Exar Corp.

views updated

Exar Corp.

P.O. Box 49007
San Jose, California 95161-9007
U.S.A.
(408) 434-6400
Fax: (408) 435-1712

Public Company
Founded: 1971
Employees: 500
Sales: $159.47 million
Stock Exchanges: NASDAQ
SICs: 3674 Semiconductors and Related Devices

Exar Corp. designs, develops, and markets analog, digital, and mixed-signal integrated circuits for use in telecommunications, data communications, microperipherals, consumer electronics products, and other goods. The circuits are sold primarily to other manufacturers who install them in various electronic systems and equipment. Exar grew rapidly during the late 1980s and early 1990s by introducing cutting-edge products and acquiring other companies.

Exar started out as a small subsidiary of the Japanese semiconductor manufacturer Rohm Company Ltd. of Kyoto, Japan. Rohm was founded in Japan in 1954. Rohm became involved in the electronics components industry and eventually began designing and manufacturing integrated circuits, which were originally developed in the United States. Rohm established the subsidiary in the United States in 1971 as a way of strengthening its presence in North American markets and getting access to U.S. technology. Throughout the 1970s and early 1980s, Exar was led by Rohm expatriates from Japan that came to the United States. During that time, Exar existed as a relatively small designer and manufacturer of analog and integrated circuits—a sort of U.S. sister company for Rohm.

Following Exar’s inception and through the early 1980s semiconductor components replaced vacuum tube gear as the fundamental building blocks used in electronic equipment. Semiconductor components were classified as either discrete devices, such as an individual transistor or diode, or integrated circuits, which incorporated numerous transistors and other elements to create a much more complicated circuit. On an integrated circuit (IC), the elements were fabricated on a small chip of silicon, which was then encapsulated in ceramic, metal, or plastic. Pins sticking out of the IC allowed it to be connected to a circuit board, which held the circuits that controlled the entire electronic device.

The semiconductor industry that developed during the 1970s and early 1980s could basically be divided into two categories: digital and linear, or analog. Digital circuits performed the functions of storing, switching, or translating data that was expressed in binary form (as 0 or 1). Linear circuits, in contrast, monitored and influenced continuously varying signals. Thus, they were used to monitor, condition, amplify, and transform varying properties like temperature, pressure, weight, light, sound, or speed. Both Rohm and Exar researched, developed, and manufactured digital and linear circuits for various applications. An important part of Rohm’s business, for example, became the design of custom circuits for automated manufacturing equipment that was custom-engineered for the fabrication of specific products.

Rohm and Exar both benefited from their alliance during the 1970s and early 1980s. The companies regularly exchanged technical information and key personnel, and partnered on certain product designs, process technologies, and manufacturing procedures. Besides sharing technical information and personnel, Rohm benefited from Exar’s access to the U.S. market. For example, Exar acted as a proxy for Rohm by purchasing certain equipment for Rohm in the United States and by negotiating certain supply and purchase agreements with U.S. companies. Exar also allowed Rohm to tap into some of the best and brightest engineers in America’s high-tech talent pool. On their side, Exar benefited from Rohm’s deep pockets, which it used to finance the development of new products. Exar also profited by having many of its circuits and components manufactured in Japan at Rohm’s highly efficient manufacturing facilities.

Semiconductor markets flourished during the 1970s and early 1980s. Importantly, many Japanese manufacturers began to assert themselves in the global semiconductor market, which was traditionally controlled by U.S. producers. By the early 1980s Rohm was generating annual revenues of roughly $300 million annually, and Exar was doing about $25 million in sales. By that time, Exar had successfully carved out a niche producing custom analog integrated circuits. It also manufactured a range of both custom and standard analog and digital integrated circuits for industries ranging from telecommunications and computer peripherals to industrial controls and scientific instrumentation. About 33 percent of Exar’s revenues were still attributable to Rohm, but it was steadily increasing sales to other companies.

Healthy semiconductor markets combined with proprietary technology developed by Rohm and Exar pushed Exar’s sales and profits much higher during the early and mid-1990s. Indeed, sales shot from $22 million in 1982 to $39 million in 1984, and then to a whopping $57 million in 1985. Net income grew nearly ten-fold during the same time frame to $4.2 million. At the same time that Exar was achieving those gains, the United States began pressuring many Japanese manufacturing sectors, in effect, to reduce their profile in the U.S. market. The pressure was especially high in the semiconductor industry, where the Japanese were rapidly consuming U.S. global market share. Partly because of that pressure, Rohm made the decision in 1984 to begin minimizing its ownership interest in Exar.

In 1985 Exar went public, selling shares valued at $11 each. The offering reduced Rohm’s stake in the company to 68 percent. It also brought about $14 million into Exar’s coffers. “We’re the first Japanese company to be a full-blooded American company,” said Nubuo Hatta, president of Exar. “We have U.S. products, people, and, now, financing.” The offering had the added benefit of increasing Exar’s ability to attract talented engineers, who often wanted stock options as part of their pay package. Although Rohm reduced its ownership in Exar, the two companies planned to sustain their long-running relationship with few changes. Top managers were still Japanese, for example, and Rohm was able to select a majority of the company’s board of directors.

