Wha ... ? Photographer: Jin Lee/Bloomberg
Wha ... ? Photographer: Jin Lee/Bloomberg

I was a capital markets banker for four years, and a corporate lawyer before that, and today I did a thing I've never done before, which is read an underwriting agreement. Nobody reads underwriting agreements! The point of an underwriting agreement is for junior lawyers to get practice using track changes in Word. Here is how an underwriting works:

  1. Company goes to Bank and says it wants to sell some stock.
  2. Bank agrees to find investors.
  3. Bank finds investors, introduces them to company, negotiates price, blah blah blah, a whole lot of stuff happens.
  4. Eventually, Company, Bank and investors agree on a deal: Investors will buy X shares at $Y a share.
  5. Then and only then, Company and Bank sign an underwriting agreement saying Bank will buy X shares at $0.93Y from Company.1
  6. Company delivers shares to Bank, and Bank pays Company.
  7. Bank delivers shares to investors, and investors pay Bank.

Steps 6 and 7 happen simultaneously, normally four days after step 5. So the underwriting agreement exists for three days. The underwriitng agreement that Vascular Biogenics Ltd. signed with Deutsche Bank and Wells Fargo is 45 pages long, and all it does is say that in four days Deutsche and Wells Fargo would wire VBL some money and VBL would give them some shares. So you can see why no one would read it. That's a lot of pages to say "we have a trade and it will settle in four days."2

That underwriting agreemeent was signed on July 30, 2014. The shares were supposed to be delivered on August 5. And now it's August 12. And the shares still haven't been delivered, and Deutsche and Wells Fargo still haven't paid for them. And on Friday, after VBL's shares had been trading for six days, VBL announced that the whole deal is off. Never mind. Here's the entire announcement:

VBL Therapeutics3 today announced that the underwriting agreement for its initial public offering has been terminated by the Company's underwriters, Deutsche Bank and Wells Fargo Securities, LLC, due to an unexpected situation in which a substantial existing U.S. shareholder did not fund payment for shares it previously agreed to purchase in the offering. As a result of the termination of the underwriting agreement, no shares will be issued pursuant to the registration statement for the offering that was declared effective by the Securities and Exchange Commission on July 30, 2014.

The termination notice is not related to the Company, its business or its prospects.

You can more or less figure out from the prospectus what happened. It says that "Certain of our existing shareholders, including certain 5% shareholders and affiliates of our directors, have agreed to purchase an aggregate of 2,916,666 ordinary shares in this offering at the initial public offering price," and names those investors. Since the whole deal was for 5.4 million shares, those 2.9 million shares sold to insiders made up more than half the deal -- and around 2.4 million of them were being sold to one insider, a fund that owned 24 percent of the pre-IPO stock and that's run by Jide Zeitlin, a director of VBL.4 Here's the Wall Street Journal:

"I have no comment, other than to say that we have extended ourselves and made every effort to support VBL," said Mr. Zeitlin, who has served on the board since 2008.

Not every effort, maybe? The deal priced at $12 a share on July 30, opened for trading at $11 the next day, and traded down to close that day at $10.25. It got as low as $9.40 on Friday, August 1, closing that day at $9.88. If you had agreed to pay $29 million to buy half of the offering, you were going into the weekend with a $5 million loss. But you hadn't put up any money yet! It would be pretty tempting to just forget about the whole thing.

On the other hand just calling up your underwriters to say "never mind," after the deal has priced, is ... well, it's something that never ever ever ever ever happens. That's probably why it took so long to pull the deal: The deal was supposed to close on Tuesday, August 5, but stretched out to Friday the 8th before it was cancelled. Presumably the underwriters spent the early part of that week in a dead faint after getting that phone call.

