Remember the days when "Made in Taiwan" meant anything sold in a five-and-dime store? In the 1970s it seemed that the tiny island off the coast of southern China churned out just about every cheap umbrella, polyester shirt, and generic plastic toy sold in the U.S. No longer. Now "Made in Taiwan" means research and development, innovation, and tough competition for American manufacturers. It is also the place U.S. corporations turn to produce high-tech goods. Taiwanese companies make a quarter of Compaq Computer Corp.'s desktop computers, 40% of IBM Corp.'s, and 60% of Dell Computer Corp.'s. The nation is the world's largest producer of monitors, desktop scanners, modems, and seven other categories of information-technology equipment. It is the top exporter of bicycles -- most of its models are sold under such brand names as Schwinn, Trek, and Specialized. As the world's fourth largest producer of semiconductors, it has earned the nickname Silicon Island. At the same time the island's entrepreneurs were furnishing the developed world with plastic and polyester, its universities were graduating mathematicians and scientists who went on to earn advanced degrees in the U.S. They found jobs at leading U.S. corporations. As its brainiest students became Silicon Valley technologists in the 1970s and early '80s, Taipei officials established several research institutes in an effort to help local enterprises boost their research capabilities, reports Chintay Shih, president of Industrial Technology Research Institute. These include the Hsinchu Science-Based Industrial Park and Shih's own institute, a government-funded organization that collaborates with business to conduct research in such areas as electronics and computers. Officials awarded generous incentives to returning expatriates willing to start businesses with the skills and experience picked up in the U.S. As a result, Taiwan in the late 1990s shares more in common with Silicon Valley than with Tokyo. The society values entrepreneurs. "Most people would prefer to own their own business rather than work for others," observes Shih. "They're always looking for new opportunities or to try a way that has not been tried before." Taiwanese start more than 100,000 companies annually, and a majority of them fail. Taiwan boasts the region's healthiest stock market -- TAIEX -- and at least a quarter of the population of 21 million invests in it. "Some people criticize the market as a money game, but stocks have performed well recently," points out Peter Chow, a professor of economics at City University of New York, who often lectures on the island's entrepreneurs. Chow sees a shift in attitude from first-generation industrialists who made money in shipping, cement, and textiles to younger U.S.-educated businesspeople who prefer technology ventures and rely on U.S. management strategies. "The second generation is much more influenced by American culture than their parents, who were affected by Japan, which occupied the country for 50 years," he says. Most companies there have sales of less than $300 million. Those that make it big are chameleons that change when markets shift and opportunities arise. To fund ambitions, they list on TAIEX, and a few have registered abroad. Their flexibility and fast decision-making have been credited with helping to cushion the island from the recession striking other Asian tigers. When local wages went up, entrepreneurs opened factories in China. At first they helped transform cities in Guangdong and Fujian provinces with new buildings and roads, but recently they have moved northward to Shanghai. Even now, as winds from Asia's financial storm threaten Taiwan's exports, entrepreneurs promise to battle harder. When manufacturers see competitors creeping into their markets, they reinvent their companies. They can do so because they pour as much as 16% of sales back into research and development. Not only are leading entrepreneurs shifting product lines based on solid research, they're implementing market changes and making decisions to grow to the next level. For some this means brand leadership, for others, increased sales. A number of U.S. manufacturers now view Taiwan as a nation of fierce competitors. Others see nimble, cost-effective producers. In some cases, entrepreneurs fill both roles. How a small Giant grew Giant Mfg. Co. Founded: 1972 Product: Bicycles and accessories 1997 sales: $300 million % of sales to R&D: 2%-3% % market share (in developed countries): 1991, 7%; 1994, 10%; 1997, 14%. One example is Giant Mfg. Co. Ltd., which in 26 years has grown into one of the world's largest bicycle makers. In Taichung, northeast of the island's midpoint, racers on Giant two-wheeled machines whiz down the city's hilly, curving roads, past rice paddies and light industrial factories. "After riding for a couple of months in Taiwan, you develop sort of a sixth sense, learn to move with traffic," reports Antony Lo, Giant's 50-year-old CEO who looks