Japan on the rebound

They once seemed invincible.

But around the turn of the century, Japanese electronics manufacturers lost their way. To be sure, they still dominate markets for many consumer electronics products, but they largely missed out on the hottest growth areas, including mobile phones and more recently MP3 music players. As a result, corporate growth stagnated. And over the past five or six years, the parent companies have been in and out of the red, which has yielded lackluster five-year growth rates. The semiconductor operations that were once the centerpiece of

corporate strategies have been scaled back, spun off and restructured. But nothing has reversed Japan's steady loss of worldwide semiconductor market share, from a peak of 51% of total sales in 1988 to 19% in 2005, the lowest percentage in more than 20 years, according to IC Insights.

"The Japanese business model of highly integrated structures and broad structures is no longer the most competitive business model. Specialization has won out over breadth," says Derek Lidow, president and CEO of market research firm iSuppli.

Japan's electronics industry executives say they have learned their lessons. They have all moved—to greater and lesser extents—to make semiconductor operations independent and more accountable, though they still believe in the benefits of ties to a larger group.

They also recognize "the need to focus on the areas we can compete in," says Satoru Ito, chairman and CEO of Renesas Technology, which was formed by a merger of the nonmemory semiconductor businesses of Hitachi and Mitsubishi Electric in April 2003. But compared with the fastest-growing semiconductor companies, the Japanese still have broad product lines.

Lidow does see bright spots in terms of companies concentrating resources on their strengths. And some market trends, such as the increasing use of semiconductors in automobiles, will put a tail wind behind the Japanese companies thanks to their ties to Japanese automakers. Whether this is enough to ignite solid growth or just allow them to preserve their current positions remains to be seen.

The gradual reversal of the fortunes of Japan's vertically integrated companies resulted from the interaction of several factors. One was an overdependence on the commodity dynamic random access memory (DRAM) sector. "It is very clear that Japanese suppliers never had a real good position [in semiconductors] except in the DRAM memory business," says Ito.

"WE NEED TO FOCUS on the areas we can compete in." —Satoru Ito, chairman and CEO, Renesas Technology

In December 1999, Hitachi and NEC merged their DRAM operations into Elpida Memories. Toshiba and Fujitsu just quit the business altogether. NEC spun off its remaining semiconductor operations into an independent company, NEC Electronics, in 2002. And in April 2003, Hitachi and Mitsubishi Electric bundled their remaining chip operations together into Renesas.

Toshiba and Fujitsu turned their chip operations into "in-house companies," which have greater autonomy and accountability while still being part of the corporate parent. All these companies are now concentrating on non-DRAM sectors, particularly system-on-a-chip (SoC) semiconductors.

The restructured operations have continued to lose ground. Bill McClean, president of IC Insights, says, "Reorganizations are never super-positive out of the gate. You can cut dead weight, but you also lose good people."

"THE JAPANESE BUSINESSMODEL of highly intergrated structures and broad structures IS NO LONGER THE MOST COMPETITIVE business model. Specialization has won out over breath." —Derek Lidow, president and CEO, iSuppli.

There were deeper problems as well. McClean notes that in the 1980s, the Japanese continued to invest capital right through industry downturns. The willingness to support semiconductor operations through thick and thin was one of the supposed benefits of the vertically integrated model. But in recent years, capital expenditures got squeezed when times were tough. "As the corporations started losing money, the semiconductor operations came under much more scrutiny as profit and loss centers," McClean says.

The consequences can be seen at Toshiba. The company pioneered the flash memory market. And when it exited the DRAM business, it converted several old factories to make flash memory. Shozo Saito, a Toshiba executive vice president, says Samsung outpaced Toshiba's investments by building 10 flash memory factories in 10 years. Samsung, which licensed the flash technology from Toshiba, is now the market leader, holding a 48.7% share in the first quarter of this year against Toshiba's 24.6%, according to iSuppli. Toshiba is bringing new flash memory capacity online this year. "But it's going to be difficult to catch up to Samsung," says Saito.

