Photographer: Hiroshi Watanabe/Getty Images

The Fuzzy, Insane Math That's Creating So Many Billion-Dollar Tech Companies

Startups achieve astronomical valuations in exchange for protecting new investors

Snapchat, the photo-messaging app raising cash at a $15 billion valuation, probably isn't actually worth more than Clorox or Campbell Soup. So where did investors come up with that enormous headline number?

Here's the secret to how Silicon Valley calculates the value of its hottest companies: The numbers are sort of made-up. For the most mature startups, investors agree to grant higher valuations, which help the companies with recruitment and building credibility, in exchange for guarantees that they'll get their money back first if the company goes public or sells. They can also negotiate to receive additional free shares if a subsequent round's valuation is less favorable. Interviews with more than a dozen founders, venture capitalists, and the attorneys who draw up investment contracts reveal the most common financial provisions used in private-market technology deals today.

The backroom agreements are becoming more common as tech companies stay private longer, according to the interviews and financial documents obtained by Bloomberg Business. The practice obfuscates the meaning of a valuation, which can become dangerous down the road because private investors aren't taking the same risks a public-market shareholder would. By the time a company does go public, the valuation it got from VCs may not align with its balance sheet. Just ask Box.

Some VCs defend the practice by saying valuations are just a placeholder number, part of an equation fueled by other, more important factors. Those can include market share, growth projections, and a founder's ego. The number is typically set by the company and negotiated alongside various provisions designed to protect a new backer's money. That often comes at the expense of employee shareholders and earlier investors, whose holdings are diluted to make room for new entrants. If you've seen the movie The Social Network, you have an idea of how that works.

"These big numbers almost don't matter," says Randy Komisar, a partner at venture firm Kleiner Perkins Caufield & Byers. "Those numbers are just a middling shot at a valuation, and then it's adjusted later" through various legal techniques, if an earlier valuation was too high, he says.

For Uber to get to $40 billion or Airbnb to $20 billion, you'd need to get a little creative with the variables underlying that logic. Since private tech companies often lack earnings or enough historical data to inform projections—or, in the case of Snapchat, any significant revenue—investors can't rely on the metrics available for public companies. If there were a math problem for determining a tech startup's valuation (for the record, there isn't), it might look like this:

Illustrator: 731

Founder's hopes and dreams

A founder often starts off with a number in mind, based on the startup's last valuation, the valuations of competitors, and, for good measure, the valuation of the company's neighbor down the street. It can become a sort of arms race. When Slack Technologies founder Stewart Butterfield decided to raise $120 million, his goal was to crack a $1 billion valuation for his corporate software startup. The other appealing thing about a big number: It means founders don't have to give away as many of their shares to raise a lot of capital. 

Billion-dollar companies join a club of "unicorns," a term used to explain how rare they are. But there are more than 50 of them now. There's a new buzzword, "decacorn," for those over $10 billion, which includes Airbnb, Dropbox, Pinterest, Snapchat, and Uber. It's a made-up word based on a creature that doesn't exist. “If you wake up in a room full of unicorns, you are dreaming," Todd Dagres, a founding partner at Spark Capital, recently told Bloomberg News.

Venture capitalists may remind founders not to get carried away because they may need to raise money again soon, perhaps in a less-favorable market. Fundraising at a lower valuation is known as a "down round." It's a major Silicon Valley no-no in terms of perception, and it can have negative effects, depending on the other stipulations of the agreement. (More on that later.)

How fast it's actually growing

A tech startup's cash flow is less important than you might think. It's something investors look at for a sense of how quickly a startup is growing its revenue, if the company has any. Financiers also look to find the number of people using the product, regardless of whether they pay for it. Investors salivate over what's called "hockey-stick" growth curves, indicating massive uptake. Costs, especially operations costs, are largely ignored for fast-growing companies. (To borrow an old aphorism, you have to spend money to make money.) Investors hope that early growth is indicative of future growth, although this can overlook the impact of early adopters.

Downside protection

Here's where things start to get tricky: Buried in their corporate filings, startups tuck away all sorts of provisions that reward investors for accepting these mega-valuations. The practice is more regular and egregious in financing rounds for mature companies. Their capital requirements tend to be much larger, so they must turn to more sophisticated investment firms that demand these kinds of terms. Startups that are generous with these guarantees can garner much higher valuations.

Each provision covers different ways to make sure new investors get paid back, even if disaster strikes, if an initial public offering gives the company a market cap far less than its private number, or, more commonly, if the startup has to raise money again at a lower valuation. One stipulation, called senior liquidation preference, ensures that a certain group gets its money back before anyone else, including employees. Another class, called downside protection or ratchets, automatically grants additional shares in the event of a declining valuation, removing a great deal of risk that the stake will ever lose value.

"When we're talking about these decacorn-type valuations, generally, the way these deals are structured is: They could be worth that much; it's not a fake number, but it's not the same as buying the stock in the public market," says Jason Lemkin, managing director at Storm Ventures. "There's always some kind of warrant, some kind of ratchet, some kind of downside protection."

Investor FOMO

In addition to minimizing the chance of getting burned, investors fear missing out on the next Google or Facebook. A severe case of FOMO can cause some to do crazy things to get into the hottest deals. But they don't just want the promise of an IPO or an acquisition someday; they want it to happen soon.

The firms giving cash to mature startups are often hedge funds, investment banks, and sovereign wealth funds, which can't afford to wait seven to 10 years for a bet to deliver returns, as a VC can. If a startup's board of directors promises a short, plausible timeline for taking the company public, investors are more likely to give leeway on eye-popping valuations. In exchange for coughing up hundreds of millions to keep the lights and espresso machines on, these investors may also arrange to get shares at, say, a 20 percent discount to the IPO price. Or they may still push to get paid back first and in full.

"When investors are demanding these provisions more, especially things like senior liquidation preferences, that tells me that there's more risk and uncertainty in the market, and they're really protecting themselves," says Barry Kramer, a partner at Fenwick & West, a law firm involved in some of the Valley's hottest deals. "I don't like to use the word bubble, but it's a sign of concern that valuations are getting too high."

There is, however, a further number that's grounded in reality. It's the valuation based on common stock, which is generally what employees receive, and it's calculated by professional auditors. That figure usually isn't anywhere close to the headline number. "It may be 30 percent less; it may be 50 percent less," Kramer says. "They don't necessarily disclose that to the public until the time of their IPO."