Tokuyama: long-term performance & turning pointsSales (revenue) and profit-margin ratio
Sales (¥ bn)Net margin (%)
1918Alkali for a country that had to import it
Revenue (¥ bn, bars) · net margin (%, line)
Source: securities reports
1918Iwai Katsujiro founds Nihon Soda Kogyo at Tokuyama; ammonia-process soda ash
1931First dividend in thirteen years
1932Exits soda ash; the same plant switches to caustic soda
1936Renamed Tokuyama Soda
The company was founded on a national shortage rather than a founder’s ambition. In February 1918 Iwai Katsujiro, who ran the trading house Suzuki Iwai Shoten, set up Nihon Soda Kogyo in the town of Tokuyama, Yamaguchi, with capital of ¥2 million, to make ammonia-process soda ash — the alkali under plate glass, soap and textiles, which Japan then bought abroad. Tokuyama offered limestone, salt, abundant process water and a harbour in one place, the four conditions a heavy chemical plant needs; the trademark, the initials N-S-K enclosed in the hexahedron of a salt crystal, fixed the founding product in the company mark.
The market broke almost immediately. Dumping by foreign producers destroyed the economics, and around 1932 the company effectively abandoned soda ash and turned the same salt-processing plant to caustic soda, catching the rising rayon industry; a 5% dividend returned in 1931 after thirteen years without one. In 1936 it renamed itself Tokuyama Soda, writing its birthplace and its product into the name, and broadened into magnesium carbonate, calcium chloride and sodium silicate. Within eighteen years the founding cause had been quietly dropped and the plant repurposed — the pattern the next century would repeat.
1938First cement kiln, fed by waste lime from alkali production
1949Listed on the Tokyo Stock Exchange
1952Electrolytic caustic soda; chlorine as co-product
1964Tokuyama-Shinnanyo petrochemical complex
1967In-house ion-exchange membrane
1970Own polypropylene production
1973Oil shock exposes a resource-intensive structure
Cement was not a new venture but a disposal problem. Alkali production left waste calcium carbonate, and in March 1938 the Tokuyama works installed its first wet-process Portland cement kiln to consume it. Kilns were added through the war and the recovery until a single site had become one of the largest cement plants in Japan, riding postwar public works until cement stood beside soda as a second earnings pillar. A waste stream had turned into a business — and the decision to pile every process onto one location set the shape of everything that followed.
Chlorine repeated the logic. A new electrolytic caustic-soda plant in 1952 produced chlorine as a co-product, and with the founding of the Tokuyama-Shinnanyo petrochemical complex in 1964 that chlorine could be piped straight to neighbouring plants instead of stranded. The build-out went through joint ventures rather than solo investment: Nihon Godo Hiryo with Asahi Glass and Toyo Soda in 1963, Shunan Petrochemical with Toyo Soda in 1964 with capital of $1.4M (¥500m), Sun Arrow Chemical with Tekkosha and Daicel in 1966 to enter PVC, and own production of polypropylene from 1970. Chemicals, cement and resins — the three pillars were assembled along a pipe network.
In 1967 the company developed its own ion-exchange membrane, which carried Japan’s salt industry through the forced shift away from the mercury process and, in time, made Tokuyama a technology licensor for chlor-alkali electrolysis worldwide — a first taste of escaping commodity competition through process patents rather than tonnage. Then the oil shocks of 1973 and 1979 struck a business built on naphtha, power and coal. Soaring input costs squeezed all three pillars at once and produced the internal consensus that would define the next decade: get out of resource-intensive products.
1984Polysilicon production begins at the East Plant
1985Kashima plant opens
1988Five-year plan: fine chemicals, electronics, medical
1989Tsukuba laboratory; General Ceramics acquired (aluminium nitride)
1994Renamed Tokuyama Corporation
2009First postwar net loss; Tokuyama Malaysia incorporated
The escape came as three consecutive years of product launches: plastic lens materials and desiccants in 1982; high-purity isopropyl alcohol and methylene chloride for the electronics industry in 1983, aimed at replacing CFCs and ethane in semiconductor cleaning; and, in July 1984, polysilicon at the East Plant. The bet was that purification know-how accumulated on soda could be aimed at niches where purity, not tonnage, set the price. A new Kashima plant followed in 1985, and the five-year plan begun in 1988 formally designated fine chemicals, electronics and medical as the businesses to grow — pulling in the dental-materials venture founded in 1978 (later Tokuyama Dental), the Tsukuba laboratory opened in 1989, and the aluminium nitride technology acquired that year with the US firm General Ceramics. Every earnings pillar of the 2020s was seeded inside this single decade.
