Are you terrified?

Photographer: Matt Cardy/Getty Images.

Unicorns Aren't So Beloved Anymore

Matt Levine is a Bloomberg View columnist. He was an editor of Dealbreaker, an investment banker at Goldman Sachs, a mergers and acquisitions lawyer at Wachtell, Lipton, Rosen & Katz and a clerk for the U.S. Court of Appeals for the Third Circuit.
Read More.
a | A

The way a bubble works is, people get enthusiastic about a thing, so they buy a lot of it, and its price goes up.  Then people realize the price is too high, so they sell a lot of it, and the price goes down. All of your favorite market things -- supply and demand, buyers and sellers, the price mechanism -- get a workout. It is a good time, unless you bought the thing near the top.

Today's Wall Street Journal has a terrific story about how mutual funds that bought stakes in large closely held technology companies are now writing down some of those stakes:

BlackRock Inc., Fidelity Investments, T. Rowe Price Group Inc. and Wellington Management run or advise mutual funds that own shares in at least 40 closely held startups valued at $1 billion or more apiece, according to securities filings analyzed by The Wall Street Journal.

For 13 of the startups, at least one mutual-fund firm values its investment at less than what it paid, the Journal’s analysis shows. Those firms are valuing the 13 companies at an average of 28% below their original purchase price.

One way to read this story is that there was a unicorn bubble and it popped. But it has played out, as it were, without markets -- or at least, without market trading of those unicorns' shares. (At least, on the way down.) Instead, the mutual funds' portfolio managers got enthusiastic about unicorn shares, so they bought a lot of them, and their prices went up. And then the mutual funds' valuation committees realized that the prices were too high, so the prices went down. Selling wasn't required: The committees just decided on the price, and that was the price. 

I mean, it wasn't exactly the price. A mutual fund valuation committee's decision that, say, Zenefits is worth 30 percent less than it was a few months ago doesn't necessarily mean that anyone bought or sold shares at the new lower price. But in other ways, those numbers really are performing the functions of a price mechanism. For instance, lower "prices" are telling investors to allocate less capital to unicorns: 

No U.S.-based, venture-backed technology companies have gone public so far this year. Last year, 16 such companies had IPOs, down from 30 in 2014. Their shares had fallen more than 30% on average as of mid-February.

The suffering stock market is likely to keep the IPO pipeline shut to companies that previously raised capital at lofty valuations and don’t want to go public at a lower price.

And the big mutual fund companies in the story -- Fidelity, T. Rowe Price, BlackRock, Wellington -- are themselves making fewer startup investments than they did in previous quarters. The regular workings of the price mechanism are unnecessary: You don't need buying and selling; pure abstract contemplation by committees at investment firms is enough to create a market price. Economists have long debated how prices could be set and resources allocated by a central planner without the use of market mechanisms to signal demand. This is called the socialist calculation problem,  and it is pleasing that late-stage unicorn capitalism has solved it. 

I mean, I kid, a little. These prices are based on market transactions. Just not transactions in the unicorns' stock:

Fidelity has cut its valuation of MongoDB in eight of the nine quarters since Fidelity made its investment in December 2013, valuing the shares 58% below what it paid. The software firm’s revenue roughly doubled to about $100 million last year, according to a person familiar with the matter.

But the company’s last valuation of $1.6 billion is a larger multiple of revenue than at publicly traded companies such as Hortonworks Inc., where revenue has been growing at a similar rate.

Hortonworks's price has fallen even as its revenue has grown, and MongoDB is, in the absence of a public trading market, just a mathematical transformation of Hortonworks. This shouldn't seem too weird: Public stocks, too, often behave like mathematical transformations of each other, as trading algorithms based on historical correlations perpetuate those same correlations.  The public stock market is a very high-powered calculation mechanism for working out those mathematical relationships, but if you can't use the market, a spreadsheet will probably work. 

One thing this suggests is that unicorns aren't much of an asset class: If their performance looks like that of public tech companies, and their multiples look like those of public tech companies, then ... what is the difference between them and public tech companies?

You know what I think! Private markets are the new public markets, and if you want to run a giant company with hundreds of employees, a multibillion-dollar valuation, and millions of dollars raised from public mutual funds, while still calling it a "startup," you can do that now. (You can even call it a "unicorn," if "startup" seems a little low-rent.) But what this means is that private companies are the new public companies, and sometimes public companies' stocks go down. In a world where venture-funded startups exist in a sort of trial-and-error, proof-of-concept phase, and go public when it turns out the concept works, private valuations shouldn't fluctuate unpredictably: You raise money, your thing works, you raise more money at a higher valuation, your thing scales a bit, and you go public at a yet higher valuation. (Or: You raise money, your thing fails, you send your venture capitalists a note with your condolences, and that is that.) The problems of running an operating business for the long term, with revenue fluctuations and competitive pressures and changing market conditions, get worked out in your stock price as a public company. Sometimes it goes down!

In a world where venture-funded startups are also mutual-fund-funded multibillion-dollar companies with massive established businesses used by millions of people, that's not the model any more. Uber works! It is worth a gajillion dollars, in part because investors have wild dreams that it might one day take over the entire multitrillion-dollar global market for transportation, but also because it makes billions of dollars in revenue right now. Zenefits works! I mean, it works the way Deutsche Bank works: There are scandals, and sometimes revenue misses expectations, but it is a real company with a real business. And when the scandals hit, or when revenue misses, its price goes down. That's the way regular companies work, and Zenefits, for all that it is a weird unicorn handing out of oversized novelty checks, is also a regular company.

