Last year it was eToys Inc. This year it's Palm Inc. A hyped-up tech company goes public, and its stock roars, doubling or tripling in hours. Then comes the fall back to earth. eToys soared from $20 to $80 in a day, but now it's dragging along at about $12. Palm zoomed to $160 one day and plunged to $80 the next. It's the IPO roller coaster. But for people who buy shares at the peak, the gut-wrenching slide makes them see a different kind of green.
One-day wonders such as eToys and Palm are hardly the exceptions, so it's time to stop worshiping the "hot IPO." Instead we should scorn it--then fix it. Academic studies--and BUSINESS WEEK's own analysis of the subsequent performance of early 1999 IPOs--show that the hottest offerings are usually the worst performers after the initial thrill is gone. According to a study published last year in the Journal of Finance, IPOs from 1988 to 1995 that rose more than 60% on Day One performed even worse over the next year than IPO stocks that didn't rise at all that day. While there are no formal studies with more recent numbers, two of the three worst Web IPOs from early 1999--Value America Inc. and Musicmaker.com--were first-day blowouts.
SECRET. What's wrong? The IPO process is broken. American securities markets work brilliantly because they share information so broadly and so fast, except when it comes to IPOs. With them, the most critical information--how much demand there is for a specific new issue at different possible prices--is a secret that only investment banks and favored clients are in on.
What's worse, underwriters set artificially low offering prices even when they know investors would buy many more shares than are available. They do it to please institutional clients, who buy at the offering price and then sell into the initial runup. "The underwriters are throwing money off a tall building to their best customers," says Dartmouth College business professor Kent L. Womack, who co-authored the Journal of Finance study. One of the consequences is that underpriced offerings take off too fast, the stocks reach peaks that are unsustainable--and their potential for long-term appreciation is nipped in the bud. In the worst cases, such as eToys, the stock sinks even below the offering price through a combination of IPO mishandling and diminished faith in the company's prospects.
WIDE OPEN. The solution is to use the Internet--an amazing mechanism for spreading information fast--to make the IPO market open, fair, and efficient. What is needed is a wholesale transformation of the underwriting system for IPOs. Picture this: In the days before the IPO, information about the advance orders placed by other investors is continuously updated on the Web. If an institution tells an underwriter it wants 100,000 shares at Price A but only 5,000 at Price B, the whole market can see exactly where the smart money will stop buying.
A system like this would let people know what kinds of investors are demanding shares, although it would not give their names. Even aggregate information could signal the direction of the stock. If a large number of initial orders is placed by investors known for quickly selling stocks and taking profits, others would know to be wary of the offering. Those investors would probably bail out fast, and the stock could have a classic up-and-down spike.
Granted, this is a radical solution. And it won't end market volatility or protect people from making investment mistakes. But using the Web to open up the IPO trading process will shine a light on deals that are being mishandled and warn investors off. At a minimum, it would mean that people who chase IPO stocks right up through the roof have enough information to weigh the risks.
Reform like this won't come easily. There's no big incentive for investment banks to change when the IPO market is going gangbusters. And the Securities & Exchange Commission would have to revise rules that impede change. But pressure is building for something to be done. Plaintiffs' lawyers are sniffing around the IPO market for securities fraud cases. "The IPO rockets of today are the bombs of tomorrow," says Samuel H. Rudman of Milberg Weiss Bershad Hynes & Lerach. "When that happens, we have a lot of business."
Underwriters agree that the IPO system has problems--but they insist it's not their fault. They say the big culprits are investors who rashly bid prices up to unsustainable levels. Michael Evans, co-head of U.S. equity capital markets at Goldman, Sachs & Co., says it's hard to predict how high an IPO will shoot or how long the spike will last. And William Brady, managing director at Credit Suisse First Boston, acknowledges that some investment bankers unfairly reward institutional investors at the expense of small fry--but he insists CSFB doesn't. "We've always been very aggressive about pushing up the [offering] price. That way, you shake out the flippers," he says.
But there's plenty of evidence that underwriters deserve some of the blame. Dozens of hot early-1999 IPOs are trading well below their first-day highs even as specific companies' fundamentals are improving. Some of the worst performers are Web content companies and e-tailers that are among the best companies in their markets. Women's site iVillage.com hit $109 its opening week last March and fell to $31.25 by June. Auctioneer priceline.com quickly reached $165 after its IPO, then dropped to the 50s five months later. Martha Stewart Living Omnimedia, priced at $18 last October, spiked to $49.50 but was back to $35.43 before Day 1 was even over. It's now about $25. In most cases, their stock price didn't fall on poor business performance. It happened, in part, because the way the underwriters set prices helped send the stocks too high in the first place.
NEW TOOL. IPOs spike up and down that way because some investors have the best information about the market while others have to guess. It's time to level the playing field. Even though the SEC requires companies that are going public to disclose all material facts about their business fundamentals, it doesn't make their underwriters divulge anything about the demand for IPO stock or the prices that institutional investors are willing to pay for it. There's something fundamentally wrong with that. In the past, there was a reasonable explanation for the SEC's way of doing things: There existed no easy way of getting timely stock supply-and-demand information into the hands of investors. Now there is: the Web.
Some investment banks recognize that the Web can help--and are trying to do something about it. Goldman Sachs says it has been working for six months on a system to disseminate more information about orders before a deal. W.R. Hambrecht & Co., the small investment bank that attracted attention for running IPOs on a modified auction system, says a coming version of its system will let bidders see summaries of other investors' bids for the first time. These changes offer the prospect of a better deal for small investors.
For a breakthrough, though, the banks say they need more flexibility from the SEC. In particular, they want to be able to react to all the information they gather by being able to change the price of a deal faster and more easily than they can now. SEC rules require issuers to make a formal filing when they change the price, which delays a deal for at least two days. Often, underwriters feel that they can't afford to wait. The result: They price the offering too low. There's no reason why updated prices cannot be disseminated to all investors, instantly, on a Web site and through e-mail to people who already have ordered shares.
A solution like this would be good for investors and good for companies going public. Martin P. Dunn, the SEC's associate director for corporation finance, says the commission is examining reform but will go slowly to avoid encouraging even more speculation. The feds should think boldly. Ever since the 1930s, securities laws have been based on the firm belief that information is power and free markets can work out the rest. More of the same will help end those sickening IPO roller-coaster rides.