Hacker News Reader

Comment by mikeash | Hacker News Reader

This seems to be a pretty common thing. Why don't IPOs use some sort of auction structure to discover the right price at the time of the IPO, rather than setting a price up-front and trying to figure out the optimal number ahead of time?

View on HN

Because then the IPO underwriters (big investment banks) wouldn't be able to make a shitload of money on top of their IPO fees.

Google used a Dutch auction in their IPO. But then again they are pretty obsessed with auctions. There is some research on auctions for IPOs, e.g. http://www.nber.org/papers/w16214, that suggests they are too complex for investors.

And Google still had a 15% first-day pop, which is just about the normal size for a traditional bank-run IPO.

It absolutely it not normal. If there was ALWAYS a 15% pop, then IPO prices would be 15% higher.

It's not a given that a stock that just popped 15% trades the same as a stock that was offered at a 15% higher price in the first place.

No, because you absolutely want that pop, so you get positive press. The last thing you want is an IPO where the price goes down

"Academics have found that I.P.O. underpricing is ubiquitous. Jay Ritter has documented underpricing over the years. According to Professor Ritter, the average underpricing for I.P.O.’s in the United States was 14.8 percent from 1990 to 1998, 51.4 percent from 1999 to 2000 and 12.1percent from 2001 to 2009."

"Over the last 50 years, I.P.O.’s in the United States have been underpriced by 16.8 percent on average. This translates to more than $125 billion that companies have left on the table in the last 20 years."

Normal over the last 50 years, maybe.

http://dealbook.nytimes.com/2011/05/27/why-i-p-o-s-get-under...

I remember the google IPO well; I had resolved to only do trades my advisor liked at the time (I was 25 or so and well meaning, really!)

The street hated Google's Dutch auction, and refused to push it to retail investors; a very unusual move at the time. The IPO estimates were at one point over $100, but if I recall correctly, the auction dropped down into the $80s, at which point at least one of the VCs pulled out of the IPO completely, preferring to hold their shares.

Anyway, despite these incredibly bullish indicators, my advisor didn't like the IPO.

It was very complicated to participate in too; you couldn't go through your normal channels. I think the complexity would be easily manageable by a modern UI/UX pro; the underlying market mechanics aren't that difficult. But hitting '90s era UI/UX with unwilling brokers was a tall order.

I'm having a hard time parsing some of what you wrote at the end there. First, Google's IPO was in 2004, not the 90's, and e.g. E-Trade was very well established and not so different from today. In any case I don't see what UI/UX have to do with it.

Google weren't that happy with how the auction IPO turned out, actually. Their second offering a year later followed Wall Street convention.

Well, they are using a sort of auction structure, just not a pure one. The "pop" is by design to signal a successful offering. The criticism of "leaving money on the table" is bogus. You want to set yourself up for a secondary where you'll raise most of your money.

That's an interesting way to look at it. How common are secondary offerings? I had the impression that they were pretty rare, but I can't think of where I got that idea.

I guess because Google's Dutch auction was considered a failure.

It was only a failure for investment bankers who want to collect the profit of a pop on the IPO.

Perhaps a naive question, but how do they benefit from the IPO pop if they are required to hold for a quarter or two?

They benefit because the price of the stock post-pop is a better predictor of the price a quarter or two later than the price of the IPO.

It seems to only work well with largely popular companies. Google's auction IPO was pretty successful, Rackspace was not so much.

Was Rackspace's trouble because of the auction, or just because of the company and market conditions? (In hindsight, August 2008 was probably not the ideal time for an IPO.) If they had done a regular IPO, might they just have had trouble selling all the offered shares at the set price?

(These are all honest questions, just to be sure. I have no preconceived idea what the answers might be! I'm not clued up enough to argue with rhetorical questions on this stuff.)

Not sure. I know Google's IPO was during a time that wasn't favorable, but it also had a strong B2C brand, which would have biased investors and made the auction more favorable.

Part of the service the banks are supposed to provide is to help with timing the IPO. (I have no idea how good they are at doing that, though).

It happens, this is true, but I wouldn't say it's common to make 60% in a few hours.

I suppose there's probably a lot of reporting bias, where IPOs which jump a lot are the ones that make the news.

Even if it's uncommon, an auction would ensure it doesn't happen to you. Maybe!

Reply on news.ycombinator.com