a pricing method established over a century ago. Trouble is, the stocks in the average have been changed over the years, based on the decisions of a small group working within the Dow Jones organization. Little is known about how these people pick stocks for the average, yet their choices can significantly change the level of this index.

On March 14,1939, one particular stock was dropped from the index. Why? It was "just another dull office equipment supplier." The name of the company was IBM. AT&T was substituted. You know the rest. With stock dividends, splits, and rights offerings, IBM moved up 635.2 times over the next four decades before die keepers of the gate saw fit to restore it to the index in June of 1979. Its replacement, AT&T, moved up 6 times in the same period.

Had IBM been kept throughout, the Dow would have been around 9,000 in 1997 (nearer to 15,000 if the appreciation were not adjusted downward after each stock split, according to the formula, to equally weight the average).

You see, in a sense, the Dow keepers are just stock pickers-and just as fallible. When the time came to restore IBM, something had to be dropped. It was Chrysler, then barely breathing. Kicked out of the index, Chrysler staged a sensational comeback. Since then, Chrysler has appreciated 1324% to IBMs 82%, or 16 times as much. Chalk up a bunch more points, had IBM not been added back. Never underestimate the ability of a committee to arrive at the wrong decision.

Okay, how about the S&P 500? Similar pattem. In the late sixties, for example, the REITs (real estate investment tmsts) were put in very near their all time highs, only to be dropped in 1974 at a fraction of tiiose prices near their lows. Gambling stocks were added in 1979 and oil service companies in 1981-both near their market peaks-which they never regained. And so it goes.

The makeup of many indexes, then, simply reflects the conventional stock-picker wisdom of a given time. Investors believe stock market indexes mirror markets completely insulated from popular fashions. Not tme. Because stock selection or deletion is an art, tiie index is not a mirror but more like a painting that differs, sometimes dramatically, from the reality it attempts to portray. This subjective element in stock indexes is something most people, including experts, are unaware of. In the end, ironically, index investors get exactiy what they attempted to avoid-human ertor.

Since there are many indexes, some are performing much better tiian others at any given time. Small stock indexes were popular well after tiie small cap markets tumed south in mid-1983. Over ten years to the early 1980s, the Value Line Index (which is a long-standing record of 1,700

Tailor-Made Performance Records

I dont want to make you totally paranoid about indexes and performance records, but I dont think it would be completely honest to hold back on two categories of bogus performance that can do you in.

The first is supercharged results. These can come out of a mutual fund, and are unlikely to be maintained. The point is so important that I think it ought to be made into a conti-arian rule:

RULE 40

Avoid the small, fast-track mutual funds. The track often ends at the bottom of a cliff.

equally weighted stoclcs) did sensationally, but only because it recovered from the sha drop in the prior ten.

You should be extra careful of specialty indexes put out by investment houses or other parties that have a vested interest in demonstrating excellent performance for a particular group of stocks. One of the leading technology houses, for example, puts out an index of high-tech stocks that outperformed the market manyfold since its inception in the early 1970s. The brokerage house uses a rather unique index-selection formula. When stocks it originally held in the index dont perform well, it drops them retroactively, sometimes ten years after the fact, and substitutes the major high-tech winners of the day instead. Naturally it puts them into the index at the prices they were trading at on the dates they were retroactively selected, which results in instant performance. The excellent record of its index indicates the firm obviously knows its stuff, and helps the company to bring in substantial business. I wrote a Forbes column on November 3, 1986, on this somewhat out-of-the-ordinary stock selection process. We checked as of mid-1997, and found that the practices had changed little.

In the end then you are not getting an index based on the performance of growth stocks of the seventies, eighties, or nineties, but one that is built using the luxury of hindsight. If we could botde this formula, wed all be rich.

The moral is clear. Steer away from these sweeping generalizations based on flimsy evidence, wherever they occur. This is particularly true of indexes that can be self-serving to the people who construct them. Which brings us, you will not be su rised to learn, to another widespread performance deception.

Forbess "1984 Annual Mutual Fund Survey" (Aug. 27) summed it up well: "A tiny fund scorches the track, brings in big money and sends late investors into the tank." Nothing has changed since that time; in fact its become worse. By late 1997, the sizzling fast-track funds of the nineties were falUng behind, with some of the former outstanding performers down 20% to 25% for the year alone.

The better the performance in a market where fast-track stocks are shooting out the lights, the more advertising the mutual fund sponsor pours into the fund (often a virtual new entry with a short but spectacular record), and the more publicity is generated in the press. Inevitably, most investors get in near the top-just in time to get blasted away. Buying fast-track funds is like paying full price for last years fashion just before the new lines are shown.

Here are some ways to avoid this shuffle. The first is not to buy a fund of $50 million if its retums have been made by some fast-track method. The record at $50 million usually doesnt continue as the fund grows into hundreds of millions or billions. Thats tme especially if a good part of the increase has come from spectacular gains at tiie beginning, when the small fund had only a minute part of present assets under management.

Second, go beyond the record itself and find out how it was achieved. If you are uncomfortable with what makes the fund tick, stay away-its likely to explode. This is also tme of a widely known practice currentiy being investigated by the SEC, of one or two funds in a fund family receiving most of the groups allotment of hot IPOs. Thats done in order to "spike" their performance. Again these are usually smaller funds in trendy industries that can attract big bucks if the record is sizzling. Allocating major portions of hot IPOs to them gives them instant performance. Nobody much looks at how tiie fund did it. The answer once more is close scmtiny of any small fund you buy.

Finally lets glance briefly at the record of the investment advisor. A good friend of mine, the late Ted Halligan, spent most of his career placing money managers with clients. "I never met a manager who didnt say he was in the top percentile in performance," Halligan once told me wryly.

So how do you separate the claims from the tmtii? Getting accurate numbers of a managers performance over a long period is more difficult than you might think. Even sophisticated investors, who depend on consultants to do their digging, often get bad numbers. A former retail broker, who once lived near me, wiped out his original clientele in the early eighties playing hot, "junior oil" stocks, just before they went into the tank. With his commission gauge rapidly moving toward empty, he