How To Do an IPO
In purely financial terms, yesterday's initial public offering from Linux
networker Andover.net will go down as just another in a series of remarkable
first-day debuts. The company's shares priced at $18, opened at close to $50,
and closed the day at $63 3/8. (Andover.net's shares jumped another 14 1/8
points today.) In other words, it looked just like a typical tech IPO in this
frothiest of markets. But in fact there was nothing typical about the success
of the Andover.net IPO. This was a deal the ramifications of which we're going
to be feeling for years to come.
That's because Andover.net was the third company to be taken public by W.R.
Hambrecht, the investment bank that is attempting to transform the IPO process
by replacing the traditional method of allocating shares and setting opening
prices with what's called a Dutch auction. (See
Slate
's "Explainer"
for more on Dutch auctions.) In a typical IPO, the lead underwriter sets an
initial filing price, then takes provisional orders from institutional
investors that it uses to gauge demand. The filing price can be adjusted upward
if demand is strong enough, but in general the offering price for a company
going public is considerably below the market-clearing price. As a result,
investors who are able to get in on an IPO have a very good chance of reaping
some easy gains.
In a Dutch auction, by contrast, the price is essentially determined by the
investors, who submit the highest price they're willing to pay and the number
of shares they want at that price. The people who bid the highest (and, if they
bid the same price, the earliest) get first dibs on the shares, which are then
allocated in order from highest to lowest bid, until they're all gone. No
matter what you bid, though, the price you pay is the lowest price that any
investor who got shares bid.
This is a superior way of pricing an IPO because no one gets shares on the
basis of who they know, and because it ensures that the company going public
isn't going to leave too much money on the table by going public at a lower
price than the one the market was willing to pay. It's also superior because it
eliminates the middleman (the investment bank) and lets the market set the
price.
Given the more general move in business toward disintermediation, it seems
inevitable that eventually this will be the way most companies go public. But
Hambrecht's Dutch-auction strategy has gotten off to a slow start because its
first two IPOs, for Ravenswood Winery and Salon.com, were relative busts, at
least in the sense that neither company has seen a significant gain over its
offering price. If investors feel they can't make money investing in IPOs
underwritten by Hambrecht, they'll pay less for them then they would if the
same companies were being underwritten by more established investment banks,
thereby defeating the whole purpose of going public with Hambrecht.
That's why Andover.net's awesome debut, with its price soaring 252 percent
on the opening day, is so important. It suggests, as I argued at
the time of Salon.com's IPO, that the problem with the earlier Hambrecht
offerings was not the Dutch-auction idea, but rather the quality of the
companies that were going public. Take a real company in a hot sector, like
Andover.net, and investors can make a lot of money, even as the company itself
raises more cash in the IPO than it would have done otherwise.
There is, to be sure, something of a paradox here, since if the point of
having a Dutch auction is for the company not to leave any money on the table,
the 252 percent jump in Andover.net's stock price might seem to suggest that
the auction didn't work all that well. But there are two things to keep in mind
here. The first, and more mundane, point is that you have to have a brokerage
account with Hambrecht to participate in its IPOs, so the number of potential
investors in any Hambrecht offering is significantly smaller than the number of
investors who might be interested in that offering. (Of course, the more IPOs
like yesterday's Hambrecht has, the more customers it'll get, since anyone who
has $2,000 can open an account.) The second, and related, point is that no
company can control the retail market for its stock. The market-clearing price
at 10 a.m., in other words, will not be the market-clearing price at noon or at
3 p.m., and that's true for all stocks, not just IPOs. Now, as investors become
more comfortable with Dutch auctions, and as more investors take part in these
kinds of IPOs, the spread between the offering price and the first day's
closing price will narrow, and that will be a good thing. But for now, the 252
percent leap Andover.net took is very good news, the first significant blow
against Wall Street's old-school IPO process.
