Saturday, April 25, 2020

Section 11 in the Age of Direct Listings

All right, it’s probably a little premature to call it an “age of direct listings”; we’ve had Spotify and Slack and I guess some company called Watford Holdings and Airbnb was reportedly considering one, and in the Before Times a lot of VCs were making noises about preferring direct listings to traditional IPOs, but in the immediate future I don’t expect to see a lot of new firms going public either way (notwithstanding the occasional ambitious SPAC), so these issues may turn out to be nothing more than a curiosity. 

BUT!  In the meantime!  We have Pirani v. Slack Technologies, 2020 WL 1929241 (N.D. Cal. Apr. 21, 2020), in which a district court refused to dismiss Section 11 claims brought by investors who purchased Slack shares after the company listed directly on the NYSE. And, it turns out, direct listings raise a lot of unsettled questions under Section 11.

 Slack, like a lot of companies these days, never formally sold its stock to the public; instead, it distributed stock in exempt transactions, subject to various securities law rules that permit these kinds of distributions but generally require the investors to hold their stock for some period of time before reselling it.  Eventually, Slack had distributed a lot of stock this way, and its investors were clamoring for liquidity, i.e., an easy way to sell their shares.  So Slack decided to list its shares for trading on the New York Stock Exchange.

That was fine for shares that had been distributed so long ago that investors were now legally permitted to resell them, but a lot of shares were distributed more recently, and investors holding those shares were were not permitted to trade them yet.  In order to allow these latecomers to sell immediately, Slack registered those shares – and only those shares – formally with the SEC. Normally, companies file registration statements when they sell new stock to the public, and the registration permits immediate trading.  But registration statements can be used – as Slack’s was – to register previously-issued shares, so as to enable holders of those shares to trade right away.  As a result, when Slack’s stock formally began trading on the NYSE, there were actually two groups of shares for sale: The shares that had been distributed so long ago that the legal holding period had expired, and the recently-issued shares that had just been registered.  In total, 118,429,640 shares were registered, and an additional 164,932,646 shares traded with them.  But, of course, it was all common stock, and therefore fungible; no specific share could be traced to one group or the other. 

Which is how we get to Section 11.

Section 11 of the Securities Act of 1933 allows any investor to sue if they purchased a security issued pursuant to a false registration statement. Damages are calculated based on the extent to which the price of the security falls below “the price at which the security was offered to the public.”

Naturally, soon after Slack’s direct listing, some investors came to believe that the registration statement contained false information (specifically, concerning Slack’s ability to avoid service disruptions), and sought to bring Section 11 claims against the company.  In the context of a direct listing like Slack’s, then, two questions were raised:  First, what is the “price at which the security was offered to the public” if the issuer is not selling any new shares to the public?  And second, how can a purchaser show that he or she bought shares tied to the registration statement?

These were the questions that confronted the court.

And first, let’s first talk about theory.

In general, the idea behind Section 11 is that when a company is selling securities to the public, it is functionally warranting that the securities are worth at least the price at which they’re being sold for, based on the information in the registration statement.  If the registration statement turns out to be false and the securities drop below that offering price, disappointed investors are entitled to recoup the amount of their overpayment.  In this scheme, the issuer is held liable to the extent it collected payments in excess of the securities’ value, as gauged by the registration statement information.

That said, the rules are not quite so simple.  First, many persons other than the issuer may be liable for false registration statements (though these persons, like corporate directors, are generally those who are in a position to correct false registration statements).  And second, registration statements are explicitly required even when persons other than the issuer are selling shares – like, control persons.  In the end, then, Section 11 remedies are analogous to disgorgement but they are not quite so strict because even persons who do not collect monies from the sale of shares may end up paying damages. 

Still, most of the time, the scheme is relatively coherent – an issuer sells securities for more than they are worth by including false information in the registration statement, therefore the issuer (and persons similarly responsible) must refund disappointed investors the amount of the overpayment.

So what to do about the Slack situation, then?

As to the first question – what counts as the “price at which the security was offered to the public” – Slack’s argument was of course that there was no such price, and therefore there could be no Section 11 damages.  The court, however, ducked the issue, holding that damages are an affirmative defense and not an element of the plaintiffs’ claim, and therefore are an inappropriate basis for dismissal.  At the same time, the court expressed some doubt as to the merit of Slack’s argument that a direct listing is entirely different from an IPO, pointing out that Slack itself admitted in its registration statement that the direct listing process would include a “pre-opening indication” that was “[s]imilar to how a security being offered in an underwritten initial public offering would open on the first day of trading.” 

In my view, Section 11 is, at least for now, one important mechanism for enforcing the securities laws.  If those laws contemplate that previously-issued shares will become freely-tradeable upon registration, then, it’s necessary that courts be able to identify some “offering price” that renders Section 11 meaningful.  Slack should not be able to avoid liability simply by claiming there is no relevant price; something must count.  

Which leaves us only to ask how that price should be determined.

