Here is a fascinating story by Katie Benner about a fallen unicorn with the puzzlingly generic name “Good Technology.” Good was formed in 2009 by the combination of a startup and a division of Motorola. It ”raised about $300 million in equity and debt,”1 hitting a peak valuation of more than $1 billion in early 2014. It filed for an initial public offering in May 2014, but then it delayed and ultimately cancelled the IPO. Its board turned down an $825 million acquisition offer in March 2015, “confident that Good would be valued at around $1 billion when it went public.” But then it ran into financial trouble and was ultimately sold for $425 million in September.
That $425 million was more or less enough to make whole Good’s venture capital investors, who owned preferred stock,2 but was a shock to employees, who had received common stock and stock options as part of their compensation, and who “discovered their Good stock was valued at 44 cents a share, down from $4.32 a year earlier.” Some of these employees had bought stock in the open market, in addition to the shares they’d received in compensation; others had paid cash taxes to exercise their options in amounts that exceeded what the shares ended up being worth. “Employees essentially ended up paying to work for the company,” says one of them.
As a further indignity, the buyer was BlackBerry.
The interesting question is, what went wrong for these employees? What caused their bad outcome? Of course there is a simple answer, which is that they were employee-shareholders of a company, and the company’s stock went down, and so their share holdings lost value. That happens to lots of employees at lots of companies, and if Good’s employees want to complain, they are unlikely to find too many sympathetic listeners at Blackberry. Here are the last five years of Blackberry’s stock:
Oh tell me more about how your unicorn died.
Still it feels like there is some difference between Good and Blackberry, some reason we should feel worse for Good’s employee-shareholders than for the average public-company employee-shareholder whose stock goes down. An obvious possibility to start with is that Good was a private company, so there was no public trading market for its stock. That has two bad effects. One is really basic: If you get stock from your employer, and there’s no market where you can sell it, then you are stuck with it, and if it loses value you are hosed. The other bad effect is that, without a public market for the stock, there is no way to know how much it is worth, or rather, how much the market thinks it’s worth.
Except that there was a market for Good stock, and Good employees — at least some of those mentioned in the article – were buyers, not sellers:
Companies that buy employee shares offered some Good workers about $3 a share for their stock in the first half of the year. But based on their belief in the company’s robust health, the employees refused. Others bought Good common stock in August, when it was valued at $3.34 a share, according to individual employee tax documents reviewed by The New York Times.
You know what I think: Private markets are the new public markets, and private companies now have many of the amenities that used to be limited to public markets. So private companies can have trading markets for their stock, letting employees get liquidity at market-derived prices. Good’s stock wasn’t listed on Nasdaq or anything, but employees could sell if they wanted to, and could get a sense of the arms-length, third-party, market-ish price of their stock. (In fact, Good was one of the “most-watched companies” on SecondMarket in 2012, suggesting that employees did pay attention to the market price.) But they chose not to sell — they chose to buy – because they thought the market price was too low.
Another possibility is that, unlike public companies, Good wasn’t subject to Securities and Exchange Commission rules about disclosure, and so perhaps Good’s investor-employees were deceived by bad or insufficient financial disclosure. This does seem to be part of the story: Benner notes that, when employees were buying in August, they “had little idea that an outside appraisal firm had valued Good at $434 million and the common stock at about 88 cents a share as of June 30,” or for that matter that “by late July, the board knew that Good would run out of cash in 30 to 60 days” and was frantically negotiating a deal with BlackBerry.
But this isn’t an easy story about bad disclosure, because Good had filed to go public, and so actually had a fair amount of disclosure. It had filed its audited financial statements as of the end of 2014. It didn’t file subsequent quarterly financials, but ”at a May company meeting,” Good’s then-chief executive officer, Christy Wyatt, “said the company missed financial projections,” so there was at least some financial update. Its IPO prospectus also included a pile of ominous risk factors, including about Good’s historical and ongoing losses from operations, and a grim discussion of its liquidity situation noting that Good’s “ability to satisfy our total liabilities at December 31, 2014 and to continue as a going concern is dependent upon either the successful completion of this offering or the timely availability of other long-term financing.”4
Good didn’t disclose its conversations with BlackBerry, but then, most public companies try not to disclose merger negotiations until there’s a deal. (On the other hand, Good “was forced to borrow $40 million” from BlackBerry while negotiating the sale, which would surely be disclosable, and worrying, for a public company.) And it didn’t disclose that lowball outside appraisal, but then, most public companies don’t get outside appraisals.4
So it’s a mixed bag, but it looks to me like Good’s employees were on notice — at least in the sort of formal, SEC-filing sort of way that public company employee-shareholders are on notice — that the cash situation wasn’t that great, and they had financial disclosure that was more or less comparable to what you’d get from a public company. But they remained optimistic, held their stock and even bought more.
