Sharp — Company History

Eleven years, and the propelling pencil that ended the subcontracting

In September 1912 Hayakawa Tokuji (早川徳次) started a one-man metalworking business at Matsuicho, Honjo Ward, Tokyo, taking subcontract work on sundries such as buckles and fountain-pen fittings. In August 1915 he invented a metal pencil that advanced its own lead, and obtained patents in both Japan and the United States. That product, named the Ever-Ready Sharp Pencil, carried Hayakawa from jobbing metal subcontractor to a stationery maker with a product of his own. In 1919 he added his own refinements to the durability of the metal fittings, introduced flow-line working in the factory and committed to volume production. Against the imported single-action type, which had to be dismantled to use, a propelling pencil that could be worked with one hand won repeat orders from stationery wholesalers and retailers. Exports to Europe and America grew during the First World War, and just before the Great Kanto Earthquake of 1923 the business had reached a factory of 300 tsubo (about 1,000 square metres), some 200 employees and monthly sales of ¥50,000.

The Tokyo factory burned down in the Great Kanto Earthquake of September 1923, and Hayakawa lost his wife and two children. Nihon Bungu Seizo (日本文具製造), his sole sales agent, demanded immediate repayment of ¥20,000, and Hayakawa answered by handing over such machinery as had survived the fire and by letting the firm use, free of charge, the 48 propelling-pencil patents registered in his own name. He attached conditions — that his trade receivables be paid, that his engineers be taken on, and that he himself serve as their chief engineer for six months — but it was a gentlemen’s agreement, with no contract exchanged. The settlement left him no manufacturing rights to the propelling pencil, and with some seventy of his injured and displaced workers still to support he had no visible route to recovery. Leaving nothing in writing later brought a seizure of assets and a lawsuit that ran two years and ten months, with the instalments continuing until April 1934.

Localising the radio: a second founding

In September 1924 Hayakawa set up the Hayakawa Metal Works Institute (早川金属工業研究所) in what is now Abeno Ward, Osaka, and began making radio receivers and their components. At the end of the previous year he had bought an imported crystal radio set for ¥7.50 at the Ishihara watch shop in Shinsaibashi, Osaka, and taken it apart; what he understood from it underlay the change of direction. The business — restarted by selling fountain-pen fittings door to door, because he could not make pencils — turned again in 1925, when he became the first in Japan to succeed in assembling a compact crystal radio set. He stamped the receivers with the mark Sharp, taken from the stationery he had given up, and by July of the same year output passed 10,000 sets a month and the works moved to a factory devoted to radios.

In 1926 flow-line working was applied to radio manufacture as well, and component exports began to China, South-East Asia and South America. In 1929 the company entered production of mains-powered valve radios, and in 1931 it became the first private firm in Japan to take up television research. Within a few years of the start of Japanese radio broadcasting in March 1925 it was producing receivers domestically in volume, and then reaching towards the next display technology, television. In June 1934 it built the Hirano plant at Kami-Fukuidocho, Higashi-Sumiyoshi Ward, Osaka, widening into a proper factory a manufacturing operation that had begun as a 37-tsubo works-and-dwelling on 235 tsubo of leased land at Tanabe.

Incorporation, and a fresh start after the war

In May 1935 Hayakawa reorganised the business as a joint-stock company with capital of ¥300,000, founding Hayakawa Metal Works Institute Ltd. In June 1936 it was renamed Hayakawa Metal Industry Co., dropping the word “institute” from its original style. In May 1942 it changed again, to Hayakawa Electric Industry Co., matching the name to what the business now did. Under the war economy it was compelled to convert to military production — optical weapons, binoculars, magnetic recorders — and could not go on making radios for civilian use. Thirty years had passed since the founding before a company that began in metalworking called itself an electrical manufacturer in its own name.

In May 1949 the company listed its shares on the Osaka Stock Exchange. By 1951 it had completed a prototype of the first domestically built television receiver. In the Osaka of that time Matsushita Electric Industrial and Sanyo Electric were gaining strength in the same period, and it was said that Osaka, long regarded as a textile city, was turning into a city of light electricals. But where Matsushita and Sanyo were both full-line light-electrical manufacturers, Hayakawa Electric took a direction closer to that of a television specialist. Narrowing the range would show its worth in the postwar television boom, while it also meant setting out behind the two of them in breadth of product and size of sales network.

