Joe Horowitz has been in venture capital for three decades, so he knows how to spot a tech bubble. Perhaps needless to say, the current market has him more than a little spooked.
His evidence? The number of start-ups that are being valued at $1 billion or more. Reaching that mark is such a feat that Silicon Valley observers have taken to calling such companies "unicorns." But according to CB Insights, there are now 42 members of the billion-dollar club.
"When we start throwing around the billion dollar valuation number in such a casual way, then it is a sign that we are losing some perspective," said Horowitz, managing general partner at Icon Ventures, which until last week was called Jafco Ventures. "Building a company that is truly worth a billion dollars or more takes a lot of work and a lot of smart people. You have to be at the right place at the right time to truly disrupt a big market relative to traditional players."
For Horowitz, that's a painful reality, because his Palo Alto, California-based firm just last year closed a $260 million fund, an increase of $110 million over Jafco's biggest prior fund. So Horowitz has to put money to work in a market that he admits is overheated.
If he were a public market investor, it would be a perfect time to go conservative, with cash, bonds and dividend-paying stocks. But playing it safe is not part of a venture capitalist's lexicon. Start-up investing is, by definition, high risk and high reward.
"We still have to be in the market and try to navigate either headwinds or even winds at our back appropriately," Horowitz said.
Venture investors poured $48.4 billion into U.S. start-ups last year, by far the most since the peak of the dot-com bubble in 2000, according to the National Venture Capital Association. Venture firms, including Icon, reeled in $29.8 billion, the largest haul since 2007, the association said.
Read More2015: The year Uber gets an IPO?
Horowitz said his firm increased the fund size because it's having to write bigger checks. In the Jafco days, the firm would invest around $8 million as part of a $15 million financing round. Now, Icon is leading rounds of up to $25 million by putting in as much as $15 million.
A big problem with the swarm of billion-dollar start-ups is not just the risk that private market investors are overpaying, but also that there's less upside for the public markets if and when those companies conduct share sales. If public investors aren't getting desired returns, they'll be less likely to participate in future offerings, which could clog the whole system.
"You need to leave room for the public to make money—that's what drives an IPO market," Horowitz said. "When we had companies going public at half a billion dollars and within a few years worth 3 to $4 billion, then the public feels rewarded for taking those risks."
Horowitz's firm, while not as well known as high fliers like Sequoia Capital, Benchmark and Greylock Partners, is on a good run of late. It backed security software companies FireEye and Proofpoint as well network security vendor Palo Alto Networks. They have gone public in the past three years and are valued at $2 billion to $10 billion.
Icon typically invests in start-ups in the second and third rounds of financing, after they've shown some level of traction and have momentum.
Horowitz says that one way he looks to play the market in this environment is by finding companies that can sell their products regardless of macro conditions.
One example: A year ago, the firm co-led a $15 million investment in Zephyr Health, which uses big data analytics to help health care companies get their drugs and devices to patients as quickly as possible. Regardless of whether there's a boom or bust, Horowitz says, Zephyr's services will be in demand from customers such as Gilead and Medtronic.
Zephyr helps drugmakers understand the competitive landscape and where diseases are spreading and assists in navigating the regulatory environment.
"Those are real solutions to real challenges that are important to their business either during good or bad cycles," Horowitz said.