company had one miUion shares to sell or, having split, had five million shares. The investor would still pay $10 a share for his hundred shares, even though he now received only a fifth of the presplit profits. It was imma[terial to these speculators, Shiller found, because they were convinced the price would go higher. Is this irrational? Of course it is. It borders on the pathological. No sane person would pay five times the price a house was valued at a few minutes earlier. But people look at IPOs differently. Again, they are heavily influenced, if not drugged, by the powerful impact of investor psychology.

The results have been disastrous to investors. In 1994, Professor J. Ritter and Tim Loughran completed the most comprehensive study of new issues made to date. The study followed the retums of 4,753 IPOs traded on the New York Stock Exchange, the AMEX, and Nasdaq between 1970 and 1990. The average retum for IPOs was 3% annually compared to 11.3% for the S&P 500. Put another way, investing $10,000 in the S&P 500 over these 20 years would have retumed 751 %, vs. 81% for the IPOs. The far safer stocks of the S&P did almost ten times as well.

But this wasnt the worst of it. The median retum for these almost 5,000 initial public offerings was down 39%. Thats right. If you couldnt get that handful of red-hot IPOs that doubled or even tripled on the first trade-and nobody but the largest money managers, mutual funds, or other major investors could-then youd lose a good chunk of your original investment. You have less chance of making money in the new issue market-and millions of investors play this game -than by betting a single number in roulette.

An earlier study by Professor Ritter showed how trendy new issues were.° For example, most new issues go public near the top of an IPO market, when the demand is the greatest and the value of the merchandise is at its lowest. Fully 61% of IPOs went public in 1983, the absolute peak of the 1977 to 1983 mania. How many went pubhc in the first five years when the quality was at its best? Try 6%. The study also showed that the hottest concept industries at the time subsequently performed horribly. After the energy crisis in the late 1970s and early 1980s, superheated oil and gas stocks retumed a horrific - 43.9%, vs. 34.7% for the benchmark of established energy companies.

Other studies support the authors results. H. Nejat Seyhun reported that the market beat a sample of 2,298 IPOs for 6 years, while Mario Levis showed that a group of British IPOs unde erformed the U.K. averages for 3 years. Further research indicated that IPO fundamentals fell after the offerings, indicating a deteriorating business picture precisely when investors were most excited by the stock.

Did any IPOs outperform? Yes, value stoclcs were the anomalies, consistendy beating the maricet according to the earlier Ritter study.

Loughran and Ritter conclude: "Our evidence is consistent with a maricet where firms take advantage of transitory windows of opportunity by issuing equity when, on average, they are substantially overvalued."* The authors might have added that the evidence does not prove that the market is efficient but that stock hucksters certainly are.

Forbes, in a December 1985 cover story, entided "Why New Issues Are Lousy Investments," came out with similar results. The magazine traced the performance of 2,800 new IPOs between January 1, 1975, and June 30, 1984. After eliminating the penny stocks (whose records are normally far worse) as did Loughran and Ritter, the study measured 1,922 IPOs. The IPOs raised $27.5 billion, of which $2 billion went to the investment bankers. Forbes calculated the average appreciation over the period was 3% annually, vs. 14.8% for the S&P 500. The 3% a year is roughly in line with the Loughran-Ritter study.

The Forbes study also found that only 4 of 10 stocks went up at all in the 10 years measured. Among the new issues, 1 in 25 ended in bankruptcy, another 1 in 8 was down 95% or more, and almost 50% were down at least 50% versus the averages. According to the Forbes study, technology was, not su risingly, one of die hottest areas. Four big IPO underwriting houses specializing in this area came to be Icnown as the "four horsemen" in 1982/1983, near the last IPO market peak. These houses dished out a multitude of new issues in biotechnology, personal computer companies, software, disk drive companies, and the like. At the time, the issues were so hot that only the largest and best paying chents (in commission dollars for other business) could hope to get even a small fraction of what they wanted. How did these issues, underwritten by the very best of the technology houses, do? The batting average wouldnt make it in Class baseball.

Alex Brown, one of the most respected technology underwriters, had only 26 of its 82 issues outperform the S&P 500 over the ten-year life of the study. Hambrecht & Quist, another widely respected technology house, showed that 25 of 94 outperformed. L&F Rothschild, Under-berg, Towbin had 21 of 88; Robertson, Coleman & Stephens had only out of 45. Remember these issues were sizzlers at the time. Most people could not buy them at the offering prices, but had to pay substantial premiums, making the results far worse. Although a large part of the Forbes sample consisted of technology IPOs, less than 20% of the top 100 performers came from this sector. Forbes concluded, "Technology, the most glamorous field of all, proved to be a trap for investors."

Here We Go Again

I thought things would cool off for a while after the devastating drop following the 1983 bubble. I was wrong. No matter how much we study the subject, we are shocked anew at the power of "the image" in markets. Even Mackay and Le Bon would have gasped to see how quickly the crowd, after absorbing gut-wrenching setbacks, came back to follow near-identical images.

At this writing, as described above, we are in the largest IPO frenzy on record-one that dwarfs 1983s in size and lunacy. In 1983, $12.5 billion was underwritten in new issues, 70% or 80% of which was lost over the next several years. By 1993, the public poured $34.2 billion into new issues, almost triple the dollar amount they had gambled in 1983, the previous peak year. By 1996, the figure had reached $48.8 billion, dwarfing any IPO mania in the past. Stephen Woodsum, a managing partner of Summit Partners, a Boston-based venture capital firm, marveled, "I dont think we have ever seen a sustained period in which the IPO market was this good."* Marvel he might. From 1990 to the end of 1995, the S&P new issue index was up 538% against a rise of 84% for the S&P 500. The gains were again in the realm of fantasy.

Were investors buying new and different concepts this time around? Heck no, the focus was on almost the identical stocks purchased in the last market. Technology, be it computer or medical, satellite or space stocks, was hotter than ever, as were biotech and fast-food stocks. HMOs and health-care companies also got a big play, as did dozens of other exciting ideas.

Billions have been made in IPOs having the least whiff of Internet. Speculators salivated over the Intemet much as the French did over the Mississippi Territory, or Eve over the apple. Netscape, a developer of a Web browser, went pubhc in August 1995 at $28 and shot up to $175,

It did not require a crystal ball to see the end of the mania of the eighties coming. When it came the results were devastating. Technology stocks were decimated, and one of the four horsemen went out of business,- while the others pared back their operations drastically-for a time. Biotechnology stocks suffered the same fate. The P/E multiples on new IPOs were chopped in half between mid-1983 and mid-1985, declining from 31.8 to 16.5.2* The 1983 to 1985 IPO plunge was the worst on record. It occurred even as the largest bull market of the century made sweeping gains, loosening even further the barrier between quality equities and pure speculation.