The RCA licence, and television as a specialist’s trade

In June 1952 Hayakawa Electric Industry concluded a technical tie-up with RCA of the United States, and in January 1953 entered volume production of television receivers. In July 1954 it built the Tanabe plant in Abeno Ward, Osaka, installing endless conveyors to raise receiver output. In March 1956 it listed its shares on the Tokyo Stock Exchange as well. Sales bases followed one after another — a Tokyo branch that April, Hiroshima in August, Nagoya in September 1957 and Fukuoka in 1958 — and in December 1957 it pressed ahead simultaneously with an extension to the Tanabe plant and the building of a second Hirano plant. Production and sales were assembled in tandem in about five years.

The first reason Hayakawa Electric posted the best results in the industry was that it had put a television expansion programme in place early, in anticipation of the television boom. Where Matsushita and Sanyo were both full-line light-electrical manufacturers, it took a direction closer to a television specialist’s. The second reason lay in not making its own tubes and valves. Matsushita, Hitachi and Toshiba all produced theirs in house, and to bring costs down they had to make more than they consumed themselves and sell the surplus outside. Because that structure required buyers, it was said that Hayakawa Electric, in not making its own tubes and valves, showed a kind of foresight. Its machining, metal pressing included, and the Hayakawa flow-line method were also recognised early within the industry.

Into full-line appliances, and the weakness of the brand

In July 1959 the company opened the Yao plant at Yao, Osaka Prefecture, as an integrated works for household electrical appliances, and in January 1960 the Nara plant at Yamatokoriyama, Nara Prefecture, making television and radio components. 1960 was also the year it began volume production of colour televisions, took up development of microwave ovens and semiconductor products, and installed IBM machines for head-office administration. In March 1961 a refrigerator factory was newly built inside the Yao plant, and in November a central research laboratory was raised in readiness for electronics development. It launched medical electronic instruments such as electrocardiographs and electrosurgical knives, and succeeded in prototyping a microwave oven. It was in that year, too, that it began developing computers and desktop calculators.

In May 1962 it established Sharp Electronics Corporation in the United States in preparation for expanded exports there. That March it extended and rebuilt the refrigerator and air-conditioning plants and entered volume production of microwave ovens, and set up a computer manufacturing department at the Nara plant. In 1963 it achieved volume production of solar cells and silicon photoelectric conversion elements, and released the industry’s first 12-inch portable television. The range widened, but it still could not win buyers who asked for it by name. In the appliance slump of 1964, when more customers chose their maker by name, it was written that listing “the makers least often asked for by name”, fewest first, gave the order Hayakawa Electric, Yaou Denki (八欧電機) and Sanyo Electric.

Giving up the computer and choosing the electronic desktop calculator

In 1960 computer development was the largest subject among electrical manufacturers, and the major firms were negotiating tie-ups with American companies. Hayakawa Electric, which in radio and television had held its own against the very largest makers, is said to have feared being left behind in this new age. Asada Atsushi (浅田篤), five years into the company, had been arguing over the future with other young employees when word of those discussions reached senior management; a calculator research group was formed, with Asada at its centre. Since almost none of them knew anything about calculators, they went to study from first principles under assistant professor Ozaki Hiroshi (尾崎弘) at Osaka University. The following year the company poured in new graduates in numbers, meaning to commit itself in earnest to developing a large computer.

That prospect was optimistic. Hitachi, Toshiba, NEC and Fujitsu had already been developing computers for more than five years, and the Ministry of International Trade and Industry was occupied with how to support them. To an inexperienced newcomer, the official in charge would not listen: Hayakawa doing a large computer — that would be impossible. The technology was too large to carry alone, so in 1962 the company changed its target and moved towards electronifying the mechanical calculator. It fixed its limits — a size that would sit on a desk, in place of the hand-cranked machine, and a price ceiling of $1,389 (¥500,000) — and designed from there. Both the arithmetic circuits and the memory had to be built up transistor by transistor; packed into a box they ran hot, which required a fan, and the fan in turn made the box bigger.

