Coverage in the press of the Facebook IPO has been sensational, with headlines about Facebook’s “stumble” at the IPO. In this post I’ll suggest a way to think about what happened, who won, who lost, and whether we should care about the decline in Facebook’s share price at the IPO. My answer: No, we shouldn’t care.
During an IPO, a firm and its shareholders wish to sell shares. Generally they want the highest possible price for these shares. Investors, on the other hand, want to pay a low price. The underwriter’s job is to strike a balance. To understand what happened with Facebook, you need to appreciate the difficulty in striking this balance.
The investors who buy in an IPO offer to buy up to a particular number of shares. Suppose Joe Average subscribes to the IPO, offering to buy 1000 shares for the offering price ($38 in the case of Facebook). If demand for Facebook is weak, that is, if there are not many subscribers, then Joe will get the full 1000 shares. However, if demand is strong, there will not be enough shares to go around and Joe will get fewer than 1000 shares. This asymmetry is the key to understanding IPO pricing. Joe thinks: When I get the full 1000 shares, the price will go down, but when I get fewer shares the price will go up. In order for Joe to willingly participate, he must expect that on average, the price will go up at the offering. This positive average return compensates him for getting fewer than 1000 shares in good times. Although the price goes up on average, sometimes it will go down. You can understand why this occurs by thinking about the strategy of those investors who actually have information. They will bid for many shares when the offering is valuable, and for few shares when it is not. This is the flip side of Joe getting all his shares in bad times and fewer in good times.
So investors lost money in the Facebook offering and we understand that sometimes this is going to happen. Did Facebook do anything obviously wrong? Surely Facebook management should have pushed for a high price and that’s apparently what they did. Does the IPO bode ill for Facebook? Why should it? Facebook is one of the most recognized names in the world. In the future, investors will judge Facebook by its financial success or lack thereof. Do you think that Facebook users will switch to Google+ because the stock fell at the IPO? If the offering price had been $32, Zuckerberg would have earned almost $200 million less in the offering. He looks to me like a smart guy.
The party on the hot seat is Morgan Stanley, the lead underwriter. Their institutional investors will want to know why Morgan Stanley agreed to a $38 price. These things happen: Morgan Stanley will make its mea culpas and be back the next time around.
Finally, what about the retail investors who participated? Were they treated fairly? Press reports made it sound like it should have been a sure thing: buy Facebook at $38 and flip the shares a few hours later at a higher price. Of course it could have happened that way. But anyone sure it was going to happen was expecting Mark Zuckerberg and Morgan Stanley to hand them free money. Here’s an Insider Tip: Zuckerberg and Morgan Stanley are not in business to give money to you. They make money from you.
Brokers and banks love to deal with investors who think otherwise.
I don’t think anybody in big media has explained what happened with Facebook’s IPO so clearly, refreshingly, and with brevity.
Bravo!
A question about Joe A., wouldn’t aggregating all ‘Joes’ would enable bankers/FB to construct a demand curve and based on it choose ‘the one’ optimal price (assuming demand behaves ‘normally’, i,e is a monotonic function) that would maximize proceeds per given number of shares (or alternatively min number of released shares per given desired proceeds)?
If that’s the case why did we see the offering price (determined by bank/FB) keep changing before IPO? (b/c of new ‘points’, that reflect new bids, that were added on the demand curve?)
-Yossi
Great question. I don’t know the answer for sure, but I can offer some thoughts.
You can imagine investors providing information like this: “I would like 100 shares at or below 38, 100 additional shares at or below 37, 300 additional shares at or below 32.50”. Now the bankers can aggregate this information and figure out how many shares they can sell at each price.
Google tried to bypass the traditional IPO, attempting something like what you suggest. They ended up offering shares at a low price, giving investors only 75% of the shares they had requested. The offering was at $85 and the share price closed at $100 on the first day.
I’m not an expert in auctions, but there can be practical problems when investors can look at bids by others. They might reassess their own valuations and alter their bids. And you still have the problem that some investors may be privately informed and might want to defer putting in their bids (or withdraw their bids) until the last feasible moment.
So in the end I suspect you are correct, that there was a continual reassessment of Facebook prior to the offer, with new bids and withdrawn bids.
The way “bookbuilding” (the main US method) works is that there’s frequent communication between the bankers and investors. During the week and a half to two weeks of the road show (Facebook began its road show on a Monday and started trading on Friday of the next week; they never start trading on a Monday), people in New York who were at the first day of the road show got a call the next day, while the road show was in Boston. On day two, those first investors had a chance to say how many shares they’d take at various prices, but the orders weren’t binding.
These “indications of interest” are recorded – the idea behind the name bookbuilding is that they’re building or filling the order book. For Facebook, the book was ‘filled’ (one time) by about Thursday or maybe even Wednesday of the first week, according to press rumors. But the New York and Boston investors that attended the road show in the first two days would keep getting calls to update them, and all investors could change their orders based on how things were going. There are some investors that won’t place orders until the offer is at least two times subscribed – those are free riders, obviously, but they may still get a few shares if they’re good overall customers. And then there are well respected investors that have their own strong idea on the price, and the issuer may have to price lower to get those investors into the offering. Last week I had dinner with someone closely involved with the Google IPO who said that Google, in the end, had 5 top investors that Google especially wanted to have in the offering, and so they chose the price in part to make sure they got those five (it was “dirty Dutch”, so they had the option of pricing below clearing).
We should remember that Facebook is only one offering. It was a mess, but it’s even trickier to price these very hot, high demand offerings. Just a few days before trading, GE announced that it was no longer going to advertise on Facebook because the adds didn’t work, and this news made no difference in the demand for the shares.
By the way, your colleague Ravi Jagannathan is an expert on IPO auctions.
Professor McDonald, Thanks for the great article. I am sharing this.