On Sunday, I pointed out that a recent column by MSN Money contributing editor Jon D. Markman was wrong in its criticism of the S&P 500 Index. In today's column I want to add some historical perspective.

In today's world of power computing and indices for virtually everything, the Dow Jones Industrial Average seems pretty quaint.

In fact it was probably the best tool for measuring the broad performance of common stocks for many decades. More important, it was as much as the computational tools of the time could handle.

Introduced in 1896 as a list of 12 "smokestack" company stocks, the Dow was a reasonable proxy for the performance of big industrial companies. Add the prices of the twelve stocks, divide by the divisor, and you had the Dow--- something anyone could do with pen and pencil. If you used the most advanced computational tool of the time, the Monroe mechanical calculator, you could eliminate all human error except clumsiness with punch keys.

I tell you this for a reason. Our tools of financial measurement have always been limited by the available computing technology.

When the Dow Jones Industrial Average was expanded to its current 30 stocks in 1928 it was hailed as a broader, more representative measure of what was happening in the stock market.

Not perfect, mind you.

Just better--- and it could still be calculated on a Monroe Calculator.

Today, no one would claim that a 'managed' index of 30 stocks, such as the Dow Jones Industrial Average, was a reasonable proxy for stock market performance. The Dow is a legacy index with brand recognition. Period. Dow Jones, the company, has moved on--- today they have sophisticated indices that are far more representative of different asset classes.

Standard and Poor's introduced its S&P index as a broader competitor to the Dow--- but its original 90 stocks didn't make it that much broader than the Dow.

Standard and Poor's didn't introduce the S&P 500 Index until 1957. It was the first modern index. Significantly, 1957 was around the time the first commercial mainframe computer was installed at M.I.T. It was also the year Ken Olsen left M.I.T's Lincoln Labs to form Digital Equipment. And it was a year before Jack Kilby created the first integrated circuit at Texas Instruments. While it required a mainframe computer to do it back then, it was possible to continuously calculate large changing figures for something as large as the S&P 500 Index.

The Monroe mechanical calculator disappeared.

The first modern index did two things. First, it captured the change in market value of America's largest companies--- about 80 percent of all market capitalization. Second, it was structured as a portfolio. Some companies that were not among the 500 most valuable were included in the index.

Why? The added companies added diversification to the "portfolio." Remember, one purpose of the index was to serve as a performance measure for the institutional pension managers who invested most of the new money that went into the market. Those portfolio managers were usually limited to a universe of 200 to 300 major companies.

The only problem with the portfolio approach to indexing is that it made some additions and subtractions a matter of human judgment. Changes in the S&P 500 stocks have been second-guessed as long as I can remember.

Whatever limitation such choices represented, the next generation of indices didn't come along for 17 years.   The Wilshire 5000 Index, introduced in 1974, was created to capture the performance of the entire U.S. stock market. By 1974 you could do that on a mini-computer. Covering the entire market eliminated the dilemma of human choices.

The Russell Company entered the fray in 1979, introducing a family of indices. Russell 3000 captures 98 percent of all domestic market value. The Russell 1000 captures 92 percent of market value of the Russell 3000, or 90 percent of all market value. Unlike the S&P 500 index, size of market capitalization, alone, determines which Russell index a stock is in.

The limitation of all of these indices is that they are, of necessity, market capitalization weighted. This results in two flaws that portfolio professionals have complained about for decades:

•       First, the largest companies account for the bulk of the value in the indices. Today, for instance, the 18 largest S&P 500 Index stocks account for one-third of the value of the entire index; the 100 largest account for 70 percent.

•     Second, booms in particular industries can make the indices very 'unbalanced'. In the late 70's and early 80's, for instance, energy companies were a large proportion of the big cap indices. The same thing happened in the recent technology bubble. According to investment strategist Steve Leuthold, technology stocks accounted for 34 percent of the S&P 500 Index at the February 2000 peak. Today technology accounts for a slightly below normal 16.2 percent. Now financial stocks account for an oversized 20.5 percent of the S&P 500 Index.

Ironically, the fact that Standard& Poor's "manages" its index to include a diversified industry portfolio acts as a minor damper on the extremes of industry concentration.

While Jon Markman criticizes index changes at Standard and Poor's and others complain that neither the S&P 500 Index nor any other index represents a truly diversified portfolio, one fact remains. Major indexes are a low cost, low turnover, tax efficient way to invest.

The proof is in the performance figures.

Over the last 20 years the Vanguard 500 Index fund--- the oldest retail fund duplicating the "ham-fisted" management of S&P 500 Index--- has been in the top 20 percent of all large cap funds. It beat the average large cap managed fund by a whopping 1.86 percent annually according to Morningstar data.

Over the three and five years ending June 30--- the period Mr. Markman criticizes--- the mother of all index funds provided a higher return than 54 and 61 percent, respectively, of all comparable large cap funds. They did this in a declining market, the kind that gives cash holding managed funds a modest advantage.

The message?

Ham-fistedness is relative. Short-term or long-term, history shows that passive ham-fistedness is kid-glove treatment compared to the active ham-fisted variety.

The Jon D. Markman Columns Berating Standard and Poor's:

7/23/2003: S&P is lousy at make-or-break bingo

8/14/2002: Behind the curtain at Standard & Poor's

6/27/2002: The S&P 500 is a Mutual Fund--- and a Bad One at That

1896/1928: History of the Dow Jones Industrial Average

1957: Information on the S&P 500 Index

1974: Information on the Wilshire 5000 Index

1979: Information on the Russell Indices