In 1964 the company developed the world’s first electronic desktop calculator, the Compet, and entered full volume production. It used 4,000 components including transistors, weighed 25 kilogrammes and cost $1,486 (¥535,000), a little above the target. Even so, a machine that calculated silently and quickly compared with the hand-cranked type won a reputation at home and abroad. In 1966 it developed a calculator using integrated circuits, and in the first model of 1967 the component count fell to a sixth, bringing the machine down to four kilogrammes and $639 (¥230,000). Sasaki Tadashi (佐々木正), head of the industrial equipment division, fixed on low-power MOS integrated circuits and negotiated with American Rockwell of the United States, then working mainly on military work, to secure a contract for their development and supply for calculators. With the MOS LSI calculator of 1969 the component count fell to less than a hundredth and the price dropped below $278 (¥100,000).

From Senri to Tenri — the ¥7.5bn the world’s fair did not get

In January 1970 Hayakawa Electric Industry changed its corporate name to Sharp Corporation. That same year the Japan World Exposition opened on the Senri hills outside Osaka, and yet the company, a leading local business, put up no pavilion. Instead, in August 1970, it built the Sharp Integrated Development Centre at Tenri in Nara Prefecture as a base for semiconductor production and research. The investment came to about $20.8M (¥8bn). At the time it committed to this outlay — “from Senri to Tenri”, as it was called — the company had a string of pioneering achievements behind it in radios, televisions, microwave ovens and calculators, and yet its management was described as having nothing that quite stands out, with president Hayakawa’s paternalism and caution given as the reason. The repeated struggle to procure integrated circuits from the United States for calculator development lay behind the move to make them in house.

The LSIs the company used were built on technology brought in from North American Rockwell, and the volumes it had been importing were switched piece by piece to its own production at the Tenri plant. The Micro Compet, an electronic abacus whose circuitry was entirely LSI, went on sale in Japan from December 1969, and from Europe and America, where it had been launched first, enquiries for the 1970 financial year reached 320,000 units. Sasaki Tadashi, who led the technology, said later: The weight of academic opinion was behind bipolar, and MOS had poor product yields. So no company was keen to produce MOS ICs. But I was convinced of the future of the MOS IC.

Scepticism dogged the tilt towards semiconductors, and the industry passed round the observation that Sharp will be destroyed by its semiconductor investment. Management took it thus: We had not expected semiconductor demand to grow this large, but we believed the direction of electronics itself was not mistaken, so we felt no anxiety. Three to five years of hardship cannot be helped. In September 1970 the founder, Hayakawa Tokuji, stepped down as president and was succeeded by Saeki Akira (佐伯旭). In January 1973 the company laid down its management philosophy, management creed and basic management policy, setting out in words what the business stood on. It was shifting from its established home-appliance field into electronics.

Eighteen makers in a war of attrition, and the turn to liquid crystal

Entrants crowded into the calculator market. By May 1969 they ranged from electrical manufacturers such as Hayakawa Electric and Toshiba to camera and office-equipment makers such as Canon and Casio Computer, and numbered eighteen. With a new model appearing at the rate of one every three months, competition turned on whether the price per digit of calculating capacity would fall below $28 (¥10,000). Intensifying discounting eroded margins, and some makers were driven to halt production. There were cases of makers resorting to unsustainably cheap exports to survive the harsh competition, and of makers announcing new products and then holding back the launch for want of a sales outlook.

The pace of miniaturisation and price cutting was striking: in August 1972 Casio Computer put a six-digit mini calculator that sat in the palm on sale for $42 (¥12,800). In less than ten years the price had come down to about a fortieth. In the middle of this war of attrition Sharp stepped down from the contest over price and drew its technology towards core devices of its own. Sasaki Tadashi and his colleagues established the COS process, which brought low-power integrated circuits and liquid crystal together on a single glass substrate, and in 1973 they put out the EL-805, the world’s first liquid-crystal calculator, which ran for a hundred hours on a single AA cell. Choosing low-power liquid crystal for the display as well was what made that long running time on one battery possible.

The price war itself was won by Casio Computer, whose weapon was cheapness. Its unit sales in the 1974 financial year reached about 4.8 million, and its share of personal calculators passed fifty per cent, far ahead of Sharp’s little over twenty. Calculator output was nonetheless enormous, and it served to bring Japan’s laggard IC and LSI industry up to speed rapidly. The single greatest reason Japan grew strong in the MOS LSIs that led on to VLSI memory was the calculator. Sasaki, for his part, pointed out that it was Intel of the United States that broke through the technical barrier beyond one-chip LSI and took the basic patents, and said: With ICs and LSIs we succeeded through decisions at the top. With microprocessors we failed, I think.

