Wall Street

Appeals Court Concludes That IPOs Are Legal

Banks are allowed to make money when initial stock sales go badly.

Of course, back at the time of the IPO, Zuckerberg would never wear a suit.

Photographer: Guillermo Gutierrez/Bloomberg

Here’s a thing that happens. When a company goes public, it sells some shares to the underwriters, and the underwriters sell more shares to the public. Let’s say the company sells 100 shares, for $10 each. The underwriters will sell 115 shares, for $10 each, and give the company $1,000 (less fees). The underwriters will keep the other $150 for themselves. They have bought 100 shares and sold 115, so they are short 15 shares. If the stock price goes down over the next few days, they use some or all of that $150 to buy the stock, so it doesn’t go down too fast. 1 This is called “stabilization.” The underwriters’ buying helps keep the stock price close to the IPO price. People think that is good.

On the other hand, if the stock goes up after the IPO -- which is more common -- the underwriters don’t buy back those 15 shares in the market. (That would be expensive.) Instead, they give the company the $150 they kept (less fees), and get back 15 more shares at the IPO price of $10 each, which they use to close out their short position. 2 This is called “exercising the greenshoe,” for historico-silly reasons. 3 The reason that the company sells the underwriters 15 more shares at $10, even though the stock is now trading at $12 or whatever, is that the underwriting agreement for the IPO included an “overallotment option,” or “greenshoe,” obliging the company to sell the banks 15 more shares, any time in the (usually) 30 days after the offering, at the deal price, at the banks’ option.

The point of the greenshoe is to reduce the banks’ risk on stabilizing the deal. The underwriters buy 100 shares and sell 115, but they aren’t really short those 15 shares: The greenshoe option lets them buy them back at the deal price. So if the stock price goes up -- as, again, it usually does -- they don’t lose any money. 4 If the stock price stays flat, and the underwriters buy back more shares at $10 to prevent it from going lower, they also don’t lose any money. (They sell at $10 and buy at $10.) But if the stock goes down a lot before the underwriters can stabilize it, then they make money: They sell those 15 shares at $10, and buy them back at $9 or $8 or whatever.

Occasionally you will find people who think all of this is very suspicious, because it kind of looks like the banks make more money from bad IPOs. If you underwrite a deal at $10, and the stock immediately falls to $5, then you make a lot of money buying in your short position. That seems like a bad set of incentives. 5 On the other hand, the banks also have lots of incentives the other way, and in practice you more often find people complaining about underpriced IPOs that go way up on their first day of trading.

But Facebook Inc.’s public offering in 2012 is one notable and controversial case of a big IPO where the underwriters made a lot of money because the price went down. Facebook’s underwriters bought 421 million shares from Facebook and sold 484 million shares to the market at $38 each, leaving them short about 63 million shares worth about $2.4 billion. The stock was up slightly on the first day of trading, a Friday. The next Monday -- May 21, 2012 -- the stock cracked, closing at $34.03. 6  The underwriters never exercised the greenshoe, and seem to have spent the bulk of their $2.4 billion buying stock on Friday and/or Monday. If they bought all of that stock at $38-ish on Friday, then they made no profit (and succeeded in stabilizing the stock for one day). If they bought all of it at $34-ish on Monday, then they made about $250 million (and didn’t stabilize too much). Or somewhere in the middle, I don’t know, but one estimate is that they “made a profit of about $100 million with the bulk of that profit” coming on May 21. 

That estimate is from a lawsuit filed against the underwriters, because of course it is. Earlier this month, the U.S. Court of Appeals for the Second Circuit dismissed the suit, but it was a fascinating effort. 7  The plaintiffs’ theory was based on Section 16(b) of the Securities Exchange Act of 1934, which says that if you own more than 10 percent of a company’s stock, and you do any short-term trading in the stock -- buying and selling shares within any six-month period -- then you have to give up all of your profits to the company. (Or, in practice, to the plaintiffs’ lawyers who sue you on behalf of the company’s shareholders.) The theory is that by selling those 63 million shares on Thursday, and then buying them back on Friday and Monday, Facebook's underwriters made a “short-swing profit” and should have to give it back. 

