Advanced Semiconductor Engineering: New leader
ASE works to maintain its top spot in packaging and test .
By Drew Wilson -- Movers & Shakers, 11/1/2004
Advanced Semiconductor Engineering (ASE) is on a roll. A series of aggressive expansion moves, combined with cost-saving measures and a little luck from the chip market, has pushed the Taipei-based company to the top of the IC package and test industry. ASE slightly edged past rival Amkor Technology at the end of 2003, reporting $1.67 billion in revenue, compared with Amkor's $1.61 billion.
 |
| Jason Chang, Chairman, ASE |
ASE's management believes that 2004 results will bolster the company's leading position. Jason Chang, the company's chairman, expects 50 percent revenue growth.
Perhaps the most significant move for ASE this year has been the acquisition of NEC Electronics Corp's IC packaging and testing operation in Japan. The deal included a four-year backend-services agreement with NEC, yet did not preclude ASE from serving outside customers.
The purchase, based on assets, was worth close to $100 million, according to the company. Andy Kung, analyst at Credit Suisse First Boston in Taipei, says the operation will generate substantial revenue while also opening a door for ASE. "Through this purchase, ASE entered the Japanese market," he says.
Only about 10 percent of Japan's backend business is outsourced, according to ASE's Chang, compared with 30 percent worldwide. The NEC deal signals that Japan's traditional reluctance to outsource is waning, which represents a big opportunity for ASE, he says. "Japan has to use lower-cost partners because its vertical integration model is not working," Chang says.
Chang is something of an unlikely founder for the world's biggest semiconductor packaging house. He earned an electrical engineering degree from National Taiwan University, but never worked at a multinational electronics firm. With the help of his brother and capital from his family's construction business, Chang founded ASE 20 years ago.
"There was a [packaging] industry on a small scale at the time, and I just felt that in Taiwan electronics would be the future," he explains.
ASE instituted some novel approaches to the backend that have proven to save the company money. In-house substrate production serves as a prime example. After buying substrate at high cost from Japan, ASE decided to produce the material itself. The move was shrewd: substrate material accounts for 30 percent of the cost of packaging a BGA (ball grid array) chip. For flip-chips, the substrate can account for 50 percent of the cost. ASE now controls supply, reduces costs, and, by extension, increases its profit margins.
ASE's net margins have been impressive. The company reported a 4.8 percent increase for the full year 2003. Meanwhile, all its major competitors were in the negative zone.
ASE still buys 50 percent of its substrate from outside suppliers in order to ensure that internal prices stay competitive with the open market, adds Tien Wu, President of ASE in the Americas, Europe, and Japan. To further reduce costs, ASE is already running substrate operations and high-volume assembly of camera modules for an unnamed IDM customer in China.
The company is also planning to set up assembly and test operations on the mainland to get closer to its customers, including the major foundries and multinationals such as Infineon Technologies, Philips Semiconductors, and Qualcomm.
Entering China, though, is complicated by politics and presents a possible threat to ASE's competitiveness. The Taiwan government has been reluctant to give blanket approval for semiconductor investment in China because of concerns over job losses, investment losses, and the transfer of advanced technology to the mainland.
ASE has in the past been unable to get permission to set up in China, while rivals like Amkor have invested hundreds of millions in mainland backend plants.
Chang says ASE's latest request has received initial approval. He is confident formal approval will come by year's end. If it goes through, Wu hopes to have volume backend operations up and running by early 2005, depending on the business climate. He expects the first two phases of the production shift will build up the critical mass of skilled workers necessary for the packaging plant. "The longest leadtime in China is in building your workforce," he says.
The backend industry has traditionally added too much capacity in good times. But the players have avoided that this year, according to Freddie Liu, CFO of ASE's test division. "The whole backend sector has been acting rationally ever since the last downcycle, so people put up their capacity with great cautiousness," Liu says.
In 2004, ASE invested more than its competitors by spending $700 million, mainly on test equipment. Amkor, the closest competitor, has forecast $500 million. Roughly 20 percent of ASE's annual revenue comes from test, and ASE wants to increase that percentage. Moreover, gross margins for testing outshine packaging work.
But companies have been reluctant to outsource test because it can expose the intellectual property of their chips. Wu estimates only 15 percent of global chip testing is outsourced, compared with up to 35 percent of assembly.
ASE argues the issue from a cost-savings perspective. Testing machines are becoming expensive, and their short lifespan means they are not cost-effective for chip companies to buy. ASE can spread the equipment cost across multiple customers, resulting in an overall lower cost. "Theoretically, we can pass these cost savings on to customers," Liu says.
The company's management team doesn't believe the widely expected chip-market slowdown in 2005 will impact harshly on ASE. Continued outsourcing will be the company's growth engine.
"We lower our costs to encourage more outsourcing," Chang says.