And then they had to go read the underwriting agreement, which I'm sure was unpleasant for them.5 They were on the hook to close the deal -- and give VBL the money -- unless they could find an escape in the 45 pages of the underwriting agreement. Could they? Can you? I mean, here it is. It doesn't say anything about being able to be terminated "due to an unexpected situation in which a substantial existing U.S. shareholder did not fund payment for shares it previously agreed to purchase in the offering." Section 1 has 48 representations and warranties of VBL, all of which have to be accurate in order for the banks to have to close. They're pretty boring and boilerplate-y though, and don't say anything about the existing shareholders funding anything. Section 6 has 12 more conditions to the banks' obligations, including getting various certificates and opinions, and section 11 has some possible reasons for termination, but again getting money from investors is not a condition to the deal, and not getting the money is not a reason to terminate.6 That would kind of be cheating; the whole point of the underwriting is that the banks are on the hook even if the investors skip town.

So if Deutsche Bank and Wells Fargo just bailed on this deal because an investor didn't stump up the money, then VBL would seem to have a pretty good case to sue them.

On the other hand! Surely that's not what happened. When you do an IPO where an existing shareholder (and company director) is buying half the deal, that's not quite the same as doing an IPO where the banks are just selling shares to investors that the banks dig up. And when that shareholder decides that he'd rather not buy, his escape is not so much refusing to pay for his shares as it is trying to get the deal pulled. It's unclear whether the precise chain of events here was more like:

  • insider didn't want shares, demanded that deal be pulled, deal was pulled, or
  • insider didn't want shares, refused to pay for them, banks demanded that deal be pulled, deal was pulled;

but they come to roughly the same thing, no? VBL is not going to get too mad at the banks for skipping out on this deal. VBL, or at least a related party of VBL, seems to be largely to blame. Between VBL and its banks, you can see why forgetting the whole thing would also be appealing.

Of course that shouldn't stop anyone else from getting mad at the banks. Felix Salmon, for instance, is mad at them for charging 7 percent fees to underwrite the deal and then not sticking to their commitment. But this I think misunderstands what the fee pays for. You're not really paying the underwriter for committing to take three days of risk that investors won't close on the deal. That never happens! The value of that commitment is .... well, if you'd asked me a week ago, I'd have said it rounds to zero. (And now I'd still say that! I mean, it was worthless here too.)

Instead, you pay underwriters for the work of putting together the prospectus, figuring out the right price, finding investors, marketing the deal, all that good stuff. And here it must be said that Deutsche Bank and Wells Fargo looked to be getting overpaid a bit. I mean, for one thing, you could not really call this deal a success on pretty much any metric. But also, the banks charged VBL a fee of 7 percent on the full size of the offering. Now 7 percent is pretty normal for a small initial public offering, but remember that over half of this deal was being sold to existing VBL shareholders. Presumably VBL could have found those guys without the banks' help. (They found them before, after all.) If you just divide the banks' $4.5 million fee by the $29.8 million that VBL was planning to raise new public shareholders, it works out to about 15 percent. Which is a pretty rich fee.

Of course now they're not getting the fee, so I guess that mitigates things a bit.

The other thing that IPO fees pay for is the underwriters' legal risk of vouching for an offering. People like to sue underwriters when things go wrong. And here things went wrong. If you bought shares on Friday the 1st at $9.40, and sold them on Thursday the 7th at $11.15, then you'd be pretty aggrieved to hear that your trades are getting unwound.7 And ... I mean, I don't want to give you legal advice, but this sure feels like the sort of grievance that a court would be interested in? The prospectus says "The underwriters expect to deliver the ordinary shares to purchasers on or about August 5, 2014," and yet they blithely let everyone trade until August 7 without mentioning that things had gone sour.

On the other hand this is not really a ton of money; only about 1.7 million shares traded after the IPO, at a volume-weighted average price of $10.28. (Even if half of those trades were at the $9.40 low on the 1st and half were at the $11.15 high on the 7th -- they weren't -- then investors would only have about $1.5 million in unwound gains.) And surely the larger group of investors who paid $12 to buy shares in the offering, and then saw their shares fall below the deal price and stay there, will be pretty pleased to see the deal unwound. This deal is an embarrassment all around, but at least everyone will be happy to forget it.

1 0.07Y is the underwriting fee. Bank buys at 93, sell at 100, and takes 7 for its troubles.

