A few weeks ago there was a decent amount of talk in the blogosphere about how the LinkedIn IPO was underpriced by its investment bankers. Henry Blodget was one of the loudest sounding the alarm bells. LinkedIn was screwed out of $175 million dollars! His assertion? The investment bankers intentionally priced the IPO well below where they knew the company was really worth so that they could put more money and profit in the pockets of its institutional clients (investors like Fidelity, Putnam, and countless others). They knew that it would pop to $94 on the first day and yet they still priced it at $45! What were they thinking? Must be greedy, right?
Continuing on this theme Hank was at it again this week following the Pandora IPO. Finally, a hot tech IPO priced correctly! They priced it at $16 and it closed the first day at $17.42, a fair and reasonable increase of 9%. Perfect!
I find Henry as interesting to read as the next guy, but in this case I think these assertions are a bit misguided. He’s obviously a very smart guy and knows finance very well, but I think the analysis ignores two key facts.
1) It’s not about the first day of trading
Yes, LinkedIn closed on day 1 at $94.25, representing a premium of 109% compared to the $45 issue price, well north of the typical 15% IPO discount. But yesterday – less than a month from its IPO – it closed at $68.27, representing just a 52% premium. Is LinkedIn really worth 28% less than just 4 weeks ago? Did something go horribly wrong with the company? Of course not.
How about the perfectly priced Pandora IPO? Yesterday in it’s second day of trading the stock closed at $13.26, 17% lower than it’s $16 issue price. How do you think Pandora feels about that? Yes, they maximized the proceeds to the company and the selling shareholders, but its long-term shareholders – the ones that didn’t flip the stock on the first day of trading – have now lost money (on paper anyways). It’s not going to make or break the company over the next 10 years, but it’s probably not how you want to start your life as a public company.
2) Retail investor demand is difficult to predict
Hank is also assuming that the investment bank responsible for pricing the IPO knows exactly where a stock will trade once it is available to trade on public markets. After all, as the lead bookrunner of the IPO they get to see the demand. They know what every investor is really willing to pay. They should be able to predict where it will go.
Intuitively that makes sense and it is often the case, but unfortunately when a company that primarily serves consumers completes an IPO you’ll often get lots of aftermarket buying by retail investors. They are users of LinkedIn or users of Pandora. They love the service. They believe in the company. So what if they don’t get to buy at the IPO price. They are in it for the long-haul. Throw caution aside and buy, buy, buy. The problem with this dynamic is that this retail demand can cause short-term price movements that have no basis in reality. Institutional investors have a better sense for what a company is really worth and ultimately a company will trade to that value. But in the short term a stock can trade wildly based on irrational and uniformed buying (or selling) by retail investors.
But why can’t the investment banks predict retail investor demand? Herein lies the biggest problem. Investment bankers will distribute IPOs through their own retail investment channels and through co-managers of the IPO, but they don’t open it up to the broad public. If you don’t have an account with one of these brokerage firms the only way you are going to buy into the company is by making purchases in the aftermarket. Further compounding the problem is the manner in which investment banks solicit indications of interest from their retail investor channel. They aren’t really going to each and every customer asking them how many shares they are willing to buy and at what price. The indications of interest are typically submitted at the branch level based on what the manager of that branch estimates they can sell. Bottom line, while the investment banks do have a pretty good sense of how their institutional investor clients will behave in the aftermarket, they have much less visibility into how their retail investors will behave, particularly for companies like LinkedIn and Pandora that are well known and loved by consumers.
The solution: a Dutch Auction IPO
So what can we do? If you can’t predict how retail investors will behave, do we just have to roll the dice knowing that sometimes a stock will likely shoot up (LinkedIn) and other times it will trade below issue (Pandora)? Are we doomed for all future IPOs attracting significant retail investor demand (i.e., Facebook)? Thankfully, I think there is a better way. I think the answer is the rarely used dutch auction IPO process.
You can read about what a dutch auction is here, but the important thing to note is that if done right it opens up the bidding process to any retail investor that is interested. Google made this structure famous back in 2004 when they used it for their IPO. Don’t have an account with one of the investment bankers managing or co-managing the IPO? No worries, because of the IPO you can set up an account, place your own bid, and actually get an allocation assuming your bid is above the ultimate issue price.
That all sounds good in theory, but how did it work in reality? Google priced their IPO at $85. On day 1 it closed at $100.34, representing an 18% increase. How about 2 weeks later? On September 2, 2004 – exactly two weeks later from it’s August 19 debut – it closed at $101.54.
There is no such thing as a perfectly priced IPO, but that ain’t bad. Everyone won in that scenario. Google maximized their proceeds. Long-term institutional investors still got the 15% IPO discount. And retail investors had access prior to a first day pop. This last point it very important. In both the LinkedIn and Pandora IPO’s the retail investors – the ones that love those services – are really the ones that get screwed. They buy at the height of the mania, and then within in weeks (LinkedIn) or days (Pandora) when the mania stops and the market value moves to intrinsic value they find themselves way underwater.
Facebook will supposedly go public in 2012. I’m hoping they – like Google – can be just as innovative with their IPO as they are with their technology innovations.
One thought on “IPO pricing”
Facebook should be innovative, but not in the way Google was – by adopting a method that had consistently failed elsewhere. Instead, Facebook should adopt a method that has been consistently successful and popular around the world. The way to open up IPOs to retail participation is through a separate public offer tranche, as is so common in Europe and Asia.
If we want to know how IPO auctions work, why focus on just one particular offering when there have been hundreds? IPO auctions have been used in more than two dozen countries over the last few decades, but the method has been rejected by issuers, due to the many problems it created. Some of these were discussed in the risk factors section of the Google prospectus, but there’s plenty of academic research detailing the problems and track record more specifically. Here’s one paper on the subject: http://ssrn.com/abstract=874344