Please enable JavaScript to view the comments powered by Disqus.

Index Funds

Reinventing the Index

One of the simplest investing strategies is getting more complicated.

As investors suffered during the bear market that began in 2000, Rob Arnott struggled to understand why anyone would call it a bear market in the first place. True, Standard & Poor's 500-stock index dropped 25% from its March 2000 peak through the end of 2001. But Arnott, a California money manager and editor of Financial Analysts Journal, observed that the average stock traded on the New York Stock Exchange rose 20% during the period. If it was an up market for so many stocks, why, he wondered, did investors lose heavily by owning funds that track the SP 500 and other popular indexes?

This question is fueling one of the fiercest debates about index investing since it first became popular 30 years ago. The dispute isn't about whether indexing, in contrast with owning actively managed portfolios, is a good idea. Both sides agree that it is.


Fund Watch Column

A Strong 2007 for Stocks?

Booyah! The Manic Universe of Jim Cramer

Rather, the argument is about the nature of the indexes themselves. Critics of the SP 500 and other widely used barometers say the indexes are inherently flawed because they are dominated by a handful of companies and because they must, by their very nature, add more shares of stocks that are rising and trim positions in stocks that are falling. Therefore, as investors move money in and out, managers of index funds are forced to buy stocks when they are overpriced and sell them when they are undervalued. If they don't trade, their funds no longer replicate the index they promise to track.

A brewing debate

Experts such as Arnott contend that the new generation of index funds, built on more sophisticated techniques, still provide diversification and low cost -- but with better returns and less risk. Defenders of the old order say the advantages of the new indexes are illusory because they are skewed toward stocks that have excelled lately but aren't certain to do well in the future.


Who's right? It's too soon to tell. But the new indexes are backed by formidable intellectual heavyweights. And funds that take the new pathways are starting to pique the interest of ordinary investors. "I think they are an interesting concept," says Donald Sonn, a Northampton, Mass., doctor who keeps virtually his entire portfolio in index funds and is considering investing in some of the new offerings. "Some companies can totally dominate an index fund," he says, "and that can put you more at risk. Spreading the risk can be beneficial."

Are new-generation indexes truly less risky? Let's take a closer look.

Like most indexes, the SP 500 holds stocks in proportion to their market value (share price times the number of shares outstanding). So its holdings of the biggest company, ExxonMobil, with a market value of $415 billion as of mid August, are roughly 729 times greater than its holdings of number 500, Gateway, a $569-million firm.

You don't need a PhD to understand why it works this way. To own a representative sample of the market -- the purpose of indexing -- you need more shares of larger companies and fewer shares of smaller ones. Often, this results in index-fund managers buying more shares of companies as their prices rise, even if those shares are already expensive. At the same time, they must sell stocks that are out of favor even if they are well priced to buy.


This was a serious problem as the bull market neared its end in early 2000, when investors drove the prices of huge companies, such as General Electric, to extremes (it wasn't just a technology bubble). At the end of 1999, the ten biggest companies in the SP 500 had a collective value of $3.1 trillion, more than one-fourth of the index's total. The ten biggest companies today are worth just $2.3 trillion, so the SP 500 is no longer as top-heavy. The top ten account for just 20% of the index's value now.

By the way, the top ten stocks from 1999 are worth just $1.8 trillion today. That $1.3 trillion in lost wealth tells you that large-company stocks (and the indexes that measure them) have been dogs for years. No wonder critics say the shortcomings of traditional indexing are chronic, even if they aren't acute at the moment. And they say the shortcomings extend to all indexes that are weighted by market value. Those include the small-company Russell 2000, the Wilshire 5000 and others. If you own a fund that tracks a market-weighted index, "you can be absolutely assured that most of your money is in overvalued companies," says Arnott.

Modified index funds are not entirely new. In 1987, Morgan Stanley launched a fund that weights every company in the SP 500 the same. So Morgan Stanley Equally Weighted SP 500 (symbol VADBX) for the Class B shares) owns, for example, ExxonMobil and Gateway in equal proportions. In 2003, SP came out with its own equally weighted index, and an exchange-traded fund, Rydex SP Equal Weight (RSP), tracks it at a fraction of the cost of the Morgan Stanley fund. But even though this equally weighted strategy has trounced the traditional SP 500 lately (15% versus 11% annualized over the past three years), it's an apples-to-oranges comparison. The equal-weight index owns a much higher proportion of small and midsize firms. If you invest now in the equally weighted index instead of a fund that tracks the regular SP 500, such as Vanguard 500 Index (VFINX), you'll be shortchanging the megasize companies just as they may be poised for a comeback.

So can you own an index fund that is less vulnerable to market manias but holds stocks in reasonable proportions? Some experts say you can. Arnott is the guiding force behind a set of so-called fundamental indexes that don't take market value into account directly. The Research Affiliates Fundamental Indexes, or RAFIs (named for Arnott's Pasadena, Cal., money-management firm), weight holdings on four factors: sales, dividends, cash flow (earnings plus depreciation and other noncash charges) and book value (assets minus liabilities). These indexes, says Arnott, "measure a company's footprint in the economy." So when stocks' market values diverge drastically from economic fundamentals, as they did during the speculative blowoff of the late 1990s, these re-engineered indexes provide a better measure of value, he argues.