For a variety of reasons, Exar began to flounder in 1985. Markets for some of Exar’s key technologies, for example, drooped in the mid-1980s. The most prominent factor contributing to Exar’s decline, though, was its decision to purchase Exel Microelectronics, Inc. Exel was a privately held manufacturer of advanced semiconductor products. Exar viewed the $6.5-million purchase as a way to get access to new technology as well as Exel’s chip fabrication facility. After Exar purchased the company it discovered that the fabrication facility needed heavy capital investments. The plant was not fully utilized for a few years, and Exar suffered heavy losses in 1987 and 1988. Exar’s revenues dipped to $52.5 million in 1986. Sales bobbed back up to $56.4 million in 1987, but the company posted a disappointing $7.9 million net loss. In 1988, moreover, Exar suffered a crushing deficit of $12.8 million on sales of $69.14 million.

To stem the tide of red ink, Rohm stepped in to help. It purchased 81 percent of Exel for $40 million and within about a year it purchased the remaining assets of the company for about $6.5 million. The buyout allowed Exar to recoup all of its original investment as well as its losses. In addition, Exar continued to benefit from technology contributed by Exel. Some of that technology, in fact, proved extremely lucrative for Exar during the late 1980s and early 1990s. Exar stepped up its research and development efforts during that period and found itself well positioned to take advantage of thriving markets for cutting-edge semiconductor technologies like EEPROM (electrical erasable programmable read-only memory) and ASICs (application specific integrated circuits).

Surging markets and leading technology helped Exar enlarge its revenue base from less than $70 million in 1988 to about $90 million annually by 1990, about $6 million of which was netted as income. That increase reflected an increase in wafer services to Rohm, as well as a 21 percent increase in the sale of custom semiconductors and a 17 percent jump in sales of telecommunications products. The turnaround in income reflected across-the-board revenue improvements in both existing and new products. Early in 1989, for example, Exar introduced the first low-power, single-supply read/write amplifier circuit for the disk-drive market. That and other product introductions carried the company to unprecedented profitability by the mid-1990s.

Meanwhile, parent Rohm continued to reduce its stake in Exar. In 1990 Rohm sold 1.65 million shares to Exar, which diminished its ownership share of the company to about 46 percent. Then, in 1992, Rohm removed Hatta, its Japanese chief executive. Exar’s board hired a new head for the company: George Wells. Wells was born in Scotland and received his bachelor’s degree in physics from the University of Glasgow, where he also studied nuclear physics for two years. After college he went to work for Fairchild Camera and Instrument from 1969 to 1980 before laboring in various management positions at ITT and GTE. Wells then served as the head of General Electric’s global semiconductor operations from 1983 to 1985 before acting as vice-chairman for LSI Logic between 1985 and 1992.

Exar continued to thrive under Wells’ leadership. As soon as he arrived, Wells launched an aggressive quality initiative. He also began repositioning the company’s product line to emphasize high-margin, high-growth products, particularly in the areas of communications, document imaging, and consumer and other computer applications. Among the most promising technologies that Exar was developing in the early and mid-1990s was mixed-signal chips, which combined digital and linear functions. Because the market was relatively specialized, the major chip producers had paid little attention to it. The chips were more complicated and more expensive to produce, but they had the effect of reducing the board size and lowering manufacturing costs for equipment manufacturers. Thus, buyers of the chips were able to produce ever smaller components with greater numbers of functions.

Going into 1993, Wells had reorganized the company’s high-tech “logic” products into five categories that reflected the diversity of its markets: 1) advanced consumer products, such as camcorders; 2) data communications products, like computer modems (importantly, Exar claimed to have introduced the first integrated fax/data modem that allowed both send and receive fax capabilities at 9600 bauds per second); 3) mass storage devices, like chips used in personal computer hard disk drives; 4) telecommunications—for example, Exar had created a chip that made “caller identification” possible; and 5) embryonic specialty products, such as chips that operated remote control car locks. By focusing on those high-tech niches, Exar was able to raise its revenues to $113 million in 1991, $140 million in 1992, and then to about $146 million in 1993. The company’s net income, moreover, grew from $7.15 million to $13.7 million during that time.

By 1994 Exar was rife with cash from surging sales and profits. Rather than sitting on the surplus, Wells decided to invest it by acquiring other companies that complemented his strategy of cultivating high-margin, high-profit semiconductor technologies. To that end, Exar purchased Origin Technology in May of 1994. Origin was a pioneer in automatic speech recognition technology that was applicable in a range of consumer and telecommunications markets. Three months later, Exar bought Micro Power Systems Inc., which was a leader in the development of high-performance data-acquisition circuits. The $25 million addition was a major boost to Exar’s flourishing mixed-signal division. Early in 1995 Exar purchased Startech Semiconductor Inc., which designed and marketed application-specific semiconductors for a range of markets. A few months later Exar bought Silicon Microstructures, Inc., a leader in the production of micron-sized mechanical parts etched in silicon for use in high-tech sensoring devices.

Rohm completely eliminated its ownership affiliation with Exar in 1994. The two companies continued to do business with each other, however. After experiencing record sales and income growth, Exar ran into turbulence in 1994, largely because of its rampant acquisition campaign but also because of problems related to its existing hard-disk chip business and an economic slowdown in Japan. Sales stagnated at about $160 million in the fiscal year that ended March 31, 1995, and earnings plunged to a negative $11.08 million. Despite the loss, Exar management remained optimistic. Wells attributed the loss to fallout from the company’s final transition from commodity-like goods to high-margin products. In fact, Exar’s strength in key growth technologies, its proven research and development expertise, and its meager debt load suggested a positive long-term outlook for the semiconductor manufacturer.

Principal Subsidiaries

Exar International, Inc.; Exar IC Design, Ltd. (United Kingdom); Exar Japan Corp. (Japan); Origin Technology, Inc.; MPS Holdings, Inc.; Micro Power Systems, Inc.; Startech Semiconductor Inc.; Silicon Microstructures, Inc.