Wednesday, November 08, 2006

Worse than a Coin Flip

I was recently reading about the possible end of Bill Miller's win streak against the S&P500 Index. His Legg Mason Value Fund has beaten the S&P500 for an impressive 15 years in a row. However, this year he's trailing the index by about 10% at this point. The manager with the next longest streak is Manu Daftary, who's Quaker Strategic Growth Fund has beaten the index for eight years in a row. However, he's also trailing the S&P500 by almost 9% this year. After him, there's a handful of funds with a seven year streak.

It got me wondering how many funds you would expect to beat the S&P500 if the results were represented entirely by chance: flip a coin, heads you beat the market that year, tails you lose. There are approximately 8,600 mutual funds out there, so if the results was entirely by chance, half (4,300) would beat the market after one year. If half of those beat the market the next year, then 2,150 would have beaten the market for two years and so on. After eight years, there would be 67 who beat the market for eight years in a row. Instead, in real life there are only two, Miller and Daftary.

True, beating the S&P500 each year isn't the only goal of all mutual fund managers, but it's still interesting nonetheless to note that only two out of thousands of mutual funds--with managers paid millions of dollars a year to beat the market--have managed to beat the market for even eight years in a row. A coin flip would have done much better.

Tuesday, November 07, 2006

Fundamental Indexation: The Next Generation of Indexing?

Underweight overvalued stocks and overweight undervalued stocks. Sounds like a simple recipe for beating the markets. Essentially, that's the idea behind what some have called the next step in the evolution of indexing.

The argument is that traditional capitalization-weighted indexes (such as the S&P 500) by their very nature will overweight stocks that are overvalued since the amount of stock that they hold in a particular company is based on the market value of that company in relation to the rest of the companies in the index. So, for example, during the tech bubble, stocks with huge market valuations, such as Cisco, dominated the index. Basically by definition, if a stock is overvalued, it is overweighted in a cap-weighted index.

The idea behind fundamental indexation is to weight the holdings by some fundamental measure, such as sales, income, book value, or even number of employees, instead of by the value that the market places on a stock. By doing so, you reduce the amount by which you are overweighting the overvalued stocks and avoiding their drag on your returns. Rob Arnott, who introduced the idea to the mainstream investing public a year or two ago, found that such an index outperformed the regular cap-weighted index by about 2% annually on average over the last 40 years or so.

The idea has enough merit behind it that individuals such as Jeremy Siegel (my old Wharton professor and author of "Stocks for the Long Run") have gotten behind the idea. He has joined Wisdom Tree, a startup investment company that offers a number of ETFs based on fundamental indexing using dividends as the weighting critieria. In addition to Wisdom Tree's offerings, there are also offerings based on the Research Affiliates Fundamental 1000 Index, an index based on Arnott's research that is weighted by a composite of several fundamental criteria.

Even William Bernstein has called fundamental indexing a promising technique, though he cautions that the advantage over cap-weighted indexing is small and could be overwhelmed in practice by fees and transactional costs.

Though it's still a new idea, it's definitely one that merits further exploration, particularly as the offerings from Wisdom Tree and those affiliated with Research Affiliates develop a real-world track record that one can examine for the effects of fees and the costs of trading.