fit enough to ride to work, but doesn't. "I'm a bit lazy," he admits. Taichung is probably the island's only city where one is likely to spot racers on Giant bicycles, and they're probably members of the company's professional team training for the Tour de Taiwan race. Almost everyone else hurries around in cars or on scooters, the preferred mode of transportation for young drivers. Giant exports 93% of its bicycles. In 1972 the bicycle market looked bleak thanks to a world energy crisis that forced down production and sales. Despite weak demand, Giant was founded that year as a contract bicycle manufacturer. Such brand names as Schwinn Cycling & Fitness Inc. turned to Giant early on, thanks to low labor costs. Giant learned a good deal about manufacturing from its clients, but management knew it could grow faster selling under its own banner. Nine years after it opened, Giant launched a line of branded bicycles. "If you don't have a brand, you don't control your destiny because you're not involved with the final consumer," explains Lo, who was hired shortly after the company was founded. Plucky Giant also recognized that if it wanted to build a global brand, it would need to establish operations abroad, not just distribute overseas. In 1986 Giant chose Europe -- where managers saw an annual market demand of about 15 million bikes and a niche -- to set up its first international sales-and-marketing presence. It selected the Netherlands for its central location. "In Europe we found two extremes," explains Lo, "cheap commuting bicycles or very expensive racing bikes." Giant introduced a model in between the two -- well-made, but affordable. Later, when Giant announced plans to open an assembly plant in the Netherlands, Lo fielded questions over why the company would build anything in a country where wages were 50% higher than in Taichung. He offered a few solid reasons: Shipping from Taiwan to Europe can take two months. Building the frame and a few basic components in Asia and then dispatching them to Europe for assembly enables Giant to "react quickly to customer demands such as color and size changes." Its Netherlands presence also helps Giant avoid strict European Union antidumping duties. To limit costs on its low-end bicycles, Giant opened two manufacturing centers in China. The company also established a U.S. sales-and-marketing unit and is considering opening an assembly plant in the U.S. for the same reasons it did in the Netherlands. Giant has devoted 2% to 3% of annual sales to R&D, which resulted in several firsts, including the first carbon-fiber composite frame and the FlexX bicycle for children designed to expand as they grow. Recently, it launched a battery-powered bike targeted at leisure riders. "We try to change ourselves every year," explains Lo. "We could rely on one product for three or four years and save a lot of money, but change is important to the company." Giant's R&D investments helped the company hold onto its contract-manufacturing clients. At first, OEM customers objected to Giant's decision to launch its own line of bicycles, and the Taichung company lost a few clients. Others remained loyal, hoping to benefit from original design manufacturing (ODM) of groundbreaking products. Specialized Bicycle Components Inc., which began working with Giant in the early 1980s, decided to stick with the Taichung manufacturer even though the California bike company knew it would be competing with the Giant brand in some markets. "Giant's factories are extremely proficient in manufacturing and efficiency," reports Bob Margevicius, Specialized's executive vice president. When Specialized wanted a manufacturer for its high-tech FSR model, a full-suspension bicycle to carry riders smoothly over bumps, it turned to Giant. Margevicius explains his company manages its competitive relationship with its Pacific rival (and partner) by filling a niche -- high-performance bicycles selling for over $500 -- that Giant hasn't targeted. Giant retains a healthy mix of OEM, ODM, and brand-name products. In 1998 contract manufacturing represented 35% of Giant's sales, which were expected to reach $350 million, a 19% jump from 1997. U.S. corporate boot camp Microtek International Inc. Founded: 1980 Product: circuit-board accessories, scanners 1997 sales: $218.4 million % of sales to R&D: 7%-10% % of market share: 16.2% Just as Lo wants Giant to mean high-tech bicycles, Benny Hsu wants Microtek International Inc. to mean high-quality scanners. A technology fanatic, the Microtek founder beams as he walks around his company's showroom in Hsinchu Science-Based Industrial Park, listing the research developments that led to his latest handheld and portable scanners. "If you ask me about priorities, my first is technology, second is technology, and third, technology. All our extra money is spent on R&D [varying from 7% to 10% of annual revenues]. Maybe we're a little weak on marketing, but that's because we're totally technology-oriented," he enthuses. When Hsu founded Microtek in 1980, Taiwan was better known for cheap umbrellas than as Asia's Silicon Valley. Hsu led the way as one of the first U.S.