut-off date stated and the 1922 starting point inferred from context; that it then built some 3,000 variants, without defining what counts as a variant; and that its strength is power density, compactness with high output, stated without a kW/L or kW/kg figure anywhere. For the share claim the release names no survey house, no reference year, no percentage and no geography, and does not define industrial. It is the company describing itself, about a product that has no line of its own in the segment tables.

The home market that funded all of this has been shrinking in one dimension and drifting in another. Japanese farm machinery shipments totalled ¥467.8 billion (US$2,856 million) in 2025, split ¥317.6 billion (US$1,939 million) domestic and ¥150.2 billion (US$917 million) export. Those shipment figures come from the manufacturers' association and cover its 58 member companies on a calendar-year basis, not the whole domestic market, which also contains non-members and imports. On the same basis domestic shipments ran ¥403.0 billion (US$2,460 million) in 2004 and ¥270.7 billion (US$1,653 million) in 2024, the low of the published series, before the 2025 figure; all nominal, and the source offers no explanation for the one-year rebound. Meanwhile the share of riding tractors rated 30 horsepower and under fell from 94% of shipments in 1975 to 37% in 2025, with 50 horsepower and above at 44% — a different body, the mechanisation association, counts those — as average cultivated area per farm entity grew from 1.02 hectares in 1978 to 3.60 hectares in 2024, on the agricultural census, national average, Hokkaido included.

Kubota's account of that year is more specific. Machinery revenue in Japan was ¥354.8 billion (US$2,166 million) for 2025, up ¥43.0 billion (US$262 million) from ¥311.9 billion (US$1,904 million), a rise of 13.8%, and because the sales are domestic the local-currency change is the same ¥43.0 billion. Management puts it down to rice prices lifting farm incomes, a replacement cycle that had stalled for years turning over, and production capacity assembled in a hurry to meet it. It also expects the reversal: the 2026 forecast has Japan down ¥14.8 billion (US$90.4 million) as the reaction to that replacement demand sets in. Kubota's ¥354.8 billion is larger than the ¥317.6 billion of domestic shipments the association counts for all 58 of its members, quoted above, and the two are counting different things: the association counts farm machinery, while a Kubota region is the machinery segment, which the company reports as farm equipment and engines plus construction machinery, and which leaves Water & Environment out entirely, because the regional split is given for machinery alone.

The first business, the one everybody skips, is not the one shrinking. Water & Environment revenue rose 3.2% in 2025 to ¥374,352 million (US$2,285 million) and segment profit rose 35.9% to ¥32,983 million (US$201 million). The company added nominal diameters of 500 to 1,000 millimetres to its GX-form ductile iron pipe line, previously 75 to 450, with the straight pipe in that range produced entirely in the Hanshin plant's electric furnace. The market underneath contracts as it goes: in the fiscal year to 31 March 2025 the population served by Japanese water utilities fell by 457,456 to 121,504,326, while penetration edged up to 98.3% — only because the total population fell faster, by 543,327.

For the year to December 2025 the disclosure does split something. Revenue was flat, ¥3,018.9 billion against ¥3,016.3 billion (US$18,430 million against US$18,414 million), up 0.1%, while operating profit fell ¥50.2 billion (US$307 million) and the margin went from 10.5% to 8.8%. The results presentation gives that fall as seven lines: exchange rates ▲¥1.4 billion (US$8.5 million); volume, mix and other ▲¥34.4 billion (US$210 million); raw materials, net, ▲¥3.2 billion (US$19.5 million); fixed costs and similar ▲¥23.4 billion (US$143 million); tariffs ▲¥65.0 billion (US$397 million); incentive rates +¥20.1 billion (US$123 million); price revisions +¥57.1 billion (US$349 million). The seven sum to the ▲¥50.2 billion exactly.

Two of those lines are larger than the number they explain. Tariffs at ▲¥65.0 billion exceed the whole decline by ¥14.8 billion (US$90.4 million); price revisions at +¥57.1 billion recover 87.8% of the tariff line, which is 57.1 divided by 65.0. Set the two self-help lines against the tariff line alone — 20.1 plus 57.1 against 65.0 — and the year was not lost to tariffs: tariffs were covered, and then some, and the profit went somewhere else. That is arithmetic on the seven lines, and not the company's arrangement of them. Kubota brackets tariffs with materials and fixed costs as cost increases of ▲¥91.6 billion (US$559 million), set against self-help of +¥77.2 billion (US$471 million), being incentive rates and price; on that pairing the cost increases are not absorbed, and ▲¥14.4 billion (US$87.9 million) remains, 28.7% of the fall. The volume-and-mix line, which sits in neither group, is 68.5% of it. The 2026 assumption is ¥90.0 billion (US$549 million) of tariffs, ¥25.0 billion (US$153 million) more, on annualisation.