As a formal matter, there are a few options.  One is the “reference price” chosen by the NYSE to guide trading; another is the “opening price” determined by the market maker based on pre-opening buy and sell orders.  Those buy and sell orders may use the reference price as a guide, but trades themselves are executed, to begin with, at the opening price set by supply and demand.  In Slack’s case, the reference price was $26, but the opening price was $38.50. 

This is sort of analogous to the IPO process.  In a traditional IPO, the offering price is used to sell shares to initial allocants, but pre-opening bidding determines the actual opening price at which trading begins.  For example, in Facebook’s IPO, its offering price was $38 per share, but due to pre-opening bidding, the stock actually began trading at $42.05 per share.  For Section 11 purposes, $38 was the relevant price.  In this analogy, the reference price in a direct listing is like a traditional IPO’s offering price and therefore the reference price should be the anchor for Section 11 claims. 

But!  Unlike a traditional IPO, no shares actually change hands at the reference price because there are no initial allocants.  If we assume that the majority of Slack’s pre-listing shares were issued in sales that count as “private” under the securities laws, the first actual sales to the public were, in fact, at the opening price, and therefore $38.50 was the “price at which the security was offered to the public” for Section 11 purposes.  Therefore, the opening price should control.

But on a third hand, we might look at the pre-open bidding history and see the very first price at which any seller offered to sell shares – likely NYSE’s reference, and possibly something else.  That first price might technically be the first “price at which the security was offered to the public,” and therefore we might treat that as the appropriate anchor for Section 11 purposes.

All of that, though, is divorced from the functional purposes of Section 11 damages, namely, to force the issuer to “stand behind” a particular stated value for the security.  Which is why I think this matter actually cannot be settled without discovery.  Slack claimed in its briefing that it had no role in setting either the reference price or the opening price, but that likely overstates matters. Its bankers appear to have assisted both the NYSE and the designated market maker in setting both prices, and it may turn out that either price was driven by the preferences of selling insiders (the plaintiffs noted that several insiders sold at prices slightly above the open).  If so, these participants’ involvement, along with their superior knowledge, may function as the same kind of “price warranty” that Section 11 ordinarily envisions, perhaps similar to when an issuer registers shares being sold by a controlling person (who necessarily must have a hand in setting the price to the public).

Which brings us to the next question, namely, how do we know that any open-market purchaser bought registered shares rather than unregistered ones?

This is actually not a problem unique to direct listings; it arises even in IPOs, when previously-issued shares become freely tradeable along with registered shares.  In the typical IPO, some number of insiders hold unregistered shares at the time of the offering, but they are subject to a contractual lockup period that prevents them from trading (usually 180 days) after the registration statement becomes effective.  After the lockup period expires, the unregistered shares commingle with the registered shares.  Courts usually hold that a Section 11 plaintiff is required to prove that his or her shares are traceable to the registration statement, Krim v., 402 F.3d 489 (5th Cir. 2005), which is literally impossible to do for any open-market purchaser who buys after the expiration of the lockup period.  As a result, the tracing requirement ends up cutting many a Section 11 claim off at the knees.  To avoid neutering Section 11 entirely, advocates have argued that courts should award damages based on some kind of statistical formula, such as assuming that plaintiffs hold registered and unregistered shares in proportion to the number of registered and unregistered shares in the market generally, see Steinberg & Kirby, The Assault on Section 11 of the Securities Act: A Study in Judicial Activism, 63 Rutgers L. Rev. 1 (2010), but courts so far have not been receptive.  Indeed, on more than one occasion, corporate attorneys have recommended that issuers intentionally shorten or eliminate the lockup period for some unregistered shares in order to defeat all Section 11 claims.

But now there’s Slack!

In Slack’s case, the court was sympathetic to the argument that because there is no lockup period in a direct listing – all shares become freely tradeable at once – a strict tracing requirement would mean no Section 11 claims are available to anyone.  As a result, the court held that “in this unique circumstance—a direct listing in which shares registered under the Securities Act become available on the first day simultaneously with shares exempted from registration,” purchasers of all shares of the registered class would be eligible to bring Section 11 claims.

On the one hand, I am (again) sympathetic to the idea that Section 11 should not be so easily defeated.  On the other hand, again, Section 11 was originally intended as a mechanism of disgorging ill-gotten gains from issuers, which is why there is so much fixation on offering prices and shares traceable to a registration statement.  In a direct listing, the concerns are a bit different – there’s no disgorgement to be had – and there’s no precise functional way to figure out how the statute should apply.

That said, if we look back to the history of Section 11, we see that the expectation was something like a fraud-on-the-market theory, where trading prices for newly-registered shares were assumed to be set based on information in the registration statement, and open-market purchasers would therefore be injured if that information turned out to be false.  See, e.g., Douglas & Bates, The Federal Securities Act of 1933, 43 Yale L.J. 171 (1933) (“At that time the registration statement will be an important conditioner of the market. Plaintiff may be wholly ignorant of anything in the statement. But if he buys in the open market at the time he may be as much affected by the concealed untruths or the omissions as if he had read and understood the registration statement.”); 78 Cong. Rec. 10186 (1934) (“the market value is fixed by the false statement of the corporation. The individual investor relies upon the investigation made by the banker. It is fair to assume that this situation continues until such time as the corporation makes available a statement showing its earnings for 12 months. Then the market value is influenced by the statement of actual earnings and not by the statements contained in the registration statement”).

