Magazine

New Enterprise Piles On The Risk


During the bubble, New Enterprise Associates was the Oldsmobile of venture capital. As rivals made billions on sexy e-commerce plays like eBay Inc. (EBAY

) and operated with a half-dozen partners, NEA's big hits came on lower-profile companies such as 3Com Corp. (COMS

) and UUNET Technologies Inc. And with a staff of 45 investment pros, huge by Silicon Valley standards, NEA's rep was like IBM's: smart and consistent, sure, but also slow and not too innovative.

No more. Last month, NEA announced a new $2.5 billion venture fund, the largest in its 28-year history and the second-largest the industry has ever seen. More surprising is that NEA will spend up to half of it on high-risk deals most VCs spurn: mega-investments in money- losing businesses, many of which haven't gotten their technologies to work yet. Some of the money will prop up cash-burning public companies that can't otherwise raise capital. In other deals, NEA will acquire unproven drugs -- many with U.S. regulatory approval still years away -- from Big Pharma outfits and build startups around them. "We've expanded the definition of what we do," says NEA managing general partner Peter Barris, a former GE Information Systems executive.

Already, NEA has put $25 million into the first round of funding for CoGenesys Inc., a spin-off from Human Genome Sciences Inc. (HGSI

) that's developing drugs for heart failure and white-blood-cell disorders. NEA's money was part of a $55 million financing, 20 times what a Web media startup might raise so early on.

This is swashbuckling stuff. If a company's technology is unfinished or hasn't landed customers, a VC firm usually puts up only a few million dollars initially. That approach shoulders what VCs call "technology risk" while minimizing financial risk. When VC outfits make big bets, they usually insist the technology be finished and the company be cash-flow positive. The size of the investment adds financial risk, but only after technology risk is resolved. NEA's fund is unusual because it takes financial and technology risks at the same time. "I think it's definitely new," says Josh Grove, an analyst at VentureOne. "We haven't seen anything quite like what they're doing."

Why do it? Partly because of the moribund initial public offering market. Only 29 venture-backed companies went public in the first half of 2006, down from 42 in early 2004, the only strong IPO year since 2000. To generate returns, NEA must take more chances. It has recruited managers from IT and life-sciences industries, including ex-Guidant Corp. chairman Jay Graf, and says it can manage tech risk with big dollars at stake. "If we don't have exactly the right person to understand an idea, that person is never more than a call away," Barris says.ACCESS IS EVERYTHING

NEA was able to sell the concept to institutional investors who put up the $2.3 billion mostly because of its own reputation. "They have a lot of [industry] expertise," says Thaddeus I. Gray, managing director of Abbott Capital Management, which advises institutions on private equity deals, "and top-tier firms like that get access to the best deals." Erik R. Hirsch, chief investment officer at adviser Hamilton Lane, says NEA can hedge risk across an 80-deal portfolio well enough to be a top performer. "We'll take risk, but we'll be compensated for that risk," Hirsch says.

Still, the idea of taking big chances on unproven technologies makes some rivals shake their heads. TA Associates Inc. does large, late-stage venture deals but avoids tech risk and major gambles on money-losers. "We do transactions like that -- by mistake,"says TA CEO C. Kevin Landry.

But the weak IPO market is making VCs rethink tactics. Oak Investment Partners has a similar strategy for a new $2.6 billion fund. For better or worse, NEA's new attitude toward risk is catching on. By Timothy J. Mullaney


Later, Baby
LIMITED-TIME OFFER SUBSCRIBE NOW

(enter your email)
(enter up to 5 email addresses, separated by commas)

Max 250 characters

 
blog comments powered by Disqus