With their initial public offering this month, Google's founders and biggest shareholders Larry Page and Sergey Brin -- the fellows who vow that theirs is "a company that does good things for the world even if we forgo some short-term gains" -- are on the cusp of becoming googolaires. Well, actually only a deca-billionaire duo. A googol is a 1 followed by 100 zeros and so they will be just 90 zeros short if the offering meets expectations. But it's a start.

Along with changing the Web, Google's founders also say they want to make the world "a better place" -- and the way they've chosen to do their IPO is a fine example of practicing their corporate mantra: "Don't be evil." Instead of playing the wicked IPO pricing game -- in which investment bankers get to buy shares on the cheap and allocate them to their buddies or favorite clients -- Google has created a level playing field. It will hold an auction and let supply and demand determine the initial price of its shares. That price will be the highest the company can get while still being low enough to ensure that there are enough bidders to gobble up all 24.6 million shares Google plans to sell now.

Doing no evil is one thing, but whether this offering makes sense for you to invest in is another. A lot of people want to bid in this IPO auction for the simple reason that they can. At long last, they are being invited to play in the big IPO leagues. But while Google has made its IPO into a level playing field, my advice to investors is: Stay on the bench.

Here's why: Auctions have been around a lot longer than the Internet (or search engines). People go to them, as in the arts and antiques markets, hoping to find bargains. Getting bargains, however, isn't so easy. As I teach my MBA students, bidding in auctions is a tricky business. If you bid too low, you might not win. And if you do win, it's likely that you bid too high. That's known as the winner's curse. Or, to paraphrase Groucho Marx, you might not want to bid in any auction that would have you as a winner.

And in this case, the risks are greater because Google has the advantage of getting to see all the bids as they arrive and then deciding when to end the auction. Investors have to bid blind; they can't see the other bids. Google can essentially say, going once, going twice, still going twice. . . .

With all of the hoopla over Google's auction, it is easy to forget that there's another choice. If you want to invest in Google, just wait a day or longer and buy the stock on the regular open market. Don't bid in the auction.

My advice has little to do with whether or not Google will turn out to be a good investment over the long run or even the short run, or whether the Google IPO is evidence that the bubble market mentality is coming back. There's reason enough to wonder about that without my expertise, which is game theory. Google estimates filed with the Securities and Exchange Commission predict that the Internet search engine company's market capitalization, its total value based on its share price, could be as much as $36 billion the day it goes public, as much as The Washington Post Co. (including Stanley Kaplan and Newsweek), the New York Times Co., Dow Jones & Co. (owner of the Wall Street Journal) and Apple Computer Inc. combined. Google's profits would have to grow nearly tenfold to bring the ratio of its stock price to earnings per share, a standard measure of whether a stock's price is justified by its profitability, down to the market average.

All the same, I'm not saying don't buy the shares. I'm just saying: Remember, there are two ways you can become a Google shareholder.

You can read the 120-page prospectus, watch the meet-the-management presentation on the Web, open up a brokerage account with one of the 21 underwriters, get a bidder ID, and then decide what price to bid and how many shares to bid for (or perhaps even make multiple bids at different prices). When the company sorts out all the bids, you'll find out whether you "won" any shares at a price you were willing to pay.

Or you can let the auction do its magic and then place an order on the open market once the stock starts trading. Unlike the folks who bid in the auction, you'll know exactly what price you'll be paying and how many shares you'll get.

What's to lose by doing that? If the underwriter purposely underprices the stock before it's issued, then there's an advantage to getting in early. You get to buy the stock at a discount (if you can get any shares). And you could make money by selling on Day One as everyone else scrambles to buy in. But with Google's open auction format, there's no reason to think that the auction price will be anything other than its market value. How could it be otherwise?

It is possible for people to bid irrationally. But in this case, the auction enables the Googlemaniacs, who think Google's stock is a must buy at any price, to drive up the initial price rather than forcing them to wait until after trading starts. The price could then fall once smart investors can sell the overpriced stock short, a promise to deliver (they hope cheaper) shares some time in the future.

Could the IPO auction lead to an underpricing? In theory it could, but only if not enough people took part in the auction. For example, if everyone followed my advice (highly unlikely) or if people weren't aware of the IPO, then the paucity of bidders could produce an artificially low price. But given the buzz around this IPO (Googling the phrase "Google IPO" turns up more than 450,000 references), it is folly to imagine that it will fail to attract attention.

You could also make money if somehow you had a better sense than others what the right price should be. But even if you've studied Google's prospectus, watched the founders' video and looked at all of the business risks and opportunities, do you really think you have any better insight than the gaggle of other bidders you'll be competing against in the auction?