In April 1994 Tokuyama Soda became simply Tokuyama. The official reasons were that the business no longer fitted the word “soda” — and that younger Japanese could no longer read the archaic characters 曹達 in the old name. Consolidated sales reached about $1.8B (¥168bn) in the year to March 1995, and a string of subsidiaries followed the customers abroad: the diagnostics company A and T in 1994, Shin Dai-ichi Vinyl in 1995, Singapore and Taiwan in 1996, Zhejiang and Shanghai in 2005, the aluminium-nitride heat-dissipation venture TD Power Material in 2007, Korea in 2008. Meanwhile soda and cement flattened and began to shrink.
Then the financial crisis produced a number the company had never seen: a net loss of $59.9M (¥6bn) in the year to March 2009, its first since the war. It showed that twenty years of new ventures had added variety without displacing the commodity cycle underneath. Two answers were alive inside the company — expand polysilicon into the semiconductor and solar boom, or restructure the commoditised chlor-alkali and cement businesses. In August 2009 it took the first: Tokuyama Malaysia Sdn. Bhd. was incorporated. The second would wait years.
2009Tokuyama Malaysia established for solar-grade polysilicon
2013Net loss of $388.3M (¥38bn) as the silicon price collapses
2014Yokota Hiroshi succeeds Kogo Kazuhisa as president
2016Net loss of $924.3M (¥101bn) — the largest in company history
2017Malaysia sold in full to OCI; complete exit
Tokuyama Malaysia was to carry the purification technology built at Tokuyama abroad and supply solar-grade polysilicon in volume, into a market that Chinese panel makers were expanding and post-crisis renewable investment was funding. Silicon purification is decided by capital cost and the price of electricity, and Malaysia offered cheap power and cheap land. Cumulative construction ran to the hundreds of billions of yen — the largest single investment in the company’s history — and the choice of commodity solar grade rather than advanced semiconductor grade would decide the size of what came next.
The plan never ran on time. Construction delays and commissioning trouble held the plant short of full output while Chinese oversupply collapsed the solar polysilicon price, so the product arrived precisely into the glut. The year to March 2013 brought an extraordinary loss of $330.9M (¥32bn) and a net loss of $388.3M (¥38bn). Losses did not stop there: after the presidency passed from Kogo Kazuhisa to Yokota Hiroshi in June 2014, the company posted a net loss of $539.6M (¥65bn) for the year to March 2015 and then $924.3M (¥101bn) — with extraordinary losses of $1.2B (¥126bn) — for the year to March 2016, against group sales that at their peak were around ¥300 billion. A plant, once started, cannot easily be stopped.
In May 2017 Tokuyama sold every share of Tokuyama Malaysia to OCI of South Korea and left the business entirely, repudiating its own 2009 design rather than holding an impaired plant and waiting. What survived the write-offs mattered as much as what was written off: advanced semiconductor-grade polysilicon at Tokuyama sat outside the Chinese price war, the Omnichroma dental composite launched in 2013 was gaining share in the United States, and aluminium nitride heat-dissipation materials were maturing. The company reorganised around two axes — legacy chemicals and cement, and growth in electronic materials and life science.
2023Returns to Malaysia in a 50:50 venture with OCI, semiconductor grade only
2026Domestic cement sales transferred to Taiheiyo Cement; production halt under study
Recovery ran on the seeds planted in the 1988 plan rather than on anything new. Research money went to advanced semiconductor-grade polysilicon and dental materials, and logistics were brought in-house with the purchase of Tokuyama Kairiku Unso in 2018. Had the crisis years consumed those lines along with Malaysia, the growth drivers of the following decade would have gone with them.
In 2023 the company went back to Malaysia with the buyer of its old plant: a 50:50 joint venture with OCI, capitalised at the equivalent of about US$218 million, to make semiconductor-grade polysilicon, with a Vietnamese subsidiary handling the final process. The scope is deliberately narrower than in 2009 — the venture makes an intermediate, while each parent keeps finishing and selling to its own customers — so the partner’s land and equipment are used without surrendering the customer. Caution is not the same as a small bet, though: the venture was set up fifteen months late, and splitting the process across borders means any one link can stop the whole chain.
The last commodity pillar is now going too. In 2026 Tokuyama agreed to transfer its domestic cement and solidification-material sales business to Taiheiyo Cement and began studying an end to production. Cement had been loss-making for years; what had kept it alive was its role in the internal loop at Tokuyama, burning other businesses’ waste as raw material. President Inoue Tomohiro’s argument for acting now was not that the business had reached the end of its life but that a buyer would not stay available forever — once the site is down to a single kiln, it can no longer do the waste-processing job, and by then no one would take it. Some 2.8 million tonnes of CO2 a year disappear with it; what replaces it on the site has not been said.
