What's The Right Way to Index?

The following article appeared in the October issue of Morningstar ETFInvestor, a monthly newsletter dedicated to making sensible use of exchange-traded funds. To learn more about Morningstar's newest publication, please click here.

Lately, a lively debate has sprung up in the indexing community over how stocks should be ranked in an index. That's not the sexiest of topics, but the debate picked up some public attention when it reached the editorial pages of The Wall Street Journal, where Jack Bogle, a respected indexing proponent and founder of Vanguard Group, squared off against Jeremy Siegel, a noted financial author and professor of finance at the Wharton School of Business of the University of Pennsylvania. Bogle argues for a continuation of the dominant practice in indexing--capitalization weighting, while Siegel contends that investors would be better off if the stocks in the indexes were ranked based on fundamental investment factors.

For years, stocks in most indexes have been ranked according to their market capitalization. (A stock's market capitalization is simply its stock price multiplied by its float, which is the number of shares freely available for trading.) So, for most index funds, the stock with the largest market capitalization in its universe is its largest holding.  ExxonMobil (XOM) is the largest U.S. corporation, as measured by market capitalization, so it has the number-one spot in the S&P 500 and the Russell 1000.

If indexes are supposed to represent the broad market, then market-cap weighting makes intuitive sense. If an imaginary and very wealthy investor set out to buy the entire U.S. stock market, he'd of necessity own the stocks in the proportion that they are represented in the Wilshire 5000, a cap-weighted index that represents almost the entire U.S. stock market.

But a poor showing by large-cap index funds during the bear market exposed a glitch in the system. In 1999, the cap-weighted S&P 500 was caught owning a large slug of pricey tech stocks that had seen their market caps skyrocket in the late 1990s. Consequently, index funds that track that ubiquitous benchmark underperformed their peers during the ensuing sell-off. That provoked charges that capitalization-weighted indexes systematically overemphasize overvalued stocks and give short shrift to undervalued names, leaving index funds vulnerable to bubbles when valuations get really out of whack. Over the past couple of years, alternatively weighted indexes have cropped up with the goal of sidestepping that problem.

Fundamental Weighting Methods
The most interesting of these are the indexes that weight stocks using fundamental factors. Perhaps the most prominent of these is the FTSE RAFI US 1000, an index that ranks stocks based on four fundamental factors: sales, cash flow, book value, and dividends. ETF investors can get exposure to this new index through  PowerShares FTSE RAFI US 1000 (PRF).

Additional alternatively weighted index funds that have really caught on with investors are those that rank stocks according to dividends.  iShares Dow Jones Select Dividend Index (DVY) was the first ETF to the market with this concept, and investors flooded it with cash. It now weighs in at a hefty $7.4 billion in assets. The popularity of that fund didn't escape the notice of other ETF providers, and it didn't take PowerShares, State Street, and Vanguard long to jump on the dividend bandwagon. What's more, an entire ETF complex has recently built up around the premise of dividend weighting: WisdomTree Investments rolled out a whole slate of dividend-weighted ETFs this summer.

These upstarts contend that their ranking systems are superior to capitalization-weighted indexes. Based on historical data, they all claim to outperform their cap-weighted counterparts by as much as 2 percentage points per year.

My Take
While those claims are eye-catching, I'm suspicious of back-tested results. No one ever got to their destination by looking in the rearview mirror. And the success of these indexes can at least partially be attributed to market trends that favored their inherent biases. Their fundamental screens favor value stocks, which have outperformed growth stocks for the past several years now. Furthermore, the RAFI index and a couple of the dividend indexes are also tilted toward the smaller stocks in their universes, which has benefited in recent years from the market's appetite for smaller names. Cap-weighted indexes, on the other hand, emphasize the largest stocks in their universes, and that's had negative implications for their relative performance lately.

The proponents of the new indexes claim that their systems should correct the cap-weighting problem of overweighting pricey stocks. But that doesn't necessarily mean that the fundamental indexes will look any better from a valuation perspective. In fact, they may very well introduce biases of their own. For example, most of the dividend-oriented indexes emphasize utility stocks because of their high yields, and that's a sector that's looking pricey at the moment. Furthermore, many of the smaller, value-oriented stocks that many of these indexes favor are looking overvalued after a lengthy rally.

What's more, the fundamental screens could cause these ETFs to miss opportunities. Because of their bias toward dividend payers, they could end up holding on to a tired industry giant with a fat yield and pass by a young upstart firm that is plowing all its cash into its growing business rather than issuing dividends. It's telling that the RAFI index dedicates 2.7% of assets to  General Motors (GM) and  Ford Motor  (F), while these stocks take up just 0.2% of the Russell 1000.

Furthermore, given all the hype that alternative indexes have gotten lately, the contrarian in me can't help but wonder if the time is ripening for cap-weighted indexes. I suspect market winds are beginning to shift in favor of mega-cap stocks, a development that would benefit large-cap cap-weighted indexes. True, academic studies have shown that smaller, value-oriented stocks have outperformed over very long stretches of time. However, you could construct a small-value-biased portfolio more cheaply than any of the fundamental ETFs by using a combination of cap-weighted ETFs.

This brings me to another issue I have with these funds--some of them are pricey. PowerShares FTSE RAFI U.S. 1000's expense ratio is contractually limited to 0.60%, but that fee cap comes courtesy of fee waivers that may or may not continue after the contract expires on April 7, 2007. Although the WisdomTree ETFs are more reasonably priced,  Vanguard Large Cap ETF (VV) still has a substantial head start with its miniscule 0.07% price tag.

Plus, these newfangled indexes could also face higher trading costs than traditional cap-weighted indexes. When the market cap of a holding in a traditional index changes, the market does the work for the fund; no trading is necessary to adjust the stock rankings. However, alternatively weighted indexes have to adjust their holdings when the fundamental factors change. Granted, the RAFI and the WisdomTree indexes make such adjustments only annually to try to keep trading to a minimum, but I suspect these ETFs will see a bit more turnover than a traditional cap-weighted counterpart.

Despite these criticisms, I'm reluctant to dismiss these funds outright. There's something to be said for keeping a focus on fundamentals, and these ETFs might make sense for certain investors. For example, an income-hungry investor might benefit from owning a dividend-oriented ETF with a generous payout. But in my opinion now's not the time to ditch your cap-weighted index funds and jump over to one of these newfangled offerings. Before I'm ready to recommend these funds, I'll need to be convinced that the portfolios they hold are attractive.