Nomura Micro Science - Company History

Nomura Micro Science was incorporated in April 1969 in Nihonbashi, Tokyo, for a single purpose: exclusive distribution in Japan and the Far East of the Nuclepore membrane, an ultra-precision filtration medium developed by General Electric of the United States. Its technical parent was the Nuclepore division of Hokko Chemical Industry, and it began life as an import agency holding rights it had not created. It moved upstream quickly. In December 1972 it joined GE in founding Nuclepore Corporation in the United States and took 23% of it; in November 1973 it absorbed the staff and assets of Hokko’s Nuclepore division, uniting manufacture and sale of the membranes and the equipment around them. Hokko Chemical was the largest shareholder from about this point and still holds a little over 10%.

The decisive change of business came in January 1974, when Nomura licensed ultrapure-water technology from Aquamedia of the United States and entered the ultrapure-water systems business. Selling a membrane once ends the transaction; a semiconductor or pharmaceutical plant can run for years afterwards without buying anything more. Building the membrane into a delivered system reverses that, because the filters and ion-exchange resins inside it wear out the more the plant runs. In March 1976 the company developed a reverse-osmosis system for removing pyrogens and delivered it to a Japanese drug maker, opening its pharmaceutical line.

In July 1980, to bring reverse-osmosis equipment into domestic production, Nomura set up Nihon Aquamedia — later renamed Namtec — as a three-way joint venture with Aquamedia and the plant engineer JGC, taking 33.3%. Localising the RO units moved supply one step further from importing toward Japanese manufacture, and bringing JGC in widened the company’s reach into semiconductor and chemical plant work. Namtec was eventually absorbed into Nomura in January 2006.

In February 1983 Nomura exported an ultrapure-water plant to Samsung Semiconductor & Telecommunications and entered the Korean market. The logic of that business is unusual: ultrapure water feeds the wash steps of semiconductor manufacture and bears directly on yield, so once a plant is installed it cannot be shut down, and the customer needs people who can service it fast and locally. Because maintenance and consumable orders keep flowing after installation, the closer Nomura stationed its staff to a fab, the fatter the account became. As Samsung overtook Japanese makers in DRAM through the 1980s, Nomura became its continuing supplier — and from then on the pattern of its overseas expansion was simply to follow Korean customers abroad. In December 1993 it set up Nomura Techno in Korea as a 50/50 joint venture to take Samsung maintenance work; it was renamed Nomura Korea in August 1999 and is now wholly owned.

The same pattern repeated elsewhere. When Samsung Electronics moved into the United States, Nomura opened a wholly owned US subsidiary in January 1996; when LG Semicon went to Britain, it opened a UK subsidiary in September 1997. Both were later wound up — the network expanded and contracted with the customers it served. Taiwan came earlier, with a delivery in July 1987 and a branch office in May 1995 that ran until October 2015. At home, the company consolidated its head office and research base at Okada in Atsugi in August 1991, after moves from Nihonbashi Hongokucho to Nihonbashi Kajicho in 1977 and Otemachi in 1981; Atsugi has been the head office ever since.

China followed the same route with more turns in it. In February 2001 Nomura formed a Shanghai trading company with Taiwan’s Hantech, taking 70% and later transferring the stake; in January 2006 the two set up a construction and manufacturing base in Shanghai, again at 70% and now wholly owned. 2006 was the year several strands ran at once: Namtec and Aglu Japan were absorbed, Nomura Pure was absorbed, a US subsidiary was re-established, and a Singapore company was opened — reorganisation and international expansion in parallel, in preparation for going public.

In October 2007, thirty-eight years after its founding, Nomura Micro Science listed on the JASDAQ exchange — now the JASDAQ Standard segment of the Tokyo Stock Exchange. Consolidated sales for the year to March 2007 were $231M (¥27bn), built on equipment orders with a substantial share from abroad. Eleven months later the financial crisis arrived. Sales fell 14.5% to ¥21.4 billion in the year to March 2009 and then 46.9% to $129.9M (¥11bn) in the year to March 2010, with an ordinary loss of ¥500 million — the first loss since listing. A rebound to ¥22.0 billion in the year to March 2011 proved temporary: revenue fell back to ¥13.3 billion in March 2013 and ¥15.0 billion in March 2014, both with ordinary losses, and ¥12.1 billion in March 2015 with an operating loss of ¥550 million. Three consecutive years of net losses made it the worst stretch in the company’s history.

Two things caused it. Equipment orders collapsed with world semiconductor investment, while the global subsidiary network and the research programme stayed on the books as fixed costs — the business is tied directly to manufacturers’ capital-spending decisions, so the amplitude of the swing lands whole on the operator. And the Asian and American subsidiaries built out in the 2006–07 expansion cost money to keep alive precisely when there were no orders to justify them. Toyosaku Senda, the longest-serving president in the company’s history, saw the environment change after listing without turning that into structural reform.