What the bridge will not do is separate quantity from mix. That line reads as sales volume, product mix and other, one number for how many machines went out and what they were, and the split is not disclosed. It is not the largest line in the bridge — tariffs, at ▲¥65.0 billion, is — but at ▲¥34.4 billion it is 68.5% of the year's decline, and it is the one item of the seven that can be attributed to neither quantity nor mix. Price is separated, and separated from discounting, which has its own line. In the previous year's securities report the company explained a much smaller fall, ▲¥13.2 billion (US$80.6 million) or 4.0%, the other way round: price increases and currency as the positives, lost volume and rising incentive costs as the negatives. The incentive line changed sign between the two years, which is what a policy change looks like in a bridge; the president dates the revision to partway through 2025 and calls its effect on share relatively limited, without giving a share figure. Currency is ▲¥1.4 billion (US$8.5 million), 2.8% of the decline, and does not appear in the chief executive's account of the year at all. For revenue the company gives the change excluding currency, +¥16.0 billion (US$97.7 million), against a reported +¥2.6 billion (US$15.9 million); subtract, and the currency effect is ▲¥13.4 billion (US$81.8 million), a figure Kubota does not print.

Below the group line the same segment appears twice with two profits. The integrated report gives Machinery ¥253.6 billion (US$1,548 million), a 9.6% margin on machinery revenue of ¥2,628.6 billion (US$16,048 million), and Water & Environment ¥33.0 billion (US$201 million), 8.8%. The results presentation gives Machinery ¥226.1 billion (US$1,380 million), 8.6%, and Water & Environment ¥27.6 billion (US$168 million), 7.4% — and prints no segment revenue at all, so the two sets of profits cannot be laid on a common denominator from that document. The margins imply revenue close to the integrated report's, but implying is not disclosing. The presentation figures are labelled management basis and restated for a reorganisation effective 1 January 2026 that moved costs formerly held in adjustment into the segments, which is the direction the two sets differ in; the integrated report's two segments sum to ¥286.6 billion (US$1,750 million) against a consolidated ¥265.5 billion (US$1,621 million), a gap of ¥21.1 billion (US$129 million) for which that report carries no reconciliation table. Neither document splits profit by region. Revenue by region is given — North America 40%, Japan 23%, Asia 22%, Europe 12% — and so is headcount, and the two do not resemble each other: Japan holds 27,129 of the 52,503 employees, 51.7%, against 23% of revenue; North America 7,631, or 14.5%, against 40%. What that arrangement costs or earns is not a disclosed line.

ROIC appears three times in one integrated report with three definitions. In the plan pages it is after-tax operating profit including equity-method income, over shareholders' equity plus interest-bearing debt: 4.2% for 2025, against a 2030 target of 7% or more. In the body of the remuneration section it is after-tax operating profit plus equity-method income, over invested capital. In a footnote to the remuneration table it is profit attributable to owners of the parent over invested capital, and on that definition the three-year average for 2023 to 2025 is 5.09%, against a standard of 5.72%, a cap of 7.22% and a floor of 4.22%, producing a payout coefficient of 79%. Invested capital is not defined anywhere in the document. The numerators differ, one operating and one net of everything, and the periods differ, one year against three, so 4.2% and 5.09% are not two measurements of the same quantity. The 2026 forecast supplies a fourth number, 4.8%. On the plan call an analyst put the current figure at 3 to 4%; the company neither agreed nor disagreed.

The targets those definitions serve were published on 13 February 2026: a 12% operating margin, 12% ROE, ROIC of 7% or more, and ¥900.0 billion (US$5,495 million) of free cash flow cumulative over five years. Kubota also publishes what it assumes its capital costs — equity at 7 to 10%, built on a 2% risk-free rate, a 5 to 7% market risk premium and a beta of 1.0 to 1.2; WACC at 5 to 6.5%, on after-tax debt at 3%. Take the top of each range and the ROE target clears its hurdle by 2 points, the ROIC target by 0.5. The starting point is an 8.8% margin and 7.3% ROE, down from 9.5% and 8.8% in 2020, and the company's own decomposition of the ROE fall is margin plus asset turnover, 0.58 turns to 0.49, with no contribution figure attached to either. On cash the plan is a reversal rather than a step up: 2025 free cash flow was ¥164.2 billion (US$1,002 million), and the five-year total divided by five averages ¥180.0 billion (US$1,099 million) a year, 9.6% above it. The five years to 2025 cumulated to a negative number, which the company states in words and does not quantify.

