Is Fundamental Indexing All that It's Made Out to Be?

It's been one of the toughest years on record for equity investors, and there aren't too many signs that things might settle down soon. Still, now may be a good time for investors to cautiously wade into stocks: Many veteran market observers agree that valuations haven't generally been this attractive in quite some time. And for many, a cheap index fund may just be the right ticket. After all, along with the great bargains available right now, there are a good many land mines as well, which makes active stock selection a nerve-wracking business. For those contemplating a sound index vehicle, now is also a good time to catch up with a continuing debate between the two main flavors of indexing: traditional, market-cap-weighted indexing versus fundamental indexing.

Weighting stocks in an index by their market caps has been the norm for decades, and (as most index purists continue to argue) for good reason. A market-cap-weighted index automatically rebalances itself because it fluctuates in perfect tandem with market prices of its component stocks, so it is easy and cheap to maintain. Also, the largest, most liquid stocks get the biggest billing, which further reduces transaction costs. Low-cost market exposure is among the most important selling points of passive investing, so those are powerful advantages.

But the new paradigm of fundamental indexing, formalized in recent years by prominent investing professionals and academics such as Robert Arnott and Jeremy Siegel, contends that it has a better approach. Essentially, fundamental indexing seeks to correct what it views as inefficiencies in traditional market-cap weighting. The prescription is to weight stocks not by their market values, but by their "fundamental" characteristics such as earnings, cash flows, and dividends.

Meet the New Boss ...
The theoretical basis for using fundamental weights as opposed to market-cap weights rests on two main claims. The first is that stock prices tend to be "noisy"; that is, rather than represent fair equity value, they are on average as likely to be overvalued as undervalued. Then, there is a case for using alternatives to market prices such as fundamental weights. The second main theoretical claim, which states that fundamental weights behave independently of market values, provides additional support for using fundamental weights. This second claim (known in the parlance as the "independence assumption") is important because if the fundamental weights used are correlated with market values, they would be just as contaminated.

Here's how the argument is fleshed out. If the market incorrectly assigns high or low prices to stocks compared with their fair values, undervalued stocks will have low market caps and overvalued stocks will have high ones. In that case, using market-cap weights will result in systematic errors in representing stocks in an index. Eventually, when the mispricing is corrected, a market-cap-weighted index will suffer a performance drag. The alternative then is to use fundamental weights. And here, the independence assumption kicks in. Proponents of fundamental indexing claim that fundamental weights are not doomed to repeat the same errors as market-cap weights because stock characteristics like earnings and dividends can stand true in their own right and are statistically independent of noisy market prices.

... Same as the Old Boss
Not so fast, say the skeptics and those who retain faith in traditional indexing. There have been some particularly illuminating articles in the last couple of years that not only counter the key theoretical and empirical claims of fundamental indexing, but also attempt to clarify the fine points of passive investing in general. Let's look at some of those arguments in more detail. The key conclusion from these papers is that fundamental indexing is prone to errors of its own, and some of those errors in fact are related to the same sources that cause noise in stock prices.

Let's first look at the fundamental indexing claim that traditional market-cap-weighted indexes must suffer a performance drag because they systematically overweight overvalued stocks and underrepresent the bargain names. Andre Perold of Harvard Business School provides a strong counterargument by demonstrating that even if market prices deviate from stocks' true fair values, it does not automatically follow that cap-weighted indexes load up on overvalued stocks. That is because there is simply no reason to conclude at the outset that a stock that commands a high valuation is overpriced. It is just as likely that the "expensive" stock deserves its premium because of its superior growth prospects, or that a "cheap" stock's fair value could be even lower than its market price. Even if stock prices are noisy as claimed, there is still the problem of knowing whether a given stock is a bargain or not, and Perold shows that fundamental indexers assume this problem away.

Morningstar's own Paul Kaplan tackles the independence assumption that underpins fundamental indexing. Fundamental indexing claims that fundamental weights are statistically unbiased estimators of fair value, which means that while fundamental weights will often err, such deviations from fair value will on average be zero. Moreover, fundamental indexing claims these errors are not correlated with market prices. This "independence" assumption is key because without it fundamental weights cannot really be superior to market-cap weights. However, Kaplan shows that this assumption is hard to justify because it leads to some far-fetched conclusions.

Kaplan develops his argument by showing mathematically that errors contained in fundamental weights are related to stocks' fair value multiples. For example, if earnings are used as fundamental weights, then the errors produced by such weights will be related to fair price/earnings multiples. Then, fundamental weights can only be independent of market prices in one of two ways. The first way is if fair value multiples are independent of actual market-assigned price multiples. That doesn't make a whole lot of sense though. True, markets can often be overly buoyant, and price/earnings and other multiples can soar much above what's fairly deserved (or vice versa in times of extreme market pessimism). But to say that fair value multiples are completely independent of market multiples is a bit of stretch. Under most normal circumstances, the market likely assigns multiples to stocks in some proportion to the fair value multiples.

The only other way left for the independence assumption to hold is if fair value multiples are constant across all stocks. This, too, is unrealistic. After all, different stocks have different risk and growth expectations, which means that they deserve to trade at different multiples. Thus, viewed from this angle, fundamental indexing in effect ignores the diversity in riskiness and growth prospects of stocks. But wait a minute. The reason for overexuberance and extreme panic in markets is also related to errors in perceiving risk and growth prospects. This means that fundamentally weighted indexes share the same sources of error as market-cap-weighted indexes, so no clear-cut theoretical claim of superiority can be made. All we have to go on is past performance, which of course famously lacks the guarantee of future performance.

So, Which One Will It Be?
That said, some impressive results based on past performance did play a key role in sparking interest in the topic of fundamental indexing. For example, back tests on the Research Affiliates Fundamental 1000 Index, the flagship index from the firm founded by Robert Arnott, show a more than 2 percent annual advantage over the S&P 500 Index between 1926 and 2004. However, detractors have pointed out that by construction, fundamental indexing introduces a small-cap and value tilt, which makes comparison to a large-cap style-neutral index such as the S&P 500 somewhat unfair. Also, large-cap growth stocks had an especially harrowing bear market in 2000-2002, so these results likely contain substantial end-point bias. Indeed, fundamental indexes have not fared nearly as well thus far in the current bear market. Mainstream market-cap-weighted index funds like  Vanguard 500 Index (VFINX) and  Fidelity Spartan 500 Index  have easily outperformed prominent fundamental index choices such as  WisdomTree LargeCap Dividend (DLN) and the aforementioned RAFI.

Conclusion
Recent studies have shown that fundamental indexing doesn't quite have the theoretical basis that its proponents claim and that fundamental indexes are ultimately exposed to the same errors as are their market-cap-weighted counterparts. Fundamental indexing may still appeal to those who seek a happy medium between passive and active investing. But most indexers have ample reason to stick with traditional choices.

This commentary is from Morningstar Mutual Funds.

		Morningstar Mutual Funds
	Morningstar Mutual Funds is a one-stop resource for serious fund investors. Every two weeks, investors receive a commentary, benchmark		data, and performance data on 1,600 funds. They also receive 160 new full-page analyses on specific funds. Learn more.
	$619.00 for 24 issues