Before the early 2000s, the tradeoff between venture and public markets was that venture got in earlier, and at more favorable valuations, but had to hold their position all the way through either IPO-plus-lockup or until an acquisition happened. That "have to hold" criterion was at both the level of positions held within a given fund, and the ownership stakes of the limited partners in that fund.

When venture was an obscure asset class with typical returns in the 25-30% range, i.e. through the late 80s, this was fine. When companies were going public five years after founding, it meant that the typical dollar invested in venture converted into something liquid in perhaps 2-3 years (since most of the funding would be later in a company's existence). By the 2000s vintage, both parts of this deal broke down: time from founding to IPO stretched out to nearly a decade, and returns dropped, too. This also made equity a worse deal for employees—they got the part of startup life where you work hard for less cash comp than your big company peers, but the part where a cash windfall shows up happened much later.¹

This led to the creation of early private company secondary markets—and the existence of those markets also meant that there was less immediate pressure to go public. There were apparently some scattered attempts to engineer this earlier, but markets need a specific confluence of factors before they can come into existence, and that confluence didn't really happen until the late 2000s and early 2010s, when Facebook was both a large-cap stock with a wide shareholder base and, technically, a private company.²

It's possible that this came about due to one of the few areas of social media that Facebook, now Meta, hasn't made many inroads: LinkedIn had a pretty comprehensive list of employees at big tech companies by this time, and conveniently displayed start and end dates. So someone who was hustling for Facebook shares could, without too much effort, find someone who'd joined early and left after they were fully vested. The growth-focused funds that had made money picking up the pieces of late-dot com detritus to find winners—some of the last dot-coms to go public, after the Nasdaq peaked, were a group of very premature Chinese Internet IPOs priced in the spring and summer of 2000: Sina, Sohu, and Netease. A few years later, with their market caps much smaller but their businesses much more real, these companies contributed to Tiger Global's early winning streak. But a few years later, there just wasn't a great inventory of very young companies that were available to public market investors, while there were plenty of private ones.

Facebook was a dominant part of this market, but it ended up creating a Schelling point. Once the concept of buying these is established, and once the path to sourcing shares and demand for them is in place, and once the legal documents have turned from new Word docs to ready-to-modify templates, onboarding another company is easy. Which is not to say that the business wasn't affected by the Facebook IPO; that seems to have been a local peak for quite a while.

But it set up financial infrastructure that led to a few changes:

1. Companies got more judicious about exercising their right of first refusal to control who ended up on their cap table, and this became part of the social contract for employee equity compensation. (At least, if you ask the issuers; the median new employee at an early company probably doesn't spend a lot of time contemplating the legal mechanics of realizing capital gains before an exit.)

2. Investors also got more judicious, and had realizations like: if every early investment happens through its own LLC, you now have a legal entity whose ownership might be subject to fewer transfer restrictions.

3. As the population of very late-stage, but very not-public-yet companies grew, some of these companies started providing liquidity by buying back shares from employees to sell to outside investors. This is a more controlled process than unfettered secondary trading, and it doesn't have the price discovery of public markets because no one has quite managed to create a private company marketplace where shorting works. But it's doable.

Each of those points touches on questions about legal structure, and those questions have gotten trickier over time. Companies like to exercise some control over who owns them, and they have very little loyalty to either former employees who left and are trying to get around transfer restrictions, or unapproved investors. So a secondary deal is typically not like a normal transaction, in that there's a period where the buyer and seller have committed but a third party may intervene. (Palantir, for example, was sued over trying to block some secondary sales as a private company.)

What we've wound up with is a continuum, where companies go public a bit at a time, and where institutional tech investors portfolios are smeared between assets with instantaneous low-cost liquidity, things that can be sold over the course of weeks or months with some haircuts and uncertainty, and assets that are early enough, or locked-down enough, that their value as measured in terms of ready conversion into cash is basically zero. Which actually makes a lot more sense than the public/private binary: some companies really are, and should be, more public than others but not as public as a company that, once a quarter, tells everyone in the world exactly how much they made and then gets on a conference call with analysts in order to tell the whole world how they did it. That has benefits, but it also has strategic costs in addition to the financial ones. If there's an option that gives companies most of the upside but little of the downside, they'll take it.

Disclosure: Long META.

The public/private distinction, and its nuances, have come up in many Diff posts over the years:

• Private markets have less fraud ($), in part because the transaction costs are high enough that a little due diligence is worth it.

• Turning a financial product into something for institutions and rich people into something for everyone is tricky, but inevitable.

• The biggest tech companies tend to stay undervalued, in both public and private markets.

• Within startups, there's a range of financial structures that work ($).

• Many debates surrounding the private/public dynamic, the rise of secondaries, and other fundamental puzzles in capital markets are a function of asset/liability duration mismatches, the traditional private equity industry has found the missing piece for theirs ($).

Join the discussion!

1  This also created a class of very nervous new rich; eight-figure net worth, but all of it tied up in a company whose gory details they can't help but be keenly aware of. Motivating! But terrifying.

2  "Technically" not just because employees were able to sell some of their stock, but because in early 2011 the company raised a round with Goldman, where Goldman offered the shares to its high net worth customers.