The “golden Showa fifties”, and the rebuild forced by the strong yen

Beginning with a large refrigerator plant built at Yao, Osaka Prefecture in January 1979, the company opened white-goods factories one after another. In March 1981 it placed the Shinjo plant at Shinjo, Nara Prefecture; in October 1984 the Fukuyama plant at Fukuyama, Hiroshima Prefecture, as a production base for electronic components; and in September 1985 an IC technology centre at Tenri. Although this was a period of low growth following the second oil crisis and the worldwide recession, the electronics strategy built up since around 1965 came into flower, and inside the company the years were called “the golden Showa fifties”. As early as 1984 there was also caution: They say a company’s life is thirty years. In another ten it will be exactly thirty years since the all-transistor calculator was developed.

The abrupt appreciation of the yen from 1985 broke the premises of volume manufacture. President Tsuji Haruo (辻晴雄) said: With the sharp rise of the yen since 1985, Japan’s wage level has become the highest in the world. Labour-intensive manufacturing is no longer viable in Japan, citing Korea overtaking Japan in colour-television output and, in radio-cassette recorders, annual production of more than 30 million units in Korea and China against Japan’s 8 million. On that basis he set out the move from simple assembly to advanced technology: The things you truly believe will be core, you absolutely have to do yourself. In our case that meant TFT — thin-film transistor — liquid crystal, semiconductor lasers, light-emitting diodes and the like.

The policy took the shape of plant. In February 1990 the company built the Nara No. 8 plant at Yamatokoriyama, Nara Prefecture, as a production base for copiers; in February 1991 the Tenri plant, as a production base for liquid-crystal panels; and in July 1995 the Mie plant at Taki, Mie Prefecture, likewise for liquid-crystal panels. The change of course was noticed from outside: in 1991 it was written that Sharp, once a byword for a discount brand with an image no better than one-and-a-half-rate, is transforming itself, and that in ordinary profit margin it now surpasses even Sony and Matsushita Electric, coming close to the first rank. Behind it, it was said, lay a distinctive management system into which the company had poured ingenuity in order to fight with scarce resources.

The LCD television declaration and the rise of “Kameyama, made for the world”

Machida Katsuhiko (町田勝彦), who became president in 1998, declared that the company would replace every cathode-ray-tube television with a liquid-crystal television. At the beginning of 2000 it ran advertising on the line What we leave behind in the twentieth century. What we carry into the twenty-first., creating an impression of an innovative company. It took first place in Japan with the AQUOS liquid-crystal television, and a company that had been placed in the middle rank of appliance makers rose from 2000 to be a top maker. Many of those involved look back on the 2000s as a dream for Sharp. What drove it to the top in the 2000s was liquid crystal.

Management made bold investments in rapid succession. By the middle of the 2000s it had put more than $7.3B (¥800bn) into television LCD panel lines alone, at the Mie No. 2 and No. 3 plants in Taki, Mie Prefecture and the Kameyama No. 1 and No. 2 plants at Kameyama in the same prefecture. The symbol was the Kameyama No. 1 plant, which started up in January 2004; “Kameyama, made for the world”, in which the place name itself became the brand, was treated as a model of Japanese manufacturing. The Kameyama No. 2 plant followed in May 2006. Consolidated sales for the year to March 2005 reached $23.1B (¥2.54tn). The strategy of making everything from panel to finished television on a single site paid off, and the year to March 2008 brought a record $33.1B (¥3.42tn) in sales and $986.2M (¥102bn) in net profit.

Behind the strong performance the finances were thin. Free cash flow was negative even in the year to March 2007, when operating profit set a fresh record, and through the first half of the 2000s, as net profit expanded, the equity ratio fell steadily. Machida himself urged caution inside the company: Liquid-crystal television may just turn out to be a business that does not make much money, and In the worst case a time may come when the margins on liquid-crystal television are poor — be prepared for that, I tell them. He also touched on selling outside: That is why I say we will sell panels to anybody. As to whether digital consumer electronics were a promising product at all, he said: I have my doubts, a little.

¥430bn at Sakai, and outside sales that never came

Katayama Mikio (片山幹雄), who became president in April 2007, announced a complex at Sakai, Osaka Prefecture, in July of the same year. The plan put about $3.2B (¥380bn) into the liquid-crystal panel plant, and ultimately some $3.7B (¥430bn), at four times the scale of Kameyama and using the world’s first tenth-generation glass substrates. The plant started up in October 2009, and in March 2010 a solar-cell factory was built at Sakai as well. Sakai had the capacity to produce panels equivalent to 13 million 42-inch sets a year. Katayama had stated plainly in 2008 that our existing panel plants alone — Kameyama No. 2 and the rest — have capacity for more than 20 million sets a year on a 32-inch basis. Given how many liquid-crystal televisions (AQUOS) we sell, that is enough for now, which made Sakai a decision to live not on television sales but on panel sales to other firms.

