And then there were… six. Over the past three years, we’ve had exactly six direct listings since Spotify pioneered their use in its current form by technology companies. It led to many questions around the what, the why, and the how of direct listings — as well as the differences and tradeoffs between all the different routes to public markets (which I’ve written about previously, including defending the traditional route of the IPO and how to improve it).

While that half-dozen direct listings barely represents 1% of the total capital markets debuts over the same time period, it includes three of the top five opening trade volumes of all time in the history of going public. The members of this direct listings “club” are also all high-profile tech companies, with products used by many consumers today.

So even though it’s a very small sample set, many people — from market watchers to bankers to policymakers to builders trying to figure out which route is best for them — have been watching each of them to try to understand, are direct listings working? What do the data tell us (albeit from a very limited sample)? And what have we learned so far when it comes to direct listings?

Before I share some data via Morgan Stanley and others, and some of my observations here, let’s first quickly summarize what a direct listing is.

A brief overview of direct listings

We’ve previously shared a very detailed primer on direct listings versus traditional IPOs, but just as quick context here: in a traditional IPO, the company going public is generally offering “primary” shares, newly issued by the company for the purpose of selling them to public investors. The implication is that the company gets to keep the money from selling those shares they issued. In a direct listing, however, the company does not sell primary shares; only secondary shares trade hands. (The SEC only recently published rules permitting the sale of primary shares in a direct listing, and none of the listings to date have utilized this.)

In the direct listings so far, the companies did not receive any cash from the secondary sales; the money went directly to the selling shareholders. Why would a company do a direct listing then? Well, if it has ample cash on its balance sheet to be able to run its business, it may not want to “dilute” its shareholders by selling more primary shares. But the shares for a direct listing need to come from somewhere, since the goal is to enable new public market investors to own the company’s stock; as such, the secondary shares come from existing investors directly to the public market investors.

So how is this market created, and what might it tell us so far about direct listings as an option for companies?

In a direct listing, there’s an opening auction by which the exchanges (e.g., NYSE, Nasdaq) or their designees (“market makers”) solicit “bids” (the price and volume at which a buyer wants to transact) and “offers” (the price and volume at which a seller wants to transact). This, by the way, is why direct listings don’t start trading with the 9:30 AM opening stock market bell, but typically open several hours later. (Some people even judge their success on how long that takes, but there are several factors at play here.) In any case, the auction occurs until a sufficient number of orders are “matched” — buyers and sellers are willing to transact at an agreed-upon price and volume to open the stock. But why not just open the stock at 9:30 AM if there is a single buyer and seller who agree to an opening price? The reason is that the exchanges are trying to create an orderly market — they don’t want ten shares to trade at, say, $50 per share to open, and then have millions of other shares trade at highly variant prices later throughout the day. It would be an inefficient method of “price discovery” — the process by which the market determines a reasonable price at which to trade the stock — and could create a whipsaw, where the stock price will suddenly change direction. It’s too volatile for all involved.

So, how’s it going?

With the small sample size of six direct listings, watchers have been wondering if the “matching” described above actually led to a good number of shares being traded. Investment bankers involved in the listings aimed to have 3-5% of the shares trading in the opening auctions, and that’s roughly borne out so far — the percentage of the total shares outstanding that have traded in the opening auctions range from a low of 2% to a high of 8%. In contrast, traditional IPOs in the tech space average around 2% trading volume on the opening trade.

This may be one of the reasons why the after-market “pops” — a phenomenon I’ve written much about, including more broadly what they do and don’t tell us about companies — are more muted in direct listings than we’ve seen recently for more traditional tech IPOs. The higher volume of opening trades in direct listings indeed leads to better price discovery, and fewer wild swings as the market adjusts. (For more on this, see my earlier pieces with Alex Rampell explaining the dynamics behind IPO pops; and the impact of supply/demand on initial trading of IPOs, given the perception of 2020 pops seeming larger than before.)

After the opening auction, the stock from a direct listing then trades like any other public stock; investors can buy or sell throughout the day. For the direct listings to date, the stocks have traded the first day in the range of 15-23% of the total shares outstanding, a first-day volume of about 3-5x the number of shares exchanged in the opening auction. Note, not all of those shares are additional secondary shares that come into the market — much of that first-day trading volume reflects the re-trading of shares that were initially exchanged in the opening price auction.