So in some ways this sounds like a traditional description of private venture-backed companies:

T. Rowe Price said in a statement: “What we are observing now is that the market appears to be bifurcating” into companies that are executing and those that aren’t.

Bifurcation is the old model: Some startups execute, and succeed wildly, and go public; others fail, and die quietly. But that isn't what the Journal's cool interactive unicorn graphic shows. The main graph shows 48 startups valued at $1 billion or more, with their price movements from the first quarter of 2013 through the first quarter of 2016. The best performer is Docusign, up 234 percent, followed by Uber, up 214 percent; the worst is Jawbone, down 69 percent, followed by Zenefits, down 53.6 percent.  There are others in the middle; it is more of a spread and less of a bifurcation. And the spread isn't even that wide! By comparison, the best performer in the Nasdaq Composite Index for that period was up 2,073 percent; the best performer in the more sedate Standard & Poor's 500 Index -- Netflix -- was up 763 percent.  The worst performer in the S&P over that period was down 80 percent; Nasdaq's worst was down 100 percent. Obviously these aren't quite fair comparisons, and there are selection issues with all of these indexes. But the point is, you know, public companies go up, and public companies go down, and private companies go up, and private companies go down, and the big tech unicorns have a long way to go before they're as volatile as the big public tech companies.

The other thing that I think is that modern financial capitalism has allowed companies to unbundle the package that used to come with being a public company. It used to be that going public meant that you could raise lots of money, including from retail and institutions, provide liquidity for investors, become a household name, and use your stock as acquisition currency. But it also meant that you had to make financial disclosures, let anyone who wanted to buy (activists!) or sell (short sellers!) your stock, have fiduciary duties to shareholders, meet with them, subject yourself to their votes. Also you had to subject yourself to the discipline of the market, to daily changes in stock prices that provide a running scorecard of how well you're doing. 

Now each of those strands is available in the private markets, and you can mix and match. More receptive markets and more relaxed regulation make it easier to raise a lot of money without an initial public offering, even with retail investors. Private secondary marketplaces make liquidity easier.  Uber has achieved its fame with ubiquity and Objectivism, not an IPO, and has been acquisitive. Meanwhile private companies can decide how much power and information they want to give outside shareholders, and can raise quite a bit of money even if the answer is "not much." 

They can even avoid the discipline of fluctuating stock prices. Just don't sell to mutual funds! It is an obvious answer, and private companies have noticed:

Some venture capitalists anticipate further markdowns by mutual funds. That could make some startups more reluctant to seek mutual-fund money, since public disclosure of their valuations is watched so closely.

But it is an answer that only goes so far. The great unicorn dream is to be as big and successful as a public company right now, but to have a startup's limitless ambition for the future. Sure, you had a lot of revenue last quarter, but your revenue doesn't define you -- your plans for world domination, your total addressable market, are what's really important. That's a tough argument for public companies to make. It's getting tougher for unicorns too.

  1. I am being super loose about the term "bubble" and hereby disavow this usage. I generally subscribe to the view that, as my Bloomberg View colleague Noah Smith has explained, a "bubble" is something more than a "pricing error":

    A pricing error is when some people think some asset is worth a lot, and they turn out to be wrong. Oops. That’s just another day in the financial markets.

    Economists think that a bubble is a lot more than that. They see a bubble as a game of the “greater fool,” where everyone buys at prices they know are inflated because they think they can find someone else to sell to at an even more inflated price. This is called a speculative bubble.

    In this post I am perhaps not doing enough to distinguish between them.

  2. This is a dopey simplification, but can I recommend to you, in the strongest possible terms, Francis Spufford's "Red Plenty"? It is my favorite novel about economics, though that is insufficient praise.

  3. "In such a world dominated by index and algorithmic funds historically logical correlations between different asset classes can remain in place long after they have ceased to be logical." The index fund is in some ways a kind of algorithmic investing: The algorithm is, whenever you buy one company in the index, buy all of them.

  4. Though I hope most regular companies don't do this after raising money:

    The next day, Zenefits trumpeted the news in a recruiting note to potential employees. “In case you missed it, we just closed our EPIC series C of $500MIL at a valuation of $4.5BIL which makes us a UNICORN! That would be awesome for any company—but for a scrappy, two year old startup it’s LEGENDARY!”

  5. This exaggerates as several of those companies don't have history going back to 2013. Docusign does, but Uber doesn't, since it doesn't seem to have sold shares to mutual funds until June 2014. The worst performer that goes back to the beginning of the chart seems to be Evernote, down 36.3 percent through the fourth quarter of 2015.

  6. I'm measuring from Jan. 1, 2013, through Jan. 1, 2016, which may not be quite the same period. The best Nasdaq performer is Anacor Pharmaceuticals. Data is from the Bloomberg terminal's MRR function.

  7. Wait, what? The Nasdaq's worst performer is FreeSeas Inc., which closed yesterday at $0.0217, and on Jan. 4 at $0.8925. It closed on Jan. 2, 2013, at $84,375. There were some reverse stock splits, as you might expect (including one this year), and these numbers are split-adjusted. Excel tells me that's down 99.9989 percent. Its current market capitalization is about $61,000.

  8. There are even blockchains!

This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.

To contact the author of this story:
Matt Levine at mlevine51@bloomberg.net

To contact the editor responsible for this story:
James Greiff at jgreiff@bloomberg.net