To secure a stable taker, in February 2008 Sharp joined hands with Sony, then second in the world in liquid-crystal televisions, agreeing that Sharp would hold 66 per cent of the Sakai plant and Sony 34 per cent, with an obligation to take panels in proportion to those stakes. Katayama said: Without a strong partner, the huge new plant at Sakai carries too much risk. In the event Sony invested $106.9M (¥10bn) at the end of 2009 to acquire only 7.04 per cent of the Sakai plant, and never raised its stake thereafter. In 2012 Sony sold its shares in the Sakai plant to Sharp, and the tie-up was dissolved.

The way the partnership was run also thinned outside sales. Immediately after Sakai started up, in the autumn of 2009, liquid-crystal televisions sold heavily in Japan under the newly introduced eco-point scheme, and Sharp repeatedly ran late on deliveries because it gave priority to panel production for AQUOS. Sony’s purchases did not grow, and the trade between them all but ceased. With other important customers such as Toshiba, too, it ran late on deliveries in periods when supply was tight, and when supply eased most of those customers went elsewhere. In the year to March 2012 outside sales accounted for only about a tenth of the Sakai plant’s output. The fall in panel prices also exceeded anything imagined: a 32-inch television panel went from about $865 in 2004 to about $149 in 2011.

Two enormous losses, and rebuilding under Hon Hai

The contraction of the television market after the Lehman shock, with the strong yen laid on top of it, left the liquid-crystal business unable to carry the weight of depreciation on its sales. The consolidated bottom line for the year to March 2012 was a loss of $4.7B (¥376bn), and the year to March 2013 added a further loss of $5.6B (¥545bn). In 2012, the centenary of the founding, the shop floor was in difficulty amid intensifying competition among appliance makers, and was trying to get through on fine-grained economies such as switching off the lights inside the buildings. Sasaki Tadashi, who had been involved in technical guidance since the early years, judged this mistaken; the aim should be to create a new industry. Emergency support from the main banks and a capital increase carried it over the immediate crisis, but its reliance on commercial paper raised through the direct markets became a burden.

Negotiations with Hon Hai Precision Industry had already stumbled once: when chairman Terry Gou (郭台銘) met president Okuda Takashi (奥田隆司) in Hong Kong in 2013, the talks broke down over terms under which investment at the market price was refused and Hon Hai could not take part in running the parent company. Sharp then took investment from Samsung Electronics and Qualcomm of the United States. The losses did not stop: $1.8B (¥222bn) in the year to March 2015 and $2.4B (¥256bn) in the year to March 2016. In January 2016 it reached an outline agreement with the Innovation Network Corporation of Japan, and yet at the board meeting of 25 February Sharp chose the Hon Hai proposal. That August, a third-party allotment of $3.6B (¥389bn) to Hon Hai and three other companies moved about 66 per cent of the voting rights, and the shares were transferred to the Second Section of the Tokyo Stock Exchange.

Tai Jeng-wu (戴正呉), sent in from Hon Hai, cut the threshold above which the president’s approval was required from ¥100 million to ¥3 million, reduced the seals on a circulating approval document from more than ten to three, and spread bonuses over a range of one to eight months’ pay. The year to March 2018 recorded a net profit of $635.9M (¥70bn), the first bottom-line profit in four years, and in December 2017 the company returned to the First Section of the Tokyo Stock Exchange. The loss-making constitution of the display business did not dissolve, however, and net losses of $1.9B (¥261bn) in the year to March 2023 and $989.4M (¥150bn) in the year to March 2024 followed. Sharp had made Sakai Display Products, the manufacturer of liquid-crystal panels, a subsidiary in June 2022; in May 2024 it announced that production there would stop, and volume manufacture ended in August. The Sakai site was sold to SoftBank for about $660.1M (¥100bn) and converted into an AI data centre. The year to March 2025 returned to profit with $241.2M (¥36bn), the first in three years, and the year to March 2026 secured a further $299.7M (¥47bn).