While we don’t have perfect data, we estimate that about one-third of the total shares that trade on opening day of a direct listing are newly introduced into the market, whereas two-thirds of the trading represents re-trading of those shares. (This does beg a question of why can’t the secondary markets accomplish the goals of a direct listing instead, since they give early investors and employees liquidity, with no additional dilution for ongoing investors. The answer our corporate development team shared previously is that secondary exchanges don’t offer true market price discovery as direct listings do: They are more of a marketplace, matching one buyer to one seller, than they are a true public market with robust supply and demand, matching many buyers and many sellers.)

Beyond re-trading, how do the shares outstanding in direct listings stack up against IPOs?

Traditional IPOs in the tech space have traded closer to 12% of total shares outstanding on the first day of trading — well below what we have seen in direct listings. Why is that? As noted above, the direct listings have done a better job of getting initial share volume into the market (~5% on the opening trade vs. 2% in traditional IPOs). Additionally, by better matching supply and demand, direct listings mitigated the magnitude of IPO pops, thus engendering better overall price discovery.

Note too that traditional IPOs, by definition, have a fixed initial entry of shares into the market via the sales from underwriters to institutional investors. Most IPOs also have lockups that prevent additional secondary shares coming into the market, sometimes for as long as six-months (although some recent IPOs have shortened this period for employees, or based on price targets for the stock). Since direct listings generally don’t have lockups (Palantir being the exception), they enable more secondary shares to be offered in the market, thus reducing price movements caused by supply/demand imbalances in a traditional IPO. Volume is key to finding the true market price in a direct listing.

Let’s now take a closer look at price discovery, as measured by reducing “intra-day” pricing volatility — what is the difference between the high and low trading price during the first day? For the six direct listings, intra-day price volatility on the first day has ranged from 9%-38%. The average volatility of similarly sized $500+ million tech IPOs over the last 10 years, meanwhile, is 29%. Intra-day volatility for tech IPOs has also been trending higher in recent years, with IPO pops on the order of 50-100% not uncommon.

Direct listings are still few and far between, representing far less than 1% of the total new issue market. But, while the data is still early, it does appear that the combination of a more formal initial auction process — coupled with larger amounts of shares being supplied to the market — is having the desired impact of improving price discovery, and thus reducing day-one price volatility.

*   *   *

So, where do we go from here? A few thoughts:

First, direct listings are still effectively limited by the secondary nature of the offerings, even though the SEC has permitted primary offerings in connection with a direct listing. No companies have chosen this route so far, and, given the backlog the SEC is facing due to the large volume of SPACs (another alternative route to public markets) it is reviewing, companies that desire to raise primary capital nonetheless may be loathe to try a bespoke primary direct listing: It could take longer in SEC review, which creates more market-timing risk.

Second, the after-market trading performance of direct listings (specifically, lesser intra-day volatility) is likely to continue pressuring companies to look at reforms to the traditional six-month lockup that accompanies most traditional IPOs. These lockups are not required by law, and we’ve already started to see some changes here: shorter lockups for employees, stock-price-based triggers of staggered early lock-up releases for non-employee investors. I expect this to continue to evolve. The data is clear that reducing artificial constraints to supply can dampen price volatility.

Third, the same after-market trading trend will continue to drive innovation on the opening-trade price-discovery process for traditional IPOs. Unity was a pioneer in this area recently, by enabling a modified auction to set its IPO price and taking a more formal role in the pricing committee discussions… but there is more to do here.

The early returns are in, and while direct listings don’t solve all the problems associated with the current IPO process, they do represent a meaningful step forward. They are an option for companies that should become increasingly viable in the future.

This is not just relevant to VCs and startups but to innovation and access as well: Publicly traded companies drive job growth, expand economic opportunity, and give more retail investors (including those planning retirement) the ability to invest in future growth. We still face a situation today where too few companies go public and, when they do so, they go public at much later stages of maturation, thus depriving more people of that growth.

Overall, it’s good news for capital formation that there are now multiple choices for companies that are considering going public, including the direct listing.

 

Acknowledgements: Thank you to the equity capital markets team at Morgan Stanley for providing data and their perspectives on the direct listings to date.

  • Scott Kupor

    Scott Kupor is a managing and investing partner at a16z, responsible for all operational aspects of the firm, and makes growth stage investments. Prior, he worked for HP, joining via the Opsware acquisition.

Join the Newsletter

Technology, innovation, and the future, as told by those building it.

Thanks for signing up.

Check your inbox for a welcome note.