One problem with that theory is that the underwriters didn’t own 10 percent of Facebook, so Section 16 doesn’t apply to them. But the plaintiffs have an answer to that. Section 16 applies not just to individual 10 percent owners, but also to any “group” of shareholders who “agree to act together for the purpose of acquiring, holding, voting or disposing of equity securities.” 8  And the underwriters did have an agreement with some shareholders to act together with respect to Facebook’s stock. Specifically, the shareholders who were selling stock in the IPO -- including Mark Zuckerberg, Facebook’s chief executive officer, who himself owned about 30 percent of Facebook immediately after the IPO 9 -- signed agreements with the underwriters promising that they wouldn’t sell any more stock for 91 to 211 days after the offering without the underwriters’ permission. 10  These lock-up agreements, like the greenshoe, are meant to stabilize the stock: Investors are more likely to buy stock in the IPO if they know that the big shareholders aren't going to dump a lot more stock next week.

But, sure: There was a group of shareholders, including the underwriters, that owned more than 10 percent of the stock, and that had “agreed to act together” in “holding, voting or disposing of” that stock. And some members of that group -- the underwriters -- sold stock on Thursday, and bought it back on Friday and Monday, and made a big profit. So if you read everything literally, they should have to give that profit back to Facebook. Or its shareholders. Or its shareholders’ lawyers.

What a neat theory! It is absurd, and the trial court and the appellate court both dismissed it. But you can’t tell that it’s absurd just by reading the law. 11 To know that it’s absurd, you have to look beyond the actual words of the law -- which seem to support the plaintiffs -- and just, like, understand how IPOs work. Every IPO -- or almost -- has a greenshoe. Every IPO -- or almost -- has lock-up agreements. If you could just combine those two elements in a way that makes them illegal, then every IPO is illegal. The law can’t mean that.

And so the Securities and Exchange Commission weighed in to say that the law doesn’t mean that. The court says (citations omitted):

We cannot avoid a larger, legitimate concern emphasized in the SEC’s amicus brief over applying Section 13(d) literally in the context of standard lock-up agreements. As the brief notes, a lock-up agreement is common, even essential, to the typical IPO, and some other public offerings as well. Such an agreement assures potential buyers of securities in the IPO “that shares owned [by pre-IPO shareholders of the issuer will not] enter the public market too soon after the offering.” These assurances lead investors reasonably to expect an orderly market free of the danger of large sales of pre-owned shares depressing the share price before the pricing of the newly offered shares has settled in the market.

Applying Section 16(b) to underwriters engaged in lock-up agreements as facilitators of a public offering would impair the market for public offerings by complicating the role of underwriters –- adding tens of millions of dollars in legal exposure to the underwriters’ costs. 

The rules can’t mean what they seem to say, because that would mess up how IPOs are done:

Far from being nefarious, these actions benefit existing shareholders and new public investors. For example, one purpose of the regulation of public offerings is to enhance relatively accurate pricing of the offering’s shares by disclosure before sales of an offering to the public are allowed. Achieving that purpose requires assurances of control over the disposition of blocs of shares owned by large pre-IPO investors, and lock-up agreements provide that control. (One effect of a lock-up agreement in an IPO is to prevent pre-IPO insiders from using nonpublic information to trade in a nascent public market.) The purpose also requires stabilization efforts by underwriters, as discussed above. Lock-up agreements are, therefore, essential to the regulation of public offerings.

Now, it is possible to quibble here. I mean, sure, let’s say lock-ups and greenshoes and stabilization are essential to IPOs. But: Why not make the underwriters give back their stabilization trading profits to the issuer? (That’s sort of what this case was about, though really it was about giving those profits to the shareholders and their lawyers.) After all, the underwriters can make those profits with very little risk, precisely because they got a free option from the issuer. 12 The underwriters sell stock short, but the greenshoe means that if the price goes up, they don’t lose any money: The company, in effect, covers their losses. If the price goes down, the underwriters make money: Why not give those profits to the company? I suppose you want the underwriters to do a good job trading the stock, and it’s no fun to trade when you have to give up your profits. 13  But it seems a little like we do things this way because this is the way we have always done things. 