On that theory, it hardly matters whether the shares were technically those issued on the registration statement or not – all shares would be affected by the same false information, and thus all purchasers should have some remedy, even if the total damages are capped at the level traceable to the number of shares issued on the defective registration statement, and divided pro rata among all claimants.

My final thought is: Section 11 may serve an important disciplining function, not only for IPOs, but even for shelf registrations that incorporate SEC filings by reference.  But, as I previously blogged, the statute has not been updated since 1933 and is not well-adapted to the realities of today’s offerings.  The best solution isn't to have courts try to mold the statute to the current environment, but for Congress to work out a solution that best meets the needs of modern markets.

Ann Lipton | Permalink


Thanks, Ann, for posting information about, and your reactions to, this interesting case. I have worked on registered secondary distributions. As a result, I saw the tracing issue immediately as I began to read the post. The pricing issue is inherent in secondary registrations, although the facts in this case make things especially difficult.

I have a number of thoughts. I will share a few here. Keep in mind that I am not a litigator as you read all this. I likely am missing something along the way . . . .

First, on public offering pricing, you say: “an issuer sells securities for more than they are worth by including false information in the registration statement, therefore the issuer (and persons similarly responsible) must refund disappointed investors the amount of the overpayment.” Well, not always (unless I am misinterpreting your words). Under Section 11(e), the damages owed by an issuer are calculated based on “the amount paid for the security (not exceeding the price at which the security was offered to the public).” So the issuer may not be responsible for paying a successful plaintiff the entire spread based on the offering price.

Second, also on pricing, the issuer must calculate, under Rule 457, the maximum aggregate offering price for the securities being registered in order to calculate the registration fee. I am not sure how that was calculated for Slack since there is no referential market price and no underwriter-fixed price. I do not have time to look that up right now. (I am grading and writing two exams.) But it may be interesting to see what price emerges if one divides that maximum aggregate offering price by the number of shares registered . . . . It’s likely not true to any market price reference if the stock did not trade before the offering. Thus, it would not fit with your Section 11 policy story. But I mention it here in case it’s useful.

Third, on tracing, I follow your argument entirely. But I am perhaps more wary than you are of expanding liability to aftermarket purchasers. Since Section 11 is functionally strict liability, I would extend liability out no further than those purchasing during the applicable period during which prospectus delivery is required under the 1933 Act. Apparently, based on what you have written, courts do not agree with me on that.

Fourth, I offer a tracing thought experiment that may lead to nowhere . . . . If, as you posit, Section 11 liability is extended to aftermarket purchasers on a fraud-on-the-market-like theory, then all information in the market at the time the investor purchased the securities should be taken into account in determining liability (even though that is not the Section 11 “way”). So, maybe that’s where Section 11 liability logically cuts off with respect to specific misstatements—when accurate and complete information is public and fully disseminated. That may make it hard to look at tracing as a standing requirement. It may then become a matter for trial. So, it does not help with pleading or other pretrial issues in which tracing may arise.

When I teach Section 11 in class, as I recently have done, I always note that it is significantly more complete and directive than other misstatement/liability provisions in the 1933 and 1934 Act. Of course, that is true. But much interpretation still must be done, as your post makes abundantly clear. Thanks, again, for a great post.

Posted by: joanheminway | Apr 25, 2020 11:56:32 AM

Hi, Joan - I'm so glad the post interested you!

Yes, you're right the issuer may not pay the entire spread - the plaintiff might buy below the offering price and maybe the issuer can prove negative causation -but the theory, I think, remains that the scheme is akin to a warranty of the correctness of the offering price based on the information in the registration statement. After all if the plaintiff buys below the offering, then someone sold below the offering - which means someone else also has a claim unless the issuer can prove negative causation.

As for Slack's fee- great thought! I checked - it said there was no aggregate offering price and so it was calculating the fee based on the book value of the shares ( So that's another option, I guess, though I don't know if it's the best from a policy perspective.

As for tracing and market information, right now, courts tend to do a kind of market analysis - they’ll use market information to determine if the misstatements were material and if losses were due to the false statements. The most famous Truth on the Market case was Wielgos, which was Section 11. And of course after a 12 months earning statement the plaintiff has to prove reliance – which is, again, because the enacting Congress thought by then the registration statement would no longer be influencing the price. So I think it's a reasonable way to think about it. That makes it not so much a tracing issue but a causation issue - if the plaintiff buys after a market correction, then there's no cause of action because of lack of causation for losses.

Posted by: Ann Lipton | Apr 25, 2020 12:16:08 PM

This all is helpful. Many thanks for your responses. I want to think about this more. But I appreciate you getting me off to a good start!

Posted by: joanheminway | Apr 25, 2020 2:17:13 PM

Post a comment