Company history

In November 1963 Tokuo Kubo and Toshio Kotaka, both from the trading house Nissho, set up Tokyo Electron Laboratories in Minato, Tokyo, on $13,889 (¥5m) of capital put up by the Tokyo Broadcasting System (TBS). It began as a minnow of a trading firm importing and exporting VTRs and car radios. The turn came in 1964, when Kotaka crossed to the United States and won agency contracts for Thermco’s diffusion furnaces and Fairchild’s IC testers. The two then made a bet no six-person firm should have made — buying a single $111,111 (¥40m) IC tester with their own money to take around on demonstration — and, on the strength of a TBS debt guarantee, delivered equipment first to NEC and then across Japan’s chipmakers, making themselves the impresarios of an IC industry most Japanese still doubted would ever be industrialised.

Watching Fairchild mass-produce industrial ICs, Kotaka wrote back to Kubo that “the age of the IC is beginning at last.” Rather than clip coupons as a pure middleman, TEL sent its own engineers to train at the American makers and carried the after-sales service itself — a “technical trading house” the big general traders could not match. It took roughly 70% of Japan’s diffusion-furnace market, and Fairchild’s co-founder Robert Noyce reportedly said the atmosphere at TEL was just like the one at his own firm when it struck out on its own.

Then came the crisis that set the company’s course. The 1974 oil shock cut hard: exports of calculators and car stereos — about 60% of sales — were shrinking structurally as big manufacturers piled in. Kubo, who took the presidency that year, and Kotaka resolved to abandon the breadwinner entirely, finishing the withdrawal by 1978. Kotaka likened it to “setting down, early, a barbell my own strength could never have lifted.” It was drastic surgery — the managing director who fought it hardest resigned, and two others left with him — but it opened the way to a single-minded concentration on IC-production equipment, chosen against Kotaka’s seven criteria: a growing market, few rivals, not a major maker’s flagship line, high value-add and high margins. The discipline that would define TEL’s high-margin character was put into words here.

Concentrated now on IC equipment, TEL listed on the Tokyo Stock Exchange’s second section in 1980 and moved up to the first section in 1984. Kotaka, as president, applied at the same time to have the company reclassified from “trading company” to “electrical machinery” — “I don’t want to be called a trading house,” he said — and was turned down, but the direction was unmistakable. It merged the prober maker Telmec in 1984, and in 1988 bought out its joint venture with Thermco to bring diffusion-furnace manufacturing in-house, importing America’s leading technology while building its own production capacity.

What emerged had no precedent in Japan: a firm that was at once trading house and manufacturer, importing and localising US technology while staying glued to its chipmaker customers. TEL stationed some 500 engineers across 22 field-engineering stations nationwide, pledged to reach a customer within an hour of a fault, and turned that service network into a sensor for the next generation’s equipment needs. On its strength the company held the number-one position in Japan across roughly 20 kinds of semiconductor-production equipment — the customer-embedded service that remains the core of its competitiveness today.

By the early 1990s the leadership of leading-edge chipmaking had passed beyond Japan — Samsung in DRAM, Intel in MPUs, TSMC in foundry. In October 1994 TEL switched its overseas selling wholesale from agents to direct sales, building a 200,000-square-metre strategic base with clean rooms in Austin, Texas, and — unusually for a Japanese company of the day — putting local nationals, not Japanese, at the head of every foreign unit. President Tetsuro Higashi argued that customers would only trust a business run with a local face. It worked emphatically: in six years the overseas share of sales rose from 34% to 70% and the US share from 8% to 34%. When the 1998–99 slump cut revenue to two-thirds, the overseas base held the company up, and by the year to March 2001 operating profit reached a record $839.4M (¥102bn); a director later said that without the 1994 decision “the company might have gone under then.”

The 2008 Lehman shock brought TEL its first loss in its history — an operating loss and a net loss in the year to March 2010, ending 46 years without red ink. It recovered the next year, but the episode drove home how violently the equipment market swings. Seeking a second leg, TEL bought the US cleaning-equipment maker FSI International and, more fatefully, Switzerland’s Oerlikon Solar, entering solar-panel production equipment — into a thin-film market already glutted with capacity. It quit solar after just two years, taking a special loss of $441.2M (¥47bn) that sank the year to March 2014 into net loss. The failure proved, from the other side, how deeply TEL’s strength was rooted in the single field of semiconductor equipment.

In September 2013 TEL agreed a merger of equals with Applied Materials, the world’s largest equipment maker — a combined holding company, “Eteris”, to be listed in Tokyo and on Nasdaq, a deal that would have redrawn the industry. Both sets of shareholders approved overwhelmingly and a name and logo were unveiled, but the US Department of Justice raised antitrust concerns and, with the gap never closing, the parties abandoned the contract in April 2015. The year and a half was not wasted: benchmarking itself against the world’s best, TEL took stock of its strengths and weaknesses and returned, deliberately, to growing on its own technology — winding up the loss-making Swiss solar unit and completing its re-concentration on semiconductor equipment.

After the merger fell through, under president Toshiki Kawai TEL turned to the offensive, announcing in 2018 a mid-term plan targeting an operating margin above 30%. The tailwind was structural: smartphones, then IoT, AI and 5G pushed the world toward a data society, and the equipment market entered an expansion without precedent. Revenue grew 2.5-fold in five years, from $7.1B (¥800bn) in the year to March 2017 to more than $15.3B (¥2tn) in the year to March 2022 — a company started on $13,889 (¥5m) of capital had become one of the highest-margin listed companies in Japan, at an operating margin near 30% and an ROE above 37%. Its own-brand ratio had passed 95%; nothing of the trading house remained.

That profitability rests on a profit-first, merit-first culture carried down from the founding — employees graded on a five-step scale with up to fivefold differences in bonus, and rapid promotion (Higashi made director at 41 and president at 46). When a young employee once asked whether it was right to take so much profit, an executive shot back that “we can take this much profit because we are this useful to our customers.” Ranked among the world’s top four equipment makers alongside Applied Materials, Lam Research and ASML, TEL spread its exposure across investment cycles from mature to leading-edge nodes and kept spending 8–13% of sales on R&D even in downturns — a structure that held its operating margin at 28% even when memory investment cooled in the year to March 2023.

In the year to March 2024 Chinese mature-node investment swelled — China took 44% of sales — while leading-edge spending stayed restrained. Then in the year to March 2025 generative AI became the real driver: demand for AI-server memory and advanced packaging lifted revenue to $16.2B (¥2.43tn). Having secured its position in front-end tools — plasma etch, coat/develop, batch deposition — TEL is now extending into the back end, where the hunger for high-bandwidth memory (HBM) is lifting bonders and debonders (revenue roughly tripled from FY2023 to about $200.5M (¥30bn) in FY2025) and advanced-packaging etch. For the 400-plus-layer 3D NAND ramping into mass production in 2026 it is bringing a cryogenic etch tool that drives ions in perfectly vertically. Its answer to the AI-era capex cycle is two-tiered — hold the front-end lead and add a new back-end revenue source — carrying, still, the concentration on one violently cyclical industry that has always been both its edge and its exposure.