Using Dutch Auction to Sell IPO
https://www.economist.com/finance-and-economics/2004/05/06/cursed
This article in Economist talks about the pros and cons of Google using “Dutch” auction to sell its IPO, and offers advice on improving the auction system. Google chooses to auction its IPO because it wants the whole process to be more public and transparent: potential shareholders wait for the price to go down and bid when they think is appropriate, and when all the shares are sold, everyone pays the lowest price. In this scenario, there is no dominant strategy. Some people might bid early, because they don’t want to wait until it reaches their true value, because at that time all the shares might have been gone. In this scenario, people are very likely to bid higher than their true value, because they expect other people to follow up and bid at lower prices, so they would expect the final price that they need to pay to be lower than what they bid themselves. Following this analysis, this auction will very likely end up in too high prices.
So what are the issues that might come with too high prices? On the one hand, one can argue that the investment bank which helps establish the IPO will benefit more from the higher prices, because it gets more service fee. On the other hand, Google doesn’t want its shareholders to suffer the “winner’s curse”. It’s a concept introduced in the article, and it means that even a buyer wins the auction, they actually get negative payoff, because they overpay. Besides shareholders losing, if prices are higher than actual values, the price of stock will drop quickly after the IPO which bring instability to Google.
After discussing pros and cons of such auction, the article also gives two potential solutions to improve the auction system. One is the “English” auction that we learned from the class. Prices in an English auction goes up, and it ends when no one offers a higher price and the winner only pays the second highest price. This is more reasonable because in this scenario the dominant strategy is for buyers to bid their true value, which prevents the bidding price from going too high. An alternative solution the article suggests is to keep the Dutch auction, but divide it into several steps: first auction 10% shares to say, institutions, then follow up to common buyers. This gives people who don’t have much experience in auction to get the chance to observe how it works and what are the better strategies, so that when it is their chance to bid, they are less likely to bid too high prices and then actually loses money.