What else? There is the fact that Good’s board of directors turned down an $825 million offer in March, gambling on doing better in an IPO, and ultimately did worse. Again, this is not unique to private companies: Public companies often turn down acquisition offers and end up regretting it when their stock goes down instead of up. But there is a difference: Most public-company boards are made up mostly of independent directors who own common stock in the company, but Good’s board – as is common for private tech companies — contained a lot of directors representing Good’s venture-capital investors, who owned a lot of preferred stock. “The six investors on the board had preferred shares worth a combined $125 million,” writes Benner, “almost the same as all 227 million common shares outstanding.”
Preferred stock is different from common: Since Good’s venture-capital investors were first in line to be paid back, turning down the deal in March was less risky for them than it was for common-shareholder employees. Both the preferred shareholders and the common shareholders would do much better at $1.5 billion than at $825 million, but the preferred shareholders wouldn’t (and didn’t) do too much worse at $425 million than at $825 million. The common shareholders did much, much worse. As Felix Salmon says, “the huge problem is divergent incentives between common and preferred,” and the board of directors was aligned with the preferred.
That is true, but it still isn’t the whole story, at least not for Good. After all, Good’s employee-shareholders faced more or less the same choice as the board: Cash out at a mediocre price ($825 million for the board, $3 per share for the employees) early in 2015, or gamble on a higher price in the IPO. And they both made the same choice: to gamble.5 The board and the employees had, on paper, different incentives, but they came to the same decision about how much risk they wanted to take. They were both pretty cool with risk.
I think that might be the biggest difference between unicorns whose employees get gored when their valuations drop and public companies like BlackBerry whose employees’ fortunes are also tied to their stock price: Public-company employees are just supposed to be risk-averse about their employer’s stock. It’s a psychological or sociological difference between public and private companies, not a structural one. After Enron, at least, people know that they’re not supposed to tie up too much of their financial wealth with their employers’ fortunes, that they’re supposed to diversify their career and financial risk. If part of your pay is stock in your own company, it’s a little weird to then buy even more stock. And if part of your pay is in stock options, it’s a little weird to pay cash taxes to exercise those options; lots of public companies let employees give up some of their shares to cover the taxes, rather than write a separate check. When you work for a public company, it is obvious enough that you shouldn’t pay to work there.
But if you work for a — startup? unicorn? venture-backed technology company? — then the expectations seem a bit different. Of courseyou’re gambling on your company; of course you should make a massive undiversified bet of your human and financial capital on your startup employer. That’s the Silicon Valley ethos, and it’s why you left your stable high-paying job at Google or Microsoft or, um, Motorola to go work for a startup.6 You are gambling on tech riches with your career, and the way that gamble pays off is when your unicorn goes public and your pre-IPO shares and options make you fabulously wealthy. Selling those shares before the IPO, to minimize your risk, misses the whole point.
But, again, you know what I think: Private markets are the new public markets, and companies with nine- and ten-digit valuations, hundreds of employees and years of operating history aren’t dorm-room startups. Private companies that are giant one day can end up smaller the next, just like public companies can; private companies that are giant one day are unlikely to go public the next day at ten times their previous valuation. The risk-reward proposition of working at a lot of unicorns looks more like that of working at Facebook in 2015 than working at Facebook in 2004. Private-company trading is catching up with public-company trading; private-company valuations are catching up with public-company valuations. But private-company expectations haven’t quite caught up yet.
- Of that, $80 million was in the form of debt with warrants, the rest in (preferred) equity. (See page 51, etc., of the initial public offering prospectus.)
- Benner’s article includes a chart of preferred-stock prices that peaked above $6 per share in early 2014, before cashing out at a bit above $3 in the September 2015 deal. I don’t quite know the mechanics of that; the preferred shares had liquidation preferences ranging as high as $4.92 per share (see page F-57), but Benner tweeted that ”even the pref got renegotiated in the end,” so rather than fully making the preferred shareholders whole, Good left some value for the common shareholders.
- Wyatt downplayed those concerns in an employee meeting — “She compared the risk section to ‘commercials for some new pharmaceutical product and it’s like 15 seconds about the product and then there’s like 45 seconds of the 15 different ways it can kill you’” — but it’s not hard to imagine a public-company CEO doing the same thing.
- On the other hand, Good did disclose a series of fair value estimates for its common stock, in connection with stock-based compensation grants and stock-based acquisitions. You can find these valuations on pages 84, 85, 88 and 91 of the IPO prospectus. The disclosed fair value starts at $4.61 in September 2012, gets up to a high of $4.92 for in March 2014, falls to $3.88 in May of 2014, and then recovers a bit, ending up at $4.26 in February 2015, the last valuation Good did. Obviously public companies don’t normally do that, but that’s because they can just use their stock’s trading price as the fair market value.
- Again, when I talk about employees here, I mean the ones Benner discusses. Maybe a lot of others cashed out, I don’t know.
- Or whatever Good was; “startup” feels like a bit of the wrong word for a company started in 2009 by buying a business out of Motorola.