In 2014 Osamu Yokokawa took over as president and Senda moved to the board. The year to March 2015 was the last of the losses; the following year brought ¥17.8 billion of sales and a ¥320 million operating profit, and the company was out of the red — though not yet demonstrably able to hold it.

The force that turned a recovery into earning power came from Yoshitaka Yamaki, president from February 2017. Born in 1957 and a graduate of the Tokyo University of Science engineering school in 1982, he had joined Sumitomo Bakelite and moved to Nomura in 1985 — an insider with a resins and materials background, which he applied to the thin-film and fluid-control technology inside the ultrapure-water plants. Sales and operating profit climbed through his tenure: ¥16.5 billion and ¥770 million in the year to March 2017, ¥21.6 billion and ¥1.24 billion in 2018, ¥25.1 billion and ¥1.21 billion in 2019, and ¥21.0 billion with ¥1.85 billion of operating profit in 2020.

The capability behind that was built close to the customers. Nomura set up NAD in Korea in November 2011 as an overseas R&D base — later folded into its Korean subsidiary — and opened a laboratory at Hwaseong in Gyeonggi Province in November 2013. Doing the development work beside the fabs is what allows the water quality specification to keep pace as semiconductor geometries shrink. In December 2018 the company obtained Authorised Economic Operator status as a designated exporter, streamlining its customs handling.

Then the cycle broke upward. Sales for the year to March 2021 jumped 44.4% to $276M (¥30bn), with operating profit of ¥3.97 billion and net profit of ¥2.62 billion, the best figures since listing; the following two years brought ¥31.9 billion and then ¥49.6 billion of sales with ¥6.55 billion of operating profit. Demand for leading-edge chips in 5G handsets, AI servers and data centres had put TSMC, Samsung, Intel and Micron into a global capital-spending programme, and Nomura’s market standing rose with it: from JASDAQ to the TSE Second Section in May 2020, to the First Section in June 2021, and to the Prime Market in the April 2022 restructuring. Yamaki handed over in April 2023 after six years that covered both the recovery from the crash and the start of the boom.

In April 2023 Makoto Uchida, born in 1958 and a career man of Mitsubishi Rayon — now Mitsubishi Chemical — became president, the second materials-industry outsider in succession after Yamaki. Putting expertise in the filters, ion-exchange resins and piping materials that make up an ultrapure-water plant at the top of the company is consistent with its origins in a chemical maker’s membrane division. Uchida promptly wrote a medium-term plan, TTT-26 (Together Toward Transformation-26), targeting for fiscal 2026 consolidated sales of ¥101.0 billion, operating profit of ¥14.6 billion, and — the point of the exercise — ROE above 25% and ROIC above 22%. A company that had posted three straight years of losses after the last downturn chose to make capital efficiency, not scale, its headline measure exactly while the numbers were expanding, in step with a rising institutional shareholder base.

The numbers did expand. Sales rose 47.2% to ¥73.0 billion in the year to March 2024 with ¥10.7 billion of operating profit, then to $643.5M (¥96bn) in the year to March 2025 with ¥15.37 billion of operating profit and a 16.0% operating margin — the best in the company’s listed history, and 95.4% of the way to the TTT-26 sales target with two years still to run. What drove it was a change in geography. The US subsidiary established in February 2006 booked ¥52.3 billion of sales, ¥8.4 billion of ordinary profit and ¥7.1 billion of net profit — 54% of consolidated sales and 63% of consolidated ordinary profit — on orders from the leading-edge American fabs of TSMC, Intel and Micron built under the CHIPS Act. A supplier that began by importing an American membrane had become a principal supplier to American chipmaking.

The rest of the network was rearranged to match. In January 2023 Nomura acquired all of a Shanghai engineering company and renamed it; in January 2025 it merged that company into its original Shanghai subsidiary, finally reducing a Chinese presence built through twenty-five years of joint ventures, stake transfers and mergers to a single legal entity. In November 2024 it opened a wholly owned Singapore company as the base for Southeast Asia, re-entering the region after a Vietnamese subsidiary set up in 2018 was wound up in 2021. The open question is the one the company has faced since 1974. Sales of ¥96.3 billion are eight times the ¥11.4 billion of the crash year, and the consumables and maintenance that follow each installed plant do cushion the swing — but the core of revenue is still equipment orders, which halve within a few years when semiconductor investment turns. Whether the ROE and ROIC targets of TTT-26 amount to real resilience, rather than good numbers taken at the top of a cycle, is what Uchida’s tenure will be judged on.