Other Japanese makers also failed to capitalize on products they initially developed and had hoped would replace DRAMs as a business. Will Strauss, president of market research firm Forward Concepts, says, "NEC was really the first true digital signal processor market leader, but Texas Instruments took the market away." The reason: NEC lagged in providing the customer support and programming services required by logic chips. Such services were never needed in the commodity DRAM business, he adds.

Perhaps the clearest example of the problems of vertical integration is the failure of Japan's electronics makers to make a dent in the global mobile phone market. "Japanese companies make very good mobile phones, but their global market share is very small, almost zero," says Gerhard Fasol, who tracks telecommunications markets for Tokyo-based Eurotechnology Japan. He adds that this obviously had "big consequences" for the chip operations.

In fact, the Japanese semiconductor manufacturers have largely failed to break into the supply chains of the increasingly important electronics manufacturing service and original design manufacturer (ODM) companies that are making most of the world's mobile phones and PCs as well. Partly it was not having the right products.

Shigeru Fujii, president of Fujitsu's Electronic Devices Business Unit, says that its mobile phone chips were designed for Japan's idiosyncratic mobile phone standard and not the Global System for Mobile (GSM) telephones, which are nearly ubiquitous throughout the rest of the world. And they had stopped making the memory that goes into PCs. "We had no solutions for the industry's main growth applications," he says. But more generally, admits Renesas' Ito , "we have not done a good job promoting our business with ODMs and OEMs."

While largely agreeing on the various reasons the Japanese semiconductor industry lost ground, analysts and Japanese industry executives are split over whether or not vertically integrated companies are a dying breed. Not surprisingly, Toshiba and Fujitsu remain true believers in the benefits of vertical integration, and they can point to the bottom line. "Within the Japanese industry, only Toshiba and Fujitsu had profitable semiconductor operations last year," Toshiba's Saito notes. Toshiba reported operating income of $1.05 billion for its electronic devices segment; Fujitsu reported $284.6 million. That's while recently spun off semiconductor companies were all in the red.

These companies still say there are synergistic benefits of strong ties to a larger electronics manufacturer. They sell about 10% of their chips in-house or to their corporate parents. "From the standpoint of sales, the impact is not so big," says Fujii. "There is a big advantage in having in-house expertise in communications, computing and security available for incorporation into our silicon products."

As an example, he points to the products they are developing for WiMax, a next-generation broadband wireless networking technology. Fujitsu has developed WiMax products and software to cover everything from the chips to the base stations to the phones. Fujii says the semiconductor division feels it is poised to sell the chips to other customers.

Other executives tell similar stories. Toshiba's Saito says the semiconductor operation gets valuable insight into consumer electronics trends because the company also manufactures TVs, DVD players and audio systems.

Toshio Nakajima, president of NEC Electronics Corp., adds, "Big industrial customers are not looking just for semiconductors, but systems and software, so it is advantageous to work with the NEC group." He also notes that the chip company still benefits from the wide recognition of the NEC brand name, though it's hard to determine just how much that contributes to the bottom line.

iSuppli's Lidow remains skeptical that vertical integration benefits either the semiconductor operations or an integrated company's finished products. He notes that the revival of Toshiba's fortunes in semiconductors is due overwhelmingly to growing demand for flash memory even though the company is not a significant player in flash memory-based products. He also thinks the failure of the Japanese to penetrate both the global mobile handset market as well as the market for mobile phone chips shows how vertical integration "holds back both the parent and the subsidiary."

"We see tremendous evidence that chip subsidiaries do worse than fully independent companies," Lidow says.

One exception to this rule is Samsung. It became the world's second-largest chip producer in 2002 while at the same time growing virtually all of its myriad electronics businesses. "Samsung's semiconductor company is left to do what they need to do, and they operate a lot more along the lines of an aggressive independent company," says Lidow.