-university-educated and company-trained technologists to return home with a few friends to start a company. He pioneered the settlement of the now-burgeoning Hsinchu Science park opened by the government in 1980 to foster high-tech businesses. Microtek first made a splash with a product it calls MICE, a debugging device for circuit-board design. By 1984 it launched the world's first desktop scanner. "Taiwan had no scanner industry. We had to teach vendors how to make parts for us," remembers Hsu. Not only did Microtek have to show suppliers how to build the product, it instructed potential customers on how to use its models. Microtek employees spent five years traveling to trade shows promoting scanners. The company didn't make money on them until 1989. Similar to the Giant strategy, Microtek strikes a balance between selling products under its own brand name and building those for other companies. Roughly 20% of its scanners are sold as ODM products for such corporations as Dell Computer. To compensate for its almost singular focus on technology, Microtek sets up alliances with companies that offer greater marketing power. These partnerships take different forms, but usually involve technology exchange. Microtek, for example, makes film scanners for Polaroid Corp. and gains attention when the Cambridge, Mass.-based camera and instrument maker mentions Microtek in press releases and at product launches. Microtek is banking on China where it holds 46% of the scanner market. In Shanghai it built a $25 million complex to churn out entry-level scanners for the local market, but there it faces a challenge similar to one it confronted at home less than a decade ago. Although Microtek was the first company in Taiwan to enter the scanner industry, local startups soon followed. One, Mustek Systems, Inc., quickly overtook Microtek and now holds 27.6% of the world's scanner market compared with Microtek's 16.2%. Hsu believes that all of his country's successful scanner manufacturers -- others include UMAX and Plustek -- were spawned by former Microtek employees, and he blames the loss of trade secrets on his slow speed in applying for patents. In China, where stolen intellectual property trades briskly, the same situation could occur. Hsu says he's not worried because it would take too long for an entrepreneur new to the market to catch up to existing producers. What's more, Hsu has taken care to arrange his company in silos so that employees learn just a portion of operations. "Manufacturing workers don't understand what the engineers do. The engineers don't know how purchasing operates, and purchasing doesn't understand systems software," he explains. FIC's firm hand in China The FIC Group Founded: 1980 Products: motherboards, PCs 1997 revenues: $1.2 billion % of sales to R&D: 2%-3% % of market share for motherboards and desktop computers: 1991, 3%; 1994, 6%; 1997, 7.5% Charlene Chien set up silos in her head for each division of the FIC Group, the motherboard and PC manufacturer she founded with her husband. She easily switches between them describing in English at a staccato pace the challenges each faces. The spouse team promotes a horizontal management structure, and many FIC divisions are run by talented Taiwanese technocrats with U.S. advanced degrees, work experience, and a good deal of responsibility. "Every one of our 11 divisions has its own president and is run like a small company," Chien explains. She oversees each division's P&L. Her husband concentrates on strategy, research (FIC devotes 2% to 3% of sales to R&D), and China operations. That structure has taken FIC past the $1.2 billion mark. Now Chien hopes for what she calls sane growth: 20% to 30% annually. "I have a little daughter, and I want to spend time with her," she says, admitting workaholism runs in her family. Charlene Chien is the daughter of Taiwan's leading tycoon, 82-year-old Y.C. Wang, who heads $10.4 billion Formosa Plastics Group. She talked business from the time she learned to speak. "You've got to respect my father because he worked 365 days a year, but I'm happy working six days a week. I need outside interests," she explains from her Taipei office at the back of the hulking Formosa Plastics complex. Being the scion of a famous businessman proved both a help and a hindrance to Chien and her husband when they started FIC in 1980. The Wang name proved key when wooing customers and negotiating loans from the country's then tight-fisted banks. At the same time, Chien felt as if her family name put her in the spotlight. "Because I'm the daughter of my father, I'm supposed to do well." Under that spotlight, Chien's company thrived. It has contracts to build PCs for some of the world's best-known brands. Bound by agreements demanded by customers who don't want consumers to