Company history

Toshiba traces to two Meiji-era ventures that would not meet for sixty years. The first began with Tanaka Hisashige — the inventor nicknamed Karakuri Giemon for his elaborate mechanical dolls and the Mannen-dokei, a masterpiece of clockwork — who in July 1875 set up a shop and workshop for telegraph equipment in the brick district of Ginza, Tokyo. Under the Meiji drive to modernize, building out the telegraph network was a national priority, and Tanaka’s precision-machine craft answered that state demand. The workshop took the name Tanaka Seizosho in 1882, and after his adopted heir moved it to Shibaura and broadened it from telegraph gear to generators and motors, it was incorporated in 1904 as Shibaura Seisakusho — the heavy-electric root of the company.

The second root lay in electric lighting. In April 1890 the engineer Fujioka Ichisuke, later called the Edison of Japan, founded Hakunetsu-sha to make incandescent bulbs at home, at a time when almost every bulb in the country was imported. Reorganized in 1896 and renamed Tokyo Electric in 1899, the firm sealed a capital and technology tie-up with America’s General Electric in 1905, and in 1911 launched the long-lived tungsten Mazda Lamp, which spread nationwide as a cheap, durable domestic bulb. Where Shibaura carried heavy current, Tokyo Electric carried light current — bulbs, vacuum tubes and communications gear — a different industrial stratum.

In September 1939 the two merged into Tokyo Shibaura Electric, a single house spanning power generation, heavy machinery, communications equipment, home appliances and bulbs — the prototype of the Japanese full-line electrical maker. War swelled it and defeat cut it down: under a 1950 reconstruction plan it spun off fourteen second-tier companies and sold or closed a swathe of plants before starting anew. Listed on the Tokyo Stock Exchange in 1949, it rode the high-growth decades on two wheels, heavy electric and home appliances, becoming the very type of the Japanese general-electric firm. That breadth was its strength — and, as later decades would show, the source of a chronic weakness, since resources spread across everything thinned the edge of each part.

The engine of the 1980s was semiconductors. In April 1984 Tokyo Shibaura Electric formally took the name it had long gone by, Toshiba Corporation. Japan’s chip industry was then sweeping the world market, and in memory chips — DRAM — Japanese makers overtook the Americans to lead global supply; Toshiba stood among them, reaching the front of the pack at the one-megabit generation. A maker of televisions and refrigerators had become a company competing at the leading edge of electronics.

Two inventions defined the peak. Toshiba’s engineer Fujio Masuoka devised a non-volatile memory that kept its data with the power off, and in 1987 Toshiba was first in the world to present NAND flash memory, beginning mass production in 1991; small, dense and shock-resistant, it would later become the storage inside digital cameras and smartphones. In June 1989 Toshiba shipped the Dynabook, the world’s first A4-size laptop, at $1,435 (¥198,000) — a price and a portability that made it a hit past a million units.

Rounding out the peak, Toshiba merged Japan Atomic in 1989 to firm up its reactor business, having built boiling-water reactors on licensed General Electric technology. The full-line maker was at its zenith across appliances, heavy electric, chips and computers. Yet the very width that looked like strength kept diluting each business’s competitiveness — the tension between the reach of the general-electric model and the focus that leading-edge fields demand, which would soon push Toshiba toward selection and concentration.

From the late 1990s Toshiba faced the weight of the full-line structure head-on. Taizo Nishimuro, president from 1996, imposed an internal-company system to expose the returns of each business; Tadashi Okamura, from 2000, brought in capital-cost measures to sort the businesses that earned from those that did not. In 2001, as the dot-com bust sent chip prices tumbling, Toshiba exited the fiercely priced commodity DRAM business — retreating from a front it had once led — and concentrated instead on the NAND flash it had invented and on system LSI. In 2003 it moved to a committee-based governance structure, separating oversight from execution ahead of its peers; how well that worked would be tested hard within the decade.

In June 2005 Atsutoshi Nishida, who had risen through the PC business, became president and set nuclear power and NAND flash as the twin engines of an ambitious plan to lift sales toward ¥9 trillion. Its centrepiece was the February 2006 acquisition of America’s Westinghouse for about $5.4 billion — a price well above what the market expected, won in a bidding war. Taking in Westinghouse’s pressurized-water reactors alongside its own boiling-water line, Toshiba now held both reactor types and the largest nuclear business in the world. Backed by strong chips, it posted a record profit for the year to March 2007 and record sales the following year, and the growth story looked to be running to plan.

It did not last. The Lehman shock of late 2008 collapsed demand for chips and PCs, tipping Toshiba to a large net loss for the year to March 2009 and forcing a public share offering to shore up its capital — yet it did not slacken its investment in nuclear and memory. Norio Sasaki, a nuclear man, became president in June 2009 and carried the Nishida line forward, even as a widening rift with Nishida over targets laid the ground, it has been argued, for the accounting fraud to come. Then in March 2011 the Fukushima Daiichi accident froze new-reactor plans around the world and knocked the premise out from under Toshiba’s pillar of growth. Sasaki held the nuclear course all the same, and Westinghouse’s reach for new demand — buying the builder Stone & Webster — would become the fuse for enormous losses.

In 2015 an accounting fraud shook Toshiba to its foundations. A third-party committee found that from fiscal 2008 to 2014 the company had inflated its profits by some $1.3B (¥156bn), through deferred loss provisions on infrastructure work, padded PC-parts transactions and delayed expense recognition in televisions. The committee traced it to top managers pressing the front lines with excessive targets — organized involvement, not isolated error. On 21 July 2015 President Hisao Tanaka resigned to take responsibility, and his predecessors Sasaki and Nishida stepped away in turn; half the board went the same day. That a company which had adopted an advanced, committee-based governance form could not prevent the fraud weighed heavily — the real fault lay in a culture that guarded appearances at any cost, and a chain of salaried managers, with no strong founding family, that could not put on the brake.

Before the accounting wound had healed, the nuclear business struck. At Stone & Webster, the builder Westinghouse had bought, construction costs ran far past plan and surfaced as a vast goodwill loss. In March 2017 Westinghouse filed for Chapter 11, and Toshiba, which had guaranteed its debt, took over ¥1 trillion in nuclear-related losses; it booked a record net loss of $8.6B (¥966bn) and fell into negative net worth, staring at the delisting that a second straight year of insolvency would bring. To escape, Toshiba raised ¥600 billion from overseas funds and, in a decision announced in 2017 and closed in 2018, sold its breadwinner — the very NAND flash memory business it had invented — to a Bain-led consortium for about $18.1B (¥2tn). The unit, Toshiba Memory, became Kioxia in 2019. Selling the crown jewel to survive marked the beginning of the dismantling.

A fight over control kept Toshiba unsettled. Activist investors poured in with the rescue financing; Nobuaki Kurumatani, brought from the fund CVC, was pushed to resign in 2021 after CVC floated a buyout that looked to shareholders like self-preservation. Toshiba’s picture of its own future would not settle either — a 2021 plan to split into three companies was pared to two and then rejected by shareholders in 2022. Under Taro Shimada, president from 2022, the company finally chose to leave the market altogether: in 2023 it accepted a roughly $14.2B (¥2tn) buyout by a domestic consortium led by Japan Industrial Partners, and on 20 December 2023 its shares stopped trading — ending 74 years on the exchange. Under a single owner, Shimada set the course on rebuilding a full-line maker that had lost its way, trading the discipline of the market for the quiet of private hands.