2 A fun, by which I mean terrible, thing to talk about is whether the settlement is T+3 or T+4. Regular-way U.S. equity markets settlement is T+3. But with a public offering, like VBL's, you price the deal (and sign the underwriting agreement) on Wednesday (day 0), it is allocated and starts trading on Thursday (day 1), and the shares are delivered the following Tuesday (T+4 from pricing but T+3 from the allocations and the first actual trades. If you're an investor who got shares allocated on Thursday morning this feels like regular-way T+3 settlement, but in some technical sense for the banks it's T+4.

Another technicality: That link was to the form of underwriting agreement that VBL filed with the Securities and Exchange Commission. It's not the executed underwriting agreement, which (according to the prospectus) was signed on July 30, but which was never filed. Usually one files that but I guess things got weird fast.

3 The boilerplate to the announcement helpfully explains that it's "Vascular Biogenics Ltd., operating as VBL Therapeutics."

4 This is not laid out simply but basically on pages 149-150 of the prospectus there's a table of principal shareholders and the footnotes show that four of them committed to buy: the Keffi Group, a New York fund that's an 24 percent pre-IPO shareholder and that's run by VBL director Jide Zeitlin; Aurum Ventures, an Israeli fund that's a 23.5 percent shareholder; Pitango Ventures, an Israeli fund that's a 11.7 percent shareholder and that's related to VBL director Ruth Alon; and VBL director Jecheskiel Gonczarowski, who seems to be an Israeli. Some rough math shows that the Keffi Group -- apparently the only "existing U.S. shareholder" that committed to buy in the deal -- was on the hook for the bulk of those shares, around 2.4 million shares for a price of around $29 million, or almost half the deal.

5 Incidentally, you can read in textbooks or wherever about "firm commitment underwritings," in which the underwriter "guarantees to purchase all of the securities being offered for sale by the issuer regardless of whether or not they can sell them to investors," but these descriptions do not correspond to lived reality. All regular-way U.S. public offerings are "firm commitment" underwritings, in the sense that once the banks sign the underwriting agreement they're obligated to pay for the shares unless the company fails to satisfy the limited conditions in the agreement. But in fact all IPOs and most follow-on equity offerings are also "marketed deals," in the sense that the banks never sign the underwriting agreement until they've already gotten firm orders from buyers for all of the shares in the deal. (And then some. Recall that Alibaba's underwriters are looking for at least four times coverage.) Once they do sign the agreement, then in theory they're on the hook for three days of risk: If investors pull their orders, the banks still have to buy the shares from the company and then go after the investors themselves. But this is usually not much of a commitment. I mean, come on, it's three days.

6 UPDATE: Pseudonymous Twitter user IvanTheK makes the good points that:

  • VBL's prospectus says that the current shareholders had agreed to buy the 2.9 million shares and subject them to a lockup.
  • When those shareholders decided to skip out on that agreement, that disclosure became, perhaps not untrue (they had agreed!), but arguably misleading: If you bought shares thinking that 2.9 million of the new shares would be sold to insiders and locked up, and then found out that they were instead languishing with the underwriters and about to flood the market, you'd be miffed.
  • The conditions to the closing in the underwriting agreement (section 6) include various legal opinions and officers' certificates certifying that the prospectus is not misleading.
  • Those opinions might be tough to get given the gap between what the prospectus says and what happened.

My gut sense is that the prospectus is not misleading -- it never says that the insiders will buy, but only that they've agreed to buy, and is careful to phrase everything about their actual purchases hypothetically -- but I can see how it would cause concern. Perhaps enough concern to give the underwriters an out.

7 From the Journal again:

On Monday, market officials began combing through trading records of the stock and contacting trading firms as they began to unwind the transactions and refund investors. Officials at the National Securities Clearing Corp., where stock trades are finalized, handled the unwinding process.

Also I guess you'd be aggrieved if you shorted the stock at like $11 and then covered lower, but you probably didn't. It's pretty hard to borrow a just-IPO'ed stock, especially one without much float or liquidity and with lots of insider holders.

To contact the writer of this article: Matt Levine at mlevine51@bloomberg.net.

To contact the editor responsible for this article: Tobin Harshaw at tharshaw@bloomberg.net.