Okay, you might be smarter than people who haven't done any homework and who bid at the top of the price range. But that won't help. Their negligence will lead to an overpriced stock and rob you of any opportunity to make money. If you are honest with yourself, you'll accept the fact that there are professionals out there who know more than you do. Later, once they set the price, you can tag along and buy some shares.

Behind my recommendation is a paradox of the IPO auction. When IPOs are underpriced, investors are rewarded for researching and investing in new issues. When IPOs are fairly priced, there is no reason for investors to do all their homework. But if no one does his homework, then how does the market get the right price?

If the price is right, there's no point in playing the game. But if no one plays, then the price won't be right.

This paradox takes us back to where we started: Is the IPO auction really good for investors? If Google had chosen to play the traditional IPO pricing game and sell its shares at a discount, Google would have received less money. Most of the discounted shares would go to the (evil) investment bankers, who would dispense them as rewards (a k a kickbacks) for customers who pay inordinately high fees for other services. But some of the discount would make its way to average investors who are able to get a small allocation of the hot stock. It's certainly not fair, but at least investors would be rewarded for paying attention. With the IPO auction, the playing field is level, but no one makes any money. Except Google.

Page, Brin and their CEO, Eric Schmidt, are super smart enough to anticipate this. Being nice people, they may reward those folks who go through the effort to participate in the IPO. It turns out that Google has the right to purposely sell its shares below the market price. That's right, they can choose to give away money. And why not? If I had a few billion dollars, I might want to thank those who made it possible by spreading a bit of the wealth.

Here's how it might work. If a price of $125 a share balances supply with demand, Google could decide to sell its shares at, say, $110. That would give initial investors a head start in earning a return. But then demand would outstrip supply and Google would have to ration shares. You wouldn't get all of the shares you bid for.

There is a history of underwriters making gifts to initial investors. W.R. Hambrecht, the firm that pioneered the OpenIPO auction and has done 10 over the past five years, has often set the IPO price just below the market-clearing number.

The question is whether you can count on this gift and, if so, how much can you count on? You can't know because Page, Brin and Schmidt can't tell you. Their gift is only worth anything if it's a surprise. Imagine that you thought that the right price for Google was $130 a share, but you also knew that Google was going to set the IPO price at 10 percent below the market-clearing price. Anticipating the bonus, you could then safely bid all the way up to $144.40 and know that you won't end up paying more than $130. In short, the new market price will factor in the discount and there won't be any bump after all.

If that all sounds complex, what about waiting? The historical evidence suggests that waiting a day or even three years might not be a bad call. A study done by finance professors Ivo Welch and Jay Ritter (an adviser to Google) has shown that most IPOs perform worse than the overall market. The professors looked at what happens starting six months after an IPO and continuing for three years after that. The average IPO lags the market averages by 10 percent. The professors' explanation: The worst IPOs are too aggressive in their forecasts and fail to meet the targets. Such stocks end up trailing the broad market by 30 percent. The best IPOs, those in the top quarter, are conservative and end up matching the market. If the best you can do is match the market, why bother?

And what if Google is the next Microsoft or Dell? Both stocks did remarkably well during their first three years. They each rose by about 250 percent. But those who bought shares three years after those firms went public didn't miss the boat; they would still have made 80 times their money with Microsoft and 100 times their money with Dell. Oh, by the way, Microsoft closed its first day (March 13, 1986) at $28, up from the underwriter's price of $21. (Adjusted for stock splits, that closing price would equal 9.7 cents a share today.) Even then, the ratio of Microsoft's stock price to its earnings per share (P/E), a common measure of stock values, was only 21, just moderately higher than most U.S. companies and justified by its fast rate of growth. Google's last 12 months of earnings were 72 cents a share (taking account of option grants). If the IPO ends up in the middle of the forecast range, say $121, that's a P/E of nearly 170. That doesn't mean Google isn't worth it. But it does mean there's no secret treasure here.

I truly love Google and I'm anxiously awaiting my chance to use Gmail, the company's experiment with e-mail that never needs to be erased. It's a brilliant company and Page, Brin and Schmidt deserve every bit of their success. I applaud their philosophy of breaking free of the investment bankers' evil IPO pricing game. I wouldn't even be surprised if they give their IPO bidders a surprise discount. Still, I'm going to wait at least a day before jumping in.

Author's e-mail:


Barry Nalebuff, the Milton Steinbach Professor at Yale School of Management, is an expert on game theory and pops up 13,000 times on Google.