As the old saying goes, there's many a slip between the cup and the lip. When the research-proven notion of Value, which is simply that the price of a stock is cheap based on the ration of the stock's market price to the company's book value (p/b)(book value is simply assets minus liabilities in the financial statements), the meaning of Value gets transformed and expanded by index providers. Index-based ETFs need to have a reference index and they take them from such providers as Morningstar, Russell, MSCI, Standard & Poors and Dow Jones/Wilshire.
The best explanation of the way Value, and other indexes such as large vs small cap are constructed, is here at the Moneychimp.
So what do we find that the index providers have added to the original p/b measure? First, there are additional historical price measures like price/earnings, price/sales and dividend payout ratio. That might not be so bad as the basic principle that the stock is somehow low-priced compared to a fundamental historical measure is still respected. But where the definition of value really starts to take on hocus-pocus falsity is those indexes that include forecasted of such numbers as earnings! What lunacy! That replaces the best guess, as expressed in the price, of the market, whose guess is better, as shown time and again by research, with the guesses of analysts who are more wrong than right,
How do the index providers stack up?
- Russell - "...the Price-to-book ratio and the I/B/E/S forecast long-term growth mean"; affects IWM, IWW
- MSCI - p/b, dividend yield and 12-months forward earnings /price; affects VBR, VTV
- Morningstar - 50% weighting on forecasted estimates; affects JKL
- Dow Jones/Wilshire - uses p/b and p/ projected earnings (see the Moneychimp link above); affects XCV
- Standard & Poors - uses p/b, p/cash flow, p/sales and dividend yield; affects IJS, IJR, RZV