Robert Arnott, editor of the Financial Analysts Journal, is circulating a paper that may one day be credited with launching the next generation of index funds. Mr. Arnott, a prolific researcher, manages money for major pension plans as well as the PIMCO All Asset fund. With co-researchers Jason Hsu and Philip Moore at his firm, Research Affiliates, Arnott has found that non-capitalization weighted index portfolios have provided returns that are 200 basis points better than traditional index funds.

Since traditional index funds--- such as the funds that try to duplicate the performance of the Standard & Poor's 500 Index--- have regularly provided higher returns than 70 percent of all actively managed funds, any improvement is enough to get your attention. But beating the index by 200 basis points is the stuff of legends. Of the 330 domestic equity funds with 15-year track records only 17 have beaten the S&P 500 index by 200 basis points over that period.

We're likely to see a new generation of index funds rather quickly.  

The primary complaint about traditional capitalization weighted index funds is that they aren't portfolios with true diversification. (Capitalization weighted means that each stock is represented in proportion to its total market value. Total market value, in turn, is figured by multiplying the number of shares outstanding by the market price.)

Instead of broadly diversified portfolios, traditional index funds have highly concentrated portfolios. A few stocks will account for most of the value (and action) in the portfolio. Recently, the ten largest stocks in the S&P 500 Index accounted for 22 percent of its total market value. The 50 largest stocks accounted for more than 50 percent. A single mega-stock like General Electric or Microsoft carries more weight than hundreds of smaller stocks in the index.

One alternative: an equal-weighted index. Earlier research shows that an equal weighted index will do better than a capitalization-weighted index. You can observe it, in real money, in the Rydex S&P Index Equal Weight Exchange Traded Fund (ticker: RSP). Launched in April 2003, the fund now has over $400 million in assets. It has beaten its namesake index by over 500 basis points in the 12 months ending September 31.

The limitation of an equal-weighted index fund, Mr. Arnott points out, is that it's a lot more difficult to buy or sell the smallest stock in the index than the largest stock. Basically, it's difficult to scale-up an equal-weighted index fund. In addition, transaction costs are likely to be much larger.

Mr. Arnott proposes what he calls "Fundamental" or "Main Street" indexation. Rather than sorting stocks by market capitalization--- a favorite Wall Street measure--- he proposes building portfolios based on fundamental measures: book value, trailing five year average operating income, trailing five year average revenue, trailing five year average sales, trailing five year average gross dividend, or total employment. In addition, he constructed a composite portfolio combining several of these measures.

The result?

Over the 42-year period from 1962 to 2003 the S&P 500 Index returned 10.52 percent annually. A capitalization weighted index of 1000 stocks returned slightly less, 10.30 percent. But the "Main Street" indexed portfolios returned an average of 12.43 percent. Main Street indexing, in other words, beat capitalization weighted indexing of the same stocks by an average of 2.13 percent annually while beating the S&P 500 Index by 1.91 percent annually.

What causes the difference?

Mr. Arnott believes that capitalization weighted indices systematically overexpose stocks that are overvalued while underexposing stocks that are undervalued. An equal weighted index, like the Rydex ETF mentioned earlier, overcomes this systematic performance drag by eliminating capitalization weighting. Using any of the Main Street index building tools, however, overcomes the systematic performance drag while allowing the construction of more liquid portfolios.

Over a 25-year period, $10,000 invested at 10 percent will grow to $108,000. Invested at 12 percent it will grow to $170,000. This is a big deal.    On the web:

Earlier columns mentioning Robert Arnott:

Tuesday, April 2, 2002: Is A Pension Plan Time Bomb Ticking?

Sunday, May 16, 2004: Flip Side of Rising Interest

Robert Arnott's biography

Search for listing of Arnott articles