Sharky - 4:39 am on May 4, 2004 (gmt 0)

In an ideal auction, everyone who is interested in buying the stock would register and place a bid for the maximum number of shares and maximum price they are willing to pay. The seller would then choose a price cut-off that produced the amount of desired revenue. Winning bidders would pay the price they bid and get all of the shares they bid for. Losing bidders would get no shares. If everyone bid the maximum they were willing to pay, everyone would be happy. There would be minimal un-met demand, and one would expect the result to be a minimum of volatility on the day after the auction (especially if the investors were all in it for the long-term).

Of course the auction will not be ideal. It will differ in many ways, including:

-- Not everyone who wants to buy the stock will register (People outside the US, for example, are excluded)
-- Not everyone who bids will bid the most they are willing to pay (anyone who has used eBay knows this)
-- Winning bidders aren't guaranteed that their allocation will match their bid (the rules for allocation as outlined in the S-1 are fuzzy at best)
-- Winning bidders pay the lowest price accepted, rather than the price they bid
-- Google can't choose what kind of investors they get. There will be tons of speculation and lots of people will be in it because they think they can make a quick buck
-- Google seems to be reserving the right to sell more shares, if they feel the demand is there. They might sell the shares during the IPO, or in the open market afterwards.

How will these differences effect the post-IPO market? There will definitely be some un-met demand. If I was willing to pay $100/share for 1000 shares, and the accepted price is $50/share, I might be willing to buy 5000 shares at that lower price. If I didn't get my full allocation, that would also leave me wanting more.

Some people will be uncomfortable with a sealed-bid auction. On eBay you can at least see what the last bidder is willing to pay. With a sealed bid, it's hard to know what a fair price really is, unless you can see the entire order book. The order books are visible on the open markets, so I'm surprised that Google and their underwriters have chosen to hide them for the auction. I would wager that an open order book would actually result in a higher net price than the planned closed one.

By opening up the IPO to more investors, I believe Google is inviting increased volatility. When you bring in the masses, that means lots of novice investors who don't have the stomach for market fluctuations.

Underwriters learned a lot from IPOs like Yahoo, Akamai and eBay. They left a ton of money on the table. It hurts the underwriters as much as the companies. Google is clearly making an effort to avoid that problem. If they do their job perfectly, the price at the end of the opening day will be at (or even slightly below) the opening price. We know they won't be perfect, but how close can they get?

What will happen when the unmet demand meets a flood of novice investors? I'll wager that the novices get creamed. Here's one possible scenario: the price goes up at the open while public orders get filled. About a half-hour after the open, it's up by 6%. Then it starts moving down with some early profit-taking. On the first sign of weakness, more experienced investors begin to sell short into the dip to help it move. The goal is to wash out the novices who don't have the stomach to ride it out. There aren't too many novices who can handle seeing the price tumble by 20%+, especially when they thought for sure that they were going to double their money. After moving down a few percent, some novices get nervous that the price is rising meteorically, and begin selling, creating a snowball effect. After a frenzied 3-hour decline, the price starts to recover, and the more experienced investors first cover their shorts and then begin to buy. After dipping down by 20%, the stock closes the day up 2%.

Does it make sense that it would be possible to get a bargain when the auction is open to everyone? The reasons past so many past IPOs went crazy were typically extremely poor pricing and artificially-induced scarcity. Neither of those factors seem to be at work here.

But hey, if you guys want to buy it that way, please do! I'll be waiting for you to sell to me at the bottom of that first big dip.