The reason for the underpricing is simple: it was the first big Dutch Auction IPO and when something is new, only the few will participate leaving the hesitant to jump in once it trades on the open market, bidding up the price.
This left the auction open to certain big institutions (Fidelity and another I forget) to badmouth the deal to lower the price while buying big themselves. Of course they were joined by the Investment Banks who really like that 7% standard IPO fee and felt threatened. And the press to whom everything is either an epic success or huge failure.
As a first, the deal also had many technical issues which lead to the August pricing and fed the negativity.
Anyway, the key win was that anyone (even a tiny investor like me) could buy in the IPO. Just try to get shares in any decent IPO run the normal way. If you do, it's likely to be a broken deal they couldn't find demand for and you'll most likely lose money at first trade.
I disagree with another HN commenter on the value of this article. Taking two backward steps through Eric Schmidt's conclusion:
- "I wouldn’t change a thing - except maybe (an interview that Playboy published putting Google in breach of the SEC-mandatory quiet period)"
- "To this day I can’t fully explain why our stock price opened so high—causing the pop we had tried to avoid [Google sold its shares at $85; three days later people were selling what they'd bought at $110]. A lot of complicated factors played a role" (none of which are discussed).
Basically, they screwed up, left a huge amount of money on the table, and he doesn't admit to the possibility of any possible improvements, the possibility that doing the IPO at that time was the wrong call, or discuss the factors behind the pop despite their strategy to try and avoid it.
Firstly, Could Google not have just stayed privately owned, but published its financial records for a few years, before then doing their IPO? From this article it doesn't sound like the IPO cash was particularly important at that point, and the meteoric rise in the share price suggests (with 20:20 hindsight) that they let far too much go, far too early on. The crucial paragraph doesn't discuss the decision process to do the IPO, it only lists the options. That's disappointing.
As for why they went for a Dutch auction - to bust the institutional racket (good on them) and because "It was consistent with the auction-based business model we used to sell our ads, so we felt we understood the underlying dynamics, and it had a strong intuitive appeal for us." Clearly that understanding wasn't worth shit - it doesn't seem to have helped them predict anything about what happened in the auction. The anti-establishment attitude was really commendable, but it seems like it was a huge strategic mistake. Some factors may of course have been out of their hands - public inexperience/wariness with the Dutch auction format leading to low initial bidding, pseudo-concerted action to keep the bids low, etc. But that's a strategic mistake which a strategy exec should be owning up to. The two first weeks in August before the bidding opened were stuffed with terrible Google news (and he sets up a surprise Playboy publication of an April interview as a paper tiger - but see the diagram in the article for the rest). They got pushed into August because of the complexity, and it is a bad month to do an IPO, but he says they didn't care.
He bemoans the press for portraying their approach as 'hubristic' and therefore worth staying away from. Yet the press was spot on with their analysis of the factual situation - it was hubristic, seemingly not particularly well thought through, or even diligently executed. What the press got wrong is in concluding that it was something to stay away from. Quite the opposite - the people who got in at the Dutch auction stage made a vast amount of money! Google didn't, though.
So I found the pomp and pride underpinning the article really quite unsubstantiated - and yes, particularly hubristic. It was innovative, done with some honest and interesting motivations, but it was a cockup, from top to bottom. I disagree that they had 'risen to the task'. They do that in their core business, undoubtedly, day in, day out, but this was a fuzzy mess and I'm always worried by CEOs who don't learn from cockups - substantially more so when they deny that the cockup even existed
Google's initial public offering took place on August 19, 2004. A total of 19,605,052 shares were offered at a price of $85 per share.[27] Of that, 14,142,135 (another mathematical reference as √2 ≈ 1.4142135) were floated by Google and 5,462,917 by selling stockholders. The sale raised US$1.67 billion, and gave Google a market capitalization of more than $23 billion.[28]
So - assuming the market established a price of $110, and the sold at $85. Google lost out 14,142,135 x $20. Now $280M is not to be sniffed at, but for a company with a market cap of $23,000M it's little over 1% of it's value. Given the shares had risen 25% on the opening day, 1% is not really material.
Given that the company was cash flow positive, the difference between raising $1.67B and and $1.93B of would probably not have any operation affect on the business.
I'm not saying a dutch auction is necessarily a worse option. Indeed it might theoretically be the best. But they spent a vast amount of money setting it up, with all the extra legal fees and transactional costs, so let's not focus entirely on the money left on the table by the strategic error to a) pick a theoretical, rather than tested, procedure that was b) seemingly, poorly executed and c) if it yielded any lessons at all, for which there is no evidence, Schmidt refuses to share.
On a pedantic point, if my understanding is correct, google made (14m * 85) about 1.2 bil on the IPO, not 1.67 - the rest being sold by other stockholders (who, incidentally, 'lost' $136.5 mil due to the pop, and made $464m, minus fees for being involved in the IPO - not insignificant)
That means that Google made 1.2 bil, but had they 'got it right' (very difficult, of course, as other commenters point out, 14% pop is usual on IPOs) they would have made $1.55 bil. You can make it seem small compared to their market cap, but the headline is: he's proud at a messy, protracted transaction, possibly the most important in Google's history, that netted them $353m more than their 1.2 bil haul. That's 20-25% on the table, well in excess of the 14% bandied about in this thread.
And all that says nothing about the overarching questions of 'Why IPO, why then?' and 'what actually went wrong?'
With regards to the Dutch Auction it may not have captured the maximum value for Google, but did it capture more than the alternative?
According to http://www.businessweek.com/1999/99_47/b3656143.htm the average IPO gained 14.1%, which would suggest they did leave money on the table, although I'd question if Google's IPO would be described as average. By comparison Akamai gained some 458% in it's first day of trading.
