Does it make sense for countries like China and India to try to compete in the ultra-expensive semiconductor industry? Indian officials seem to think so. They’re putting together new policies designed to make investors willing to plunk down huge sums – the price tag for a low-end fab is in the hundreds of millions and well over $1 billion for more cutting-edge ones. In March, the government unveiled its “Special Incentive Package Scheme,” a package of incentives for chipmakers. And Indian officials will be revealing more details of this plan soon.
China caught the chip bug well before India did, with the Chinese government helping new companies like SMIC and Grace get started in the early ’00s. And if the Chinese experience is any indication, the Indians might be better off thinking twice before pushing ahead with their chipmaking ambitions. Grace, which was launched with support from the son of former Chinese President Jiang Zemin, is perhaps known best in the U.S. for its role in the unpleasant divorce of Neil Bush, the brother of George W. Bush and member of the Grace board. Grace was supposed to have had an IPO by now but breaking into the chip business isn’t easy and that IPO still hasn’t happened.
SMIC did make it to market, with a spectacular IPO in 2004 in New York and Hong Kong that raised $1.8 billion. SMIC has grown quickly, with fabs in Shanghai, Beijing and Tianjin and projects underway with local governments to operate new fabs in Wuhan and Chengdu, but the company is having trouble translating sales into profits. Last Friday, SMIC revealed its Q1 earnings of $8.8 million. But that was largely the result of a one-off sale of equipment that brought in $27 million. Strip that out and SMIC is in the red – again. “The only reason the company has been able to report profits at the operating level is simply because of non-recurring asset disposal,” Warren Lau, an analyst with Macquarie in Hong Kong told me. But what about the idea, popular among many boosters of Chinese chipmakers, that Chinese chipmakers have a special advantage in catering to the country’s huge demand for semiconductors? Lau doesn’t buy it. “Location is not the single most important consideration. It really doesn’t matter if you are based in the U.S., Taiwan, Singapore or China.” (Maybe he should add India to that list, too.)
In China, at least, there's no shortage of governments willing to throw around money to get into the chip business. “The entry barriers to build a fab are not that high," Lau adds. "As long as you have money you can build a fab. The important question is whether you can load the fab.” In other words, turn it into something that can make money. To do that, new chipmakers need customers, tech, and talent. “These are the high entry barriers that a lot of people don’t see," he says, "rather than ability to build a white elephant."