A comparison of the holdings of the flagship RAFI US 1000 with those of the SP 500 illustrates RAFI's contrarian ways. Overall, the two indexes hold 20 stocks in common among their top 25 (though often in differing order). But RAFI's top 25 includes struggling General Motors, Ford, Merck and Time Warner. It also holds Berkshire Hathaway, which is excluded from the SP 500 because its high-priced shares trade too infrequently. The SP 500, on the other hand, holds Cisco and Intel in its top 25, as well as Home Depot, PepsiCo and UPS.

Arnott's firm has created 24 fundamental indexes that track U.S. and foreign stocks. An ETF launched last December, PowerShares FTSE RAFI US 1000 (PRF), tracks the U.S. index. In the first seven months of 2006, the ETF returned 6%, about twice the gain of Vanguard 500 Index. ETFs that seek to track other RAFIs are working their way through the regulatory approval process.

It's hard to draw conclusions about RAFIs' performance from such a short time span. For a clue as to how the approach might fare over the long term, check out a study by Arnott and two coauthors published last year in Financial Analysts Journal. The study found that a hypothetical 1,000-stock fundamental index would have outpaced the SP 500 by about two percentage points per year, on average, from 1962 through 2004, and with less volatility.

Value against growth

Not everyone buys the idea that RAFIs are better than traditional measures. Skeptics argue that the RAFI performance edge derives from a tilt toward small-company stocks and "value" stocks -- shares of struggling and lesser-known companies that are trading at a discount to their true worth. Small-company and value stocks have performed particularly well over the past few years. Arnott doesn't dispute that but says that even if you disregard the recent past, the RAFI US 1000 looks superior to the SP 500. Besides, Arnott notes, market-value-weighted indexes have their own biases -- toward big growth stocks.


Not fair, counters George "Gus" Sauter, who used to manage Vanguard's flagship SP 500 fund and now oversees all of Vanguard's index products. Sauter contends that you essentially bet against the market if you invest in an index fund that isn't market-value-weighted. "Why kid yourself and think you have some kind of superior performance when all you're really doing is getting exposure from a different segment of the market?" he asks. Counters Arnott: "If it works, use it."

Over the long haul, investors in the RAFI funds should earn fairly comparable returns to those of traditional indexes. But that may not hold true for other new-generation index funds making their way into the marketplace.

New York City's WisdomTree Investments, for example, weights companies by their cash dividends. In June, WisdomTree launched 20 ETFs based on its own dividend-weighted indexes. Weighting by dividends steers these indexes sharply away from fast-growing companies, which often don't pay dividends. But WisdomTree's backers, who include Wharton School professor and Kiplinger's columnist Jeremy Siegel, argue that dividend-paying stocks beat the market over time. If you're high on dividend-paying stocks -- and there are many reasons to be (see "Dividends With Room to Grow," Feb. 2006) -- these indexes are worth considering. Just don't use them as direct substitutes for broad-based index funds. For example, WisdomTree Dividend Top 100 (DTN) has 58% of assets in financial and utility companies.

Wheaton, Ill.-based PowerShares takes selective indexing a step further, using 25 factors to construct so-called intelligent indexes. The goal of these Intellidexes, created with the American Stock Exchange, is not to match the market but to beat it. The factors, which include measures of value, price momentum, risk and timeliness, are intended to highlight companies with "the greatest investment merit," says Bruce Bond, PowerShares' president. An ETF based on its large-company index, the 100-stock PowerShares Dynamic Market Portfolio (PWC), returned an annualized 16% over the past three years to August 1, compared with 11% for the SP 500 (24 Intellidex-based ETFs already trade; 31 more await regulatory approval).

But it's impossible to know whether these indexes can retain their edge. Although PowerShares discloses generally how it determines its indexes, exact formulas are secret. You're essentially buying a "black box" strategy with a short track record.

Also keep in mind that fundamental indexes are frequently rebalanced because of changes in factors such as dividends and cash flow. That means these funds buy and sell more often, driving up trading costs. Moreover, the expense ratio of the RAFI ETF is 0.6% annually, nearly seven times that of the lowest-cost SP 500 ETFs.

Arnott argues that the performance of funds based on RAFIs will more than offset the higher expenses. But his confidence is based on his indexes' performance in hypothetical tests using historical data. It is possible that the markets will continue to act as they have in the past. But then again, markets love to change course when you least expect them to do so.

The index boom: How three market measures compare

Understanding what makes indexes tick isn't just an academic exercise. Investors, who have poured billions of dollars into traditional SP 500 index funds, are being wooed by funds that track new kinds of indexes that may -- or may not -- improve on the tried and true. Consider these three indexes that track large-company stocks.

SP 500
ExxonMobil (4%) Gateway (0.004%) $86.6 billion 6% 18 2.8 1.92% Financials (22%)
SP 500 Equally Weighted
Kerr-McGee (0.3%) PMC-Sierra (0.1%) $24.6 billion 42% 18 2.9 1.58% Financials (18%)
RAFI 1000
ExxonMobil (4%) First Citizens Bancshares (0.01%) $85.0 billion 12% 16 2.4 2.17% Financials (23%)

Data to August 1. *Based on trailing 12-month earnings. Sources: Research Affiliates, Rydex Investments, Standard Poor's.