San Jose, Calif. One of the biggest stories of 2006 seemed to come out of nowhere but really didn't: the surge in private-equity buyouts of major chip vendors.
Texas Pacific's buyout a decade ago of Zilog seemed an aberration. After all, stocks generally were up and to the right, and the Internet and communications-sector bubbles weren't yet fully inflated. In retrospect, however, Zilog was a harbinger for the industry at large.
It took the bubbles' bursting and deep semiconductor recession to make matters clear. The industry was maturing and its growth rate slowing. Wall Street stopped valuing chip companies, even as the color returned to the sector's cheeks in 2002-03.
Flash forward to 2006. Investors have billions burning holes in their pockets, and rather than push that money into the startup end of the business, they've been looking to take companies like NXP and Freescale private to find efficiencies. In addition to those two high-profile deals, Advanced Semiconductor Engineering Inc. (ASE), Jazz Semiconductor Inc., Toshiba Ceramics and others have been acquired by private-equity firms in recent months or are headed down that road. Some companies are simply waiting for the right offer. Said one venture capitalist: "The smart money is rolling up the industry."
The dramatic changes in the equity markets are expected to alter the semiconductor landscape, though whether the shift is for the better or worse depends on whom you ask. While some suggest that the trends in the equity sector are creating wealth and providing better returns for shareholders, others fear the changes point toward further consolidation, a slowdown in startups and, worse, less innovation.
Also up for debate is whether the industry's capital engine just needs an overhaul or is broken beyond repair. But without doubt, a sea change is under way in the three main pillars of the equity sector: the public markets, private equity and venture capital. Simply put, private-equity firms--high tech's new power brokers--are hot. VC funding and IPOs are not.
But VCs and private-equity firms alike are lukewarm on semiconductors. That confounds analysts, since chip demand "is stronger than ever," said Bill McClean, president of IC Insights Inc. (Scottsdale, Ariz.).
Raising money in traditional equity markets--via venture funding and IPOs--is maddening in today's business environment, said Dipanjan Deb, managing partner at private-equity firm Francisco Partners (Menlo Park, Calif.). "It takes a lot of money to take a company public," Deb said at a recent MicroVentures conference in Silicon Valley. "There is a lot more money in the private-equity sector. So the motivation to go private is higher than ever."
Deb expects a new wave of equity buyouts, especially for midsize chip makers with little or no leverage.
In sharp contrast, U.S. venture funding for IC companies fell in 2006 for the third straight year, with fewer deals having come to the table since the IC bloodbath of 2002, VentureOne, the VC research arm of Dow Jones & Co. (New York), reports.
Moreover, only two domestic companies filed for a public offering in 2006, down from four a year ago, VentureOne said. Since an offering must generate $500 million in proceeds to grab Wall Street's attention, the decline isn't surprising.
These are ominous trends for both established chip makers and startups, many of which count on the VCs and public markets to fund operations, product development and R&D.
"It's a very tough market," Pierre Lamond, a semiconductor veteran and high-profile venture capitalist with Sequoia Capital (Menlo Park), said at the MicroVentures conference.
"It would be impossible to start and invest in a new Nvidia today," Lamond said, referring to Nvidia Corp. (Santa Clara, Calif.), the only graphics chip maker to retain its independence. "To get a company going, you need a lot of money. If you made a chip to compete with Nvidia, you would have to spend $30 million on the first product alone."
The barriers to VC funding--soaring design costs, lackluster returns and uncertain public markets--are growing, Lamond said. Chip design costs are approaching $50 million to $75 million at the 65- and 45-nm nodes.
What really depresses Lamond is that 62 percent of U.S. semiconductor startups, having "no place to go" in the public markets, must continue to rely on VC funding even in "later rounds."
Not long ago, to be considered successful, a chip startup was expected to go public after "five to seven years" and raise at least $100 million in the IPO, Lamond noted. Now that startup is expected to go public "in four years or less" and raise no less than $500 million.
Venture capital funding has not dried up for semiconductor startups with innovative ideas, said William Tai, general partner for Charles River Ventures. But "we are very, very selective," Tai acknowledged. And he conceded that VCs, believing they can't make a decent return on semiconductor startups, generally are seeking more promising opportunities in hotter sectors, such as clean energy and Web 2.0.
The U.S. chip industry saw 148 venture-capital deals or transactions in 2005, said Jessica Canning, senior research manager for VentureOne, with U.S.-based startups raising $1.9 billion. This year, VentureOne estimates, 135 U.S. deals raised $1.8 billion.
Scourge or savior?
In a panel discussion at the MicroVentures conference, Francisco Partners' Deb predicted the number of private-equity buyouts will accelerate in the IC sector, causing seismic shifts that will roil the remaining players.
"I think you'll see more Freescales,'' Deb said. ''I also think you'll see a plethora of sub-$1-billion buyouts."
-- Additional reporting by Brian Fuller