There’s a lot of that generally in the IPO process. Everything I described above seems ... illegal? I mean, just the act of “stabilizing” the stock looks like market manipulation, in that the underwriters are buying the stock primarily with the purpose of keeping its price up. Everyone knows this, and it is well disclosed in the IPO prospectus, and there are SEC rules around how the underwriters can use their “stabilizing bid,” so it’s not too manipulative. 14  But it’s a little manipulative. It’s just legal manipulation. It’s allowed because otherwise IPO prices would be more volatile, and IPO investing would be riskier, and investors wouldn’t like that. Or that is the theory, though it’s not exactly subjected to frequent and rigorous empirical tests. Every IPO has a greenshoe, and a stabilization agent, and a lock-up, so you’d be crazy to try to do an IPO without one. The banks, the SEC and the courts all say so.

Most of the time, in stock trading, the market is the market: Collusion and coordination and manipulation are frowned upon, and cold impersonal forces of supply and demand rule. But initial public offerings are an exception, a protected place where prices are set by agreement and where tradition supports coordination between banks and companies to maintain the price. It’s all so standard that no one involved gives it much thought. But it’s also just a bit askew from the rest of the market, and if you do think about it too hard, it looks pretty weird. Every so often, an enterprising lawyer will notice.

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

  1. Schematically:

    1. They buy 100 shares from the company in the IPO for $10 (spending $1,000).
    2. They sell 115 shares in the IPO for $10 (receiving $1,150).
    3. They buy back 15 shares in the market for, say, $9 (spending $135 or whatever).

    Net, they own no shares, and have made $15. This ignores fees.

  2. Schematically:

    1. They buy 100 shares from the company in the IPO for $10 (spending $1,000).
    2. They sell 115 shares in the IPO for $10 (receiving $1,150).
    3. They buy 15 more shares from the company for $10 (spending $150).

    Net, they own no shares, and have made zero dollars. This, again, ignores fees.

  3. Capital markets lore is that the first overallotment option was done for a company called Green Shoe Manufacturing Co. Heres a good introduction to greenshoes.

  4. In fact they make money: The deal gets bigger, and they get their underwriting fees on the larger deal.

  5. Not only that! The greenshoe is normally limited to about 15 percent of the stock, but the underwriters short position is not so limited. And so on tough deals the underwriters might take a naked short. (Not to be confused with the other, more illegal, sort of “naked short.) So they buy 100 shares, sell 130, and have a greenshoe option for 15. If the deal cracks and the price goes down, they buy back all 30 shares, supporting the price. If the stock goes up, they can buy back 15 shares at the deal price with the greenshoe; they have to buy back the other 15 in the market, at their own risk.

    This doesnt happen all the time, but it is a standard piece of technology, and well disclosed. Here is Facebooks IPO prospectus:

    In order to facilitate our initial public offering of the Class A common stock, the underwriters may engage in transactions that stabilize, maintain or otherwise affect the price of the Class A common stock. Specifically, the underwriters may sell more shares than they are obligated to purchase under the underwriting agreement, creating a short position. A short sale is covered if the short position is no greater than the number of shares available for purchase by the underwriters under the over-allotment option. The underwriters can close out a covered short sale by exercising the over-allotment option or purchasing shares in the open market. In determining the source of shares to close out a covered short sale, the underwriters will consider, among other things, the open market price of shares compared to the price available under the over-allotment option. The underwriters may also sell shares in excess of the over-allotment option, creating a naked short position. The underwriters must close out any naked short position by purchasing shares in the open market. A naked short position is more likely to be created if the underwriters are concerned that there may be downward pressure on the price of the common stock in the open market after pricing that could adversely affect investors who purchase in our initial public offering.

  6. It got better. All of this stuff is short-term stuff; usually the greenshoe is exercised, or the stabilization is finished, within a few days after the IPO. The fact that Facebook is now way, way, way above its IPO price is irrelevant to how it was supposed to trade four and a half years ago.

  7. It is discussed in this Paul, Weiss, Rifkind, Wharton & Garrison LLP memo

    Oh, disclosures: The underwriters were led by Morgan Stanley & Co., but also included Goldman Sachs & Co., where I used to work. Also, the judge who wrote the opinion, Ralph Winter, was my securities law professor in law school.

