Everyone is always astounded when new IPOs begin trading at double or triple the offering price in the matter of a few minutes. "These are great companies," the collective gasp seems to say. Broadcast.com (BCST) a few weeks ago, GeoCities (GCTY) this week. Maybe they are great companies, maybe they are not. Because for every Amazon.com (AMZN) - Get Report and Yahoo! (YHOO) there are scores of General Magics (GMGC) , 3DOs (THDO) and Cybershops (CYSP) .

Early trading in IPOs is bogus. It's rigged, and it is a completely inefficient market. The reason is simple: If a company sells 2 million shares out of a total of 12 million outstanding after the deal, then less than 17% of the company's shares are freely tradable. The rest are locked up. Every investor, manager, director and employee who owns even a single share of the company's stock is asked to sign a lockup agreement, requiring the shareholder not to sell shares for 180 days after the offering. The threat is that if you don't sign, there is no deal, no IPO, no public shares to trade. Usually, everyone signs. I try like hell to not sign, but always, the call comes in, "Mr. Kessler, stop being a jerk and sign the damn thing, or no deal, etc."

Underwriters insist on a lockup so they can control the early trading in shares, and be able to carefully place those shares in the best hands. Remember, investment banks get 7% for underwriting IPOs. This onerous fee was intended to compensate for the risk. What risk? Underwriters actually own the shares overnight, they buy them at a 7% discount the night before and then sell them to investors the next morning. But in the markets of the 1990s, this has become a riskless transaction. Even if the IPO becomes really screwed up (like when the

Eagle Computer

CEO crashed his Ferrari), they cancel the deal and everyone gets their money back.

OK, back to trading. The lead investment banker, the one whose name is listed first on the cover of the prospectus, the one on the left (and known in the business as "on the left"), controls the trading. That firm's syndicate desk places shares with investors, mostly institutions, and then its trading desk attempts to make a natural and stable market for those placed shares. Many people known as "flippers" sell immediately, while others may be buying in the open market to increase their positions. It takes about 15 minutes to an hour for the price to reach an equilibrium between sellers and buyers.

It is hard to do, but the trader knows where all the shares are. He placed them with investors, so he moves them from one account to the next and can more or less keep track in order to set the equilibrium price. The last thing this trader wants is 5 million shares to come out of left field, from existing investors and company secretaries. Hence the lockup.

But c'mon. I'll give the guy 180 hours, but 180 days? By the end of the first week he has lost track of where the shares are, and can probably use some more shares to place with institutions hungry for 250,000 shares of a great company, instead of the measly 5,000 shares they got on the deal.

What annoys me is that the lockup makes an artificial market. The stock's price in the first 180 days will most likely not come close to reflecting reality. Sure, you can figure out a company's market capitalization by multiplying the stock price by the total shares outstanding, but the

trading

in the stock is artificial -- there may be lots of buyers, but not a lot of sellers because, as in the above example, only 17% of shares are trading.

Then comes the real annoying part. At day 181, the fire hose hits. Everyone and his brother is unlocked. If the stock is at lofty levels, they will blow out, or try to, all at the same time. It is uncommon for any stock to trade up when shares are unlocked. Shorts try to outguess the market and short stocks ahead of the unlock date, figuring they will buy distressed shares lower.

If I am an investor in a private company that goes public, I carry the shares at 80% of their public value until they are unlocked, the discount for lack of liquidity. But in the meantime, I am forced to play a guessing game of whether the stock is overvalued preunlock, or undervalued postunlock when I am finally allowed to sell (we usually don't sell, since we stick to long-run, three-to-five-year holdings). I can't listen to the market to tell me the companies' true value, because it has been a bogus market for six months.

If I bought a stock on the IPO, I have a similar problem. At 181 days, the stock often gets whacked. Should I sell it and incur a short-term gain, figuring it will go lower and I'll buy it back? Should I just ride it out and stomach the volatility? The best bet is to buy more if you have conviction in the fundamentals and the stock drops because of new supply.

Personally, I vote to do away with lockups, or stagger them, say, six 30-day unlocks. The stock market is the world's greatest voting mechanism, a market where buyers and sellers set fair prices every day, every hour, every minute, every tick of the tape. With IPOs, the first six months are rigged, and prices are not true economic values. Let's change this tradition.

Andy Kessler is a partner at Velocity Capital, and runs a technology and communications fund out of Palo Alto, Calif. This column is not meant as a solicitation for transactions; it is instead meant to provide insight into the methods of venture capital, technology and investing. Under no circumstances does the information in this column represent a recommendation to buy or sell stocks. Kessler appreciates your

feedback.