know their products are made in Taiwan, FIC is tight-lipped about its client list, but says it designs and builds PCs for two of the top four brands. In 1994 FIC turned to China to manufacture motherboards and PC products because of skyrocketing labor costs at home. That year it set up a large plant in Shenzhen, which borders Hong Kong, and a smaller one in Shanghai. The company quickly learned that just because they spoke the same language didn't mean their Chinese employees followed FIC's work ethic. "We learned how to handle them -- treat them like the military would," reports Chien. "Set strict rules, but pay workers well." With management under control, and the realization that China offered engineers with solid skills, FIC not only transferred standard manufacturing to the mainland, but also set up a motherboard repair center there and plans to expand further. As a model of what she wants her company to become, Chien looks to another successful company founded by a Taiwanese: Milpitas, Calif., contract manufacturer Solectron Corp. Customers have pushed FIC in that direction. Increasingly they ask the company to handle "manugistics." The term covers design and production, as well as logistic and distribution services. For Chien, manugistics means snap decisions. "The technology changes so fast," she says. "It used to be that one computer would last 18 months -- now it lasts four. We're used to constant change." A glass ceiling and a good idea Macronix International Co. Ltd. Founded: 1989 Product: nonvolatile memory and "system-on-a-chip" ICs 1997 sales: $327 million % of sales to R&D: 14%-16% % of market share in semiconductors: 1993, 2.8%; 1995, 3.8%; 1997, 5.4% Miin Wu also thrives on change. Like his parents' generation who emigrated from China to Taiwan during the Communist Revolution, Wu and many of his friends migrated, temporarily, to the U.S. Wu left home to study engineering at Stanford University and then went on to Intel Corp. By 1988 he tired of what he perceived as a glass ceiling for Asians with strong technical backgrounds, and he wanted to start a company. "When I was at Intel, the top Chinese was at the director level. He spoke fluent English, had MIT and Stanford degrees, but could only reach that level. The limits were obvious," he remembers. Wu turned his complaints into action and talked two dozen Taiwanese working in Silicon Valley into returning home. He persuaded a number of well-paid executives to leave stable positions for a high-risk venture by raising the promise of a company where employees could be a part of the vision and Silicon Valley-type rewards such as a form of stock options. In 1989 the group founded Macronix International Co. Ltd. in Taiwan's Hsinchu Science Park to specialize in nonvolatile memory and "system-on-a-chip" ICs. Since its founding, Macronix has moved seven times, always to larger quarters. This year it will move again. "We just keep expanding," observes Wu. Wu's commitment to R&D has fueled that growth. Even during the semiconductor industry slump the last two years, Wu has plowed 14% to 16% of sales back into R&D. "Bad times like this are a perfect time to invest more in research to remain competitive," he explains. "A lot of other companies are scaling down, so after the slump we can outgrow them as long as we maintain positive cash flow." Wu savors the day when Taiwan will overtake Japan as the region's largest chip producer, but at the same time he links up in manufacturing alliances with companies -- such as Matsushita Inc. -- that may one day prove rivals. Such ties have helped Macronix achieve a good cash flow during tough times. Macronix also fueled growth by listing on Nasdaq, the first Taiwan manufacturer to do so. In 1997 it recorded net income of $59.6 million on sales of $327 million. Despite his company's growth, Wu fears losing the entrepreneurial edge. Recently he hired consultants McKinsey & Co. to "create a new organization" and to restore the enthusiasm fostered when he and a bunch of engineers plotted market strategies over late dinners. Because the company has grown so fast, Wu believes it lacks experienced managers. "We have so much activity going on that it's hard to find time to train younger people, to give them authority so they can learn important skills," he explains. With McKinsey's help, Macronix is reorganizing into three business units, each of which will have P&L responsibility to encourage risk taking. Wu expects this reorganization will move Macronix to the next level: sales of $1 billion. Still, perhaps his biggest trouble is keeping talent in Taiwan. After a few years at Macronix, a number of executives left, citing the lifestyle they enjoyed in California or other American states where they owned large houses and sent their children to their alma maters. Even his own parents refuse to leave the U.S. "We just continue to bring back new people," shrugs the entrepreneur, determined to remake his home into Silicon Valley East.