Google IPO Analysis

Several clients have asked me to comment on the upcoming Google IPO. Since this particular auction seems to be attracting significant mindshare, I thought I’d post the  “Auctus Perspective”. 

The situation in brief:

  • The auction method to be used is a multiple unit, homogenous, modified Vickrey auction.
  • The offering price will be determined after the auction closes and will use the auction clearing price as the “principal” factor, but not the only factor in determining the offering price.

A pure Vickrey auction is also known as a “closed second-price” auction by the academics. In the financial community it is also called a “Dutch” auction. However, academics use the term “Dutch” auction to refer to an entirely different type of auction. Be careful what you read, reporters have been known to get these mixed up.

A pure Vickrey works as follows:

  • Everybody submits the number of shares that they would like to buy and the MAXIMUM price that they are willing to pay for each of those shares.
  • After all bids have been submitted, they are opened and sorted into descending order by price and a running total of shares requested is calculated.
  • The clearing price is the price at which the running total of shares requested equals the number of shares offered. There are some nuances related to whether the clearing price should be the lowest winning or the highest losing price, but they are insignificant in this particular context.
  • All bidders who bid at or more than the clearing price receive the shares that they requested and they pay the clearing price (which will always be less than or equal to their maximum bid price).

For example:

If 1,000 shares are offered for sale and the following five bids were tendered:

            Bidder A                      300 shares  @            $100 each        300 running total

            Bidder B                      300 shares  @            $  80 each        600 running total

            Bidder C                      200 shares  @            $  75 each        800 running total

            Bidder D                      300 shares  @            $  70 each     1,100 running total

            Bidder E                      200 shares  @            $  50 each     1,300 running total

Then Bidders A, B and C would get all the shares they requested and Bidder D would get 200 of the 300 he requested. All four bidders would pay $70 for each share regardless of the fact that some bidders were prepared to pay more.

In a pure Vickrey auction the optimal strategy for bidders is to bid their true maximum value. So if a bidder is willing to pay $117 per share then that’s what he should bid. This is different from the optimal strategy in a traditional English auction or a “seal-bid” auction.

So if Google’s IPO auction were a pure Vickrey (and all bidders were rational and knowledgeable) then the post-IPO market price would equal the IPO clearing price. There would be no “pop” (or drop) in the stock price on the first day.

However, the auction is not a pure Vickrey and all the bidders will not be rational and knowledgeable.

Not pure Vickrey: Google reserves the right to set the offering price LOWER than the auction-clearing price. The shares requested at or above the offering price will be allocated (roughly pro rata) among all the bidders who bid at or above the offering price. This will tend to cause the post-IPO price to rise from the offering price to the clearing price. Since Google plans to use the auction clearing price as the “principal” factor in determining the offering price, I expect the offering price to be lower than the clearing price, but not by much.

Bidders not rational and knowledgeable: Many bidders, particularly retail bidders, will not fully understand this particular auction process and not understand their optimal strategy.  

  1. Some will bid more than they think it’s worth thinking that they will pay less than their bid. In most cases that will be true. However, if the offering price ends up being more than they think it’s worth but less than their bid, they will pay too much. I suspect that these bidders will hang on to their newly acquired stock and not affect the short-term post-IPO market.
  2. Other bidders will bid less than they are willing to pay for a variety of specious reasons. This is a common phenomenon on eBay, which also uses a modified Vickrey method (albeit a differently modified method). These bidders could decide to buy shares in the post-IPO market at more than their original bid, but at or less than their value. This would tend drive the price up.

In conclusion, I expect Google’s post-IPO stock price to show a gain in the short-term after the IPO. I do not expect it to be on the order of 25% or more as we saw in the Dot-Com boom days. Instead I expect it to be positive, but in the single digit percentages, say on the order of 5%.

My conclusions are based on “all other things being equal.” Significant news and non-auction related market factors could affect Google’s trading prices, the technology sector or the market as a whole. “A rising (or falling) tide raises (lowers) all boats.”