  8. Section 16 applies to "every person who is directly or indirectly the beneficial owner of more than 10 percent" of the stock. Securities and Exchange Commission Rule 16a-1(a) defines "beneficial owner" to include "any person who is deemed a beneficial owner pursuant to section 13(d) of the Act and the rules thereunder." Section 13(d) says that "When two or more persons act as a partnership, limited partnership, syndicate, or other group for the purpose of acquiring, holding, or disposing of securities of an issuer, such syndicate or group shall be deemed a 'person' for the purposes of this subsection." Rule 13d-3 says:

    For the purposes of sections 13(d) and 13(g) of the Act a beneficial owner of a security includes any person who, directly or indirectly, through any contract, arrangement, understanding, relationship, or otherwise has or shares:

    (1) Voting power which includes the power to vote, or to direct the voting of, such security; and/or,

    (2) Investment power which includes the power to dispose, or to direct the disposition of, such security.

    And Rule 13d-5 says: "When two or more persons agree to act together for the purpose of acquiring, holding, voting or disposing of equity securities of an issuer, the group formed thereby shall be deemed to have acquired beneficial ownership."

  9. Actually Zuckerberg maybe doesn't count, since he owned only high-vote Class B shares, the underwriters sold only low-vote Class A shares, and Section 16 applies by class of security. But for instance Accel Partners sold stock in the offering, and owned 21.4 percent of the Class A stock afterwards, and signed a lock-up agreement. 

  10. As the IPO prospectus puts it:

    We, all of our directors and executive officers, and the selling stockholders have agreed that, without the prior written consent of Morgan Stanley & Co. LLC on behalf of the underwriters, we and they will not, during specified periods of time after the date of this prospectus: offer, pledge, sell, contract to sell, sell any option or contract to purchase, purchase any option or contract to sell, grant any option, right or warrant to purchase lend or otherwise transfer or dispose of, directly or indirectly, any shares of common stock or other securities convertible into or exercisable or exchangeable for common stock;

    Et cetera. There were different lock-up periods for different selling shareholders. 

  11. To be fair, the court tries. Judge Winter writes:

    A plain language argument suggests application of Section 13(d), but we have explicitly avoided holding that such an agreement, without more, forms a group under Section 13(d).


    Our reluctance to recognize the existence of a “group,” notwithstanding a contractual arrangement explicitly limiting the disposal of shares, reflects the fact that lock-up agreements, rather than being agreements “to act together,” are generally one-way streets keeping certain shareholders out of the IPO market for a specified period of time or without compliance with other restrictions, as discussed immediately below.

    That is, you don’t have to interpret the language of the rules to apply to lock-up agreements. Maybe they’re not agreements to act together. They’re just agreements to ... not ... act. It is not an overwhelmingly convincing textual analysis, but it doesn’t have to be.

  12. Every so often, when I underwrote securities, someone would ask me why the banks don’t pay for the greenshoe option. And I’d explain: Well, look, the greenshoe is for the company’s benefit. The purpose is to get the deal done, not to make a trading profit. The underwriters don’t trade the greenshoe like a real option; they use it to stabilize the deal, and usually don’t make a profit. This is true. But it does feel weird that the company just gives up a valuable option for free.

  13. Though profit might be the wrong incentive here. You want the underwriters to stabilize the stock, which they can do by buying when others are selling at right around the deal price. But if the price is trying to drop, the underwriters -- who are short -- can make more money by letting it drop, and then buying back at lower prices.

  14. Here are the rules. Most notably: “Stabilizing is prohibited except for the purpose of preventing or retarding a decline in the market price of a security.” For instance: “After the opening of quotations for the security in the principal market, stabilizing may be initiated in any market at a price no higher than the last independent transaction price for the security in the principal market if the security has traded in the principal market on the day stabilizing is initiated.” You can’t put in a stabilizing bid at a higher price than the last independent trade, because that looks like manipulation up.

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

To contact the editor responsible for this story:
Brooke Sample at bsample1@bloomberg.net

Before it's here, it's on the Bloomberg Terminal.