The opening scale — 200 tonnes a year of polysilicon — says a good deal about the character of this entry. The proposal came not from a business division but from the planning department, and Murakami Shoji pointed at inorganic polysilicon while openly admitting that his own field was organic chemistry and that the theme did not lie there. All the company had to hand was a byproduct feedstock and the purification experience accumulated on soda. Even so it fixed a deadline of two and a half years before anything else, choosing speed over completeness — which conveys how urgently it wanted out of resource-intensive products.
The entry did not produce only good outcomes. Aluminium nitride, commercialised in the same period, stayed stuck in narrow applications and ran below cost for long enough to be called a burden inside the company. Polysilicon itself, once volume was pursued in commodity grade through the Malaysian investment from 2009, ended in withdrawal. What was chosen in 1984 was the single point of high purity; which applications to serve, and how far to chase volume, is a question later managements have had to answer again and again. The choice made at entry did not settle the forks that followed.
At the centre of this decision lay a fork: hold on to a plant that survived impairment, or crystallise the loss and let it go. The confidence of a period when solar demand was swelling and polysilicon was a material with margins above 50% carried the company into the largest overseas investment in its history without preparing for the downside of mass Chinese entry and price collapse. Equipment once brought on line cannot easily be shut down, and losses kept accumulating after the assumptions behind the market had vanished. Repudiating the 2009 design itself and stepping out of the solar-grade market is what gives this withdrawal its weight.
That said, the decade in Malaysia left more than losses. In the shadow of the enormous write-downs, two other seeds — silicon for advanced semiconductors, and dental materials — survived to become the leads of the rebuilding years. Within the same polysilicon, whether to chase scale in commodity grade or to narrow the application at the leading edge was what separated profit from loss, and that fact became the starting point for the portfolio shift that followed the exit. How a process-plant business should hold itself at a distance from commodity markets remains an open question for a company that still carries chemicals and cement.
The first move was to team up with the buyer. OCI of South Korea, which took the plant off Tokuyama’s hands in 2017, is an equal partner six years later. It looks like the same ground, the same business and the same counterparty with only the positions exchanged, but the substance is not the same. Where the earlier venture was aimed at commodity solar grade, this one is confined to a semi-finished product for semiconductors, with finishing and sales left in each parent’s hands. Using the partner’s land and equipment while refusing to give up the final process and the customer is the line drawn from the previous loss.
A cautious design is not the same as a small bet, however. The joint venture is capitalised at the equivalent of about US$218 million, its establishment ran fifteen months behind the original schedule, and it has got no further than a groundbreaking ceremony. The Vietnamese subsidiary handles only the final process, so dividing the process across borders means that if any one part stalls, the whole thing stops. Whether the stance taken to avoid the last failure has invited a different kind of fragility cannot be judged until there is an operating record.
On profit and loss alone, cement had long since sunk. The reason given in 2022 for not closing it was the internal loop at the Tokuyama works, which turns other products’ waste into raw material — a benefit said to outweigh the losses. The decision to fold it comes from getting ahead of the single point at which that benefit disappears. Once the site is down to one kiln it can no longer perform the waste-processing role, and by then no buyer will be left: the sense of urgency Inoue Tomohiro described puts the presence of a counterparty, not the life of the business, on the time axis.
Even so, this tidying leaves things unresolved. The kilns are to be left retired in place for the time being, concrete plans for conversion are still to come, and the use of the site is undecided. Waste accepted from outside is to be discussed item by item with the companies that generate it, so not every destination has been settled. Against the clarity of the figure — some 2.8 million tonnes of CO2 a year gone — the outline of what remains in the equipment and the region is still faint. What is to be put in place of the second founding business at a lightened Tokuyama works has not, as of this writing, been shown.
Each heading links to the full Japanese analysis — background, decision and outcome, with sources.
This is a condensed English edition. The full, source-by-source history — with the detailed narrative, financial tables, shareholders and executives — is maintained in Japanese: 日本語版（詳細）— Tokuyama full history in Japanese →
Tokuyama Corporation — 有価証券報告書 (annual securities reports) and company disclosure materials.
This page is provided for general information only and is not investment advice, nor a recommendation to buy or sell any security.
Figures are compiled independently and include our own estimates, approximations and machine-processed data; we make no warranty as to their accuracy or completeness.
Sources are primarily each company’s securities reports and other public filings, but errors and omissions may remain.
Any use of this information is at the reader’s own risk. Past performance does not indicate future results.
Company names, logos and other marks belong to their respective owners.
Data API
Tokuyama’s history, financials, executives and
shareholders are published as static JSON — no key, plain GET.