  
"TOSHIBA WANTS TO EXPAND (it's customer base) to local businesses, and we are targeting China and India." —Shozo Saito, executive vice president, Toshiba Corporation

Samsung and Sony show that vertical integration can work, according to McClean says. "There is nothing inherently wrong with vertical integration. A company has to be dedicated to whatever business model it picks to follow."

Rather than worrying about their degree of independence, Japanese executives believe they are positioned to learn from recent experience and move ahead with new strategies. "We had very serious discussions internally last year to set up a new three-year plan to turn us around, and we are now back into growth," says Renesas' Ito.

Thanks to the missed opportunity in mobile phones, the Japanese semiconductor makers all recognize they need to grow their customer base beyond their groups and beyond Japanese companies. Most of these companies have been increasing their sales outside of Japan and especially in Asia. At Toshiba, for example, sales in Asia last year overtook sales in North America to become the second-largest geographical segment after Japan.

But Toshiba's Saito admits that most of these "overseas" sales are in fact to transplanted Japanese factories. "Toshiba wants to expand [our customer base] to local businesses, and we are targeting China and India," he says. To do so, it has been beefing up sales channels and customer support offices in both countries.

Toshiba is hardly alone. This past April, NEC opened the China Digital AV Application Center in Beijing to support the adoption of its Enhanced MultiMedia Architecture SoC by designers. The center hopes to ramp up to 30 engineers within this fiscal year and add another 80 by 2008. "We are building a market among the design houses and ODMs that produce designs and products for China's household electronics makers," says Nakajima.

Renesas' Ito acknowledges that the Japanese face cultural factors in moving into these new markets. "Most of the business leaders there were educated and trained in the U.S., and their thinking about financial markets [and] their corporate philosophies are pretty much American," he says. There is also a language barrier for the Japanese. "But we have to overcome those issues," he says.

The Japanese companies have also realized that they can't do everything. For Toshiba, one big target is flash memory. Its Cell processor is also being used in Sony's PlayStation 3, and it is used in TVs and DVDs. NEC Electronics is targeting SoCs and microcontrollers. Nakajima says NEC has the top share of the market for high-end 32-bit microcontrollers and the second-largest share of the world market for LCD drivers.

Ito cites Renesas' radio frequency ICs as providing one area of growth. He adds that the company is finally starting to crack the mobile phone market with a chip that enables mobile phones to receive digital TV broadcasts.

iSuppli's Lidow says the narrowing of the focus is showing positive results. Toshiba gained 7.6 percentage points of market share in flash memory in the first quarter. Elpida also increased its share of the market for high-performance DRAMs. Lidow notes that Renesas' sales of 16-bit microprocessors grew faster than the overall market.

Forward Concepts' Strauss says that Japanese companies have also found a "back door" into the mobile phone market in audio chips that play ring tones and MP3 music files. Strauss says a recent study by his firm indicates that mobile phones now represent the largest market for audio chips. The market leader is niche chip maker Yamaha, with 48 percent of a nearly billion-dollar market. The rest of the market is split among Oki, Rohm, NEC Electronics and a Taiwanese company.

Another bright spot for the Japanese is the increasing use of semiconductors in automobiles and the growing clout of Japan's automotive industry. The Japan Electronics and Information Technology Industries Association (JEITIA) estimates that worldwide demand for semiconductors used in vehicles grew 19% in 2003, to $14 billion, and will rise to $20.9 billion in 2007, making this possibly the fastest-growing segment of the chip market.

The JEITIA's report notes that the average number of microcontrollers in each vehicle has jumped from 10 to 50 over the past decade. Not surprisingly, the Japanese makers are happy to see Japan's automakers expanding their presence in the global market. "We have a close relationship with Japan's car manufacturers," says Ito.

IC Insight's McClean says it's not clear yet if all this activity will be enough to initiate a full turnaround or just allow the Japanese to tread water. But whichever way they move, they won't ever again be the companies they were 10 years ago.

Are Japan's electronics companies poised for a comeback? Send your thoughts to feedback@eb.reedbusiness.com.