There appear to be three main theories on how to create style indexes. Some indexers, like S&P, argue that book value is the most important determinant, while others (like Dow Jones/Wilshire) focus on earnings growth. Still others (MSCI, Intellidex and Russell) stand somewhere in the middle. Where an index stands along this spectrum will play a key role in what kind of performance these indexes provide in particular market situations.
In the example given above, the Morningstar index posted the best returns among traditional indexes, while the Dow Jones Total Market index lagged. But under different market conditions, those returns clearly could be reversed.
There are mathematical formulae (quadratic optimization from William Sharpe is the most popular) that can provide the exact "style" score analysis of an index or a fund manager. But ultimately, there is no one "right" answer: Determining what style analysis is best depends on the needs and desires of each individual investor or each individual fund.
Addendums
--One Man's Indexing Idea: Absolute Style Indexes--
The seven style indexes covering the United States differ markedly from one another. One thing they all share in common, however, is the assumption that, on any given day, half of the stocks on the market are "growth stocks" and half are "value stocks." In all of the systems, the total market capitalization of the "growth" and "value" sides must be equal for each capitalization category.
There are solid philosophical reasons for this assumption, but from the author's perspective, the absence of an "absolute" style index is glaring. It is clear from recent history that the valuation of the market varies widely from the historical mean. For instance, the P/E ratio of the S&P 500 Index neared 30 during the Internet bubble, whereas its long-term historical average stands at about 16.
Rather than dividing the market equally into growth and value, what if an index took an absolute approach towards defining style? What if, using the example of the S&P 500, an "absolute large cap value index" was created that included all of the companies with a P/E ratio of less than 16. (Similarly, one could use book value, or some combination of the two).
This index would allow us to track the long-term performance of "true value investing" - buying stocks that are cheap in the long view - instead of showing the relative performance of growth and value stocks. It might even attract a few investors of its own along the way.