What the plan does not do is bridge its own headline ratio. Total asset turnover goes from 0.49 to 0.64, a 31% improvement, by way of three levers — profitability, working capital, and holding down retail finance receivables — of which exactly one carries a number: North American product inventory down 30% in months of cover by 2030. There is no group inventory figure, no receivables figure, no asset disposal target, so the remaining 0.15 turns cannot be assembled from the document. Shareholder returns are given as an order rather than an amount: progressive dividend first, balance-sheet health second, buybacks third and flexible with cash flow. No total return ratio, no payout ratio, no DOE. The 2024 securities report had a number — total returns of 40% or more of profit attributable to owners, counting dividends plus the value of cancelled treasury stock — and the plan does not restate it; asked why, the chief executive said the cash has to show up first. Two months later, in April 2026, the board authorised up to ¥30.0 billion (US$183 million) and 15.0 million shares, which the disclosure puts at 1.3% of shares outstanding excluding treasury; the reference figure printed for that denominator, 1,138,716,846 shares, is dated 31 March, three weeks before the decision. The reason given runs to one sentence: part of shareholder returns, to raise the value per share.

One quarter of the plan now exists. In the first quarter of the December-2026 financial year, results published 8 May 2026, revenue was ¥810.0 billion (US$4,945 million), up 13.7%, operating profit ¥98.0 billion (US$598 million), up 59.1%, and the margin 12.1% against 8.6% — the first quarter of the first year printing 12.1% against the 9.5% assumed for the full year, and against the 12% the plan sets for 2030. The quarterly deck draws neither comparison: it does not mention the plan, or 2030, or ROIC, and its only reference to a medium-term target is a footnote on fixed costs. The bridge keeps the same seven lines: currency +¥23.9 billion (US$146 million), sales volume, mix and other +¥27.5 billion (US$168 million), the volume contribution alone again undisclosed, price +¥18.7 billion (US$114 million) and incentive rates +¥7.7 billion (US$47.0 million), against fixed costs ▲¥16.1 billion (US$98.3 million), tariffs ▲¥24.6 billion (US$150 million) and materials ▲¥0.8 billion (US$4.9 million). Currency is 65.7% of the ¥36.4 billion (US$222 million) increase, which is 23.9 divided by 36.4; excluding it the increase is ¥12.5 billion (US$76.3 million). The quarter's tariff line is a year-on-year increment, and four times the quarter is ¥98.4 billion (US$601 million) of increment against the ¥25.0 billion (US$153 million) of increment assumed in the full-year bridge; the ¥90.0 billion (US$549 million) quoted earlier is a different quantity, the assumed full-year cost of U.S. tariffs in total, and the quarterly line names no country at all. The single plan target with a quarterly read is the cap on fixed cost growth of 3% or less, and Kubota reports it missed, while noting that the controllable part, excluding logistics and accounting effects, grew ¥11.0 billion (US$67.2 million). Free cash flow was ▲¥12.7 billion (US$77.5 million) against ▲¥27.5 billion (US$168 million) a year earlier, improved by spending less on plant rather than by generating more from operations.

The two melt shops carry ¥17,333 million (US$106 million) of book value and 373 people at Onkajima, ¥15,226 million (US$93.0 million) and 583 people at Hanshin — parent-company figures, inside a consolidated group of 52,503 employees. One of them casts ductile iron pipe for waterworks that reached 98.3% of Japan's population as of 31 March 2025. The other casts crankcases. In the securities report's table of principal facilities, each of them gets a single line.

On currency: yen amounts are converted at ¥163.8 = US$1, the rate for 24 July 2026 as published by TradingEconomics. Unless otherwise noted, that single rate is applied to every period cited, regardless of when the yen figure was reported. The dollar figures are therefore restatements at one rate, not historical translations; because the rate does not move, the conversion adds no exchange-rate effect to year-on-year comparisons, though whatever currency effect is embedded in the underlying yen figures remains. Ratios, percentages, multiples, unit counts, headcounts and horsepower or kilowatt figures are left unconverted.