The dominant driver of performance in an exchange-traded fund (ETF) is the index with which it is paired. So, if you can select the "best-fit" index for your personal needs, you are on the road to ETF success. However, deciphering indexes is one of the more demanding tasks in the investment field, and it isn't getting easier.

Indexing has become an innovation-driven science, and investor acceptance of new concepts has fueled a "great debate" about which indexing method will produce superior performance over time. In this article, we will discuss key points in this debate while comparing three genres of indexing: market cap weight, equal weight and fundamental indexing.

Defining Investment Performance
Investment indexes are designed to mirror the average performance of financial markets. For example, the most successful index in history is the Standard & Poor's 500 Index (S&P 500). According to Standard & Poor's, more than $1.5 trillion of capital is linked to this benchmark.

The S&P 500 index assigns 500 component stocks a weight that reflects its market capitalization and the total value of all shares of stock outstanding. One advantage of such a market cap weighted index is simplicity. As stock prices change daily, the weightings of index components automatically adjust. (To learn more on indexes, see Don't Judge An Index By Its Cover and An Inside Look At ETF Construction.)

However, there also are drawbacks to this method, starting with a fairly heavy concentration in the largest stocks. The top 10 stocks typically account for about 20% of the S&P 500 index weight and performance, as shown in the table below.
 

Top 10 Components of S&P 500 Index By Weight  As of March 2008

Company	Float Adjusted Market Cap (Millions of dollars)	Index Weight
Exxon Mobile	453	3.93%
General Electric	370	3.21%
AT&T	231	2.01%
Microsoft	227	1.97%
Procter & Gamble	216	1.87%
Johnson & Johnson	184	1.60%
Chevron	177	1.54%
Bank of America	168	1.46%
International Bus. Machines	159	1.38%
JPMorgan Chase & Co.	146	1.27%

Top Ten Total: 20.2%

Source: Standard & Poor's.

Many other leading U.S. equity indexes have the same market-cap weighting and top-heavy bias toward large companies. They include the Russell 1000, Russell 3000 and Wilshire 5000 Total Market Index (TMWX). However, each of these indexes includes more components than the S&P 500, so their top 10 holdings represent less than 20% of total market cap.

Because market cap weighted indexes tilt toward large companies, they generally perform best when large caps outperform mid/small caps. Also, these indexes perform well in momentum-driven markets. During the late 1990s bull market, the S&P 500 became heavily weighted toward tech stocks and thus became more vulnerable to the crash in prices that followed in 2000-2002. According to Standard & Poor's, the allocation of the S&P 500 Index increased from 11% in the technology sector at the start of 1995 to a peak of 34.3% in March 2000.

The Value of Equal-Weight Indexes
An alternative to a market cap weighted benchmark is the Standard & Poor's 500 Equal-Weight Index. Instead of tilting toward the largest companies, this index assigns the same weighting. For example, the S&P 500 gives a 0.2% weight to each of the 500 components, and rebalances quarterly to adjust for changes in market values. An ETF that tracks this index is Rydex S&P Equal Weight (AMEX:RSP).

An equal-weight index tends to perform better than a market cap weighted index in environments that favor mid/small-cap stocks. The equal-weight index also has the advantage of avoiding excessive valuations during momentum-driven markets.

Some investors believe the choice between a market cap weighted index and an equal-weight index is a nuance. However, the performance differentials can be surprisingly large. The broadest of all U.S. stock market indexes are published by Wilshire Associates and include more than 5,000 component stocks, which account for more than 99% of all U.S. publicly traded equities. The table below compares performance of the equal-weight version of this index with the market cap weighted version. Both versions hold the same components.

Performance of Wilshire 5000 Indexes For Periods Ending 3/31/08

Index	1 Year	3 Years Annualized	5 Years Annualized	10 Years Annualized
Wilshire 5000 Equal Weight	-18.77%	3.44%	20.75%	12.94%
Wilshire 5000 (market cap-weighted)	-5.76%	6.37%	12.45%	3.95%

Source: Wilshire Associates. Periodic updates of this information can found at Wilshire.com.

The equal-weight 5000 index has superior long-term performance because longer periods are more favorable for small/mid-cap stocks. However, over the short periods shown, the trend reverses and large caps return to favor.

Rebalancing Cost Drag
Unlike the Russell and Dow Jones Wilshire indexes in which components are selected mechanically based on "quant" statistics, the S&P 500 is driven by a committee approach to stock inclusion and turnover. (For related reading, see Rebalance Your Portfolio To Stay On Track.)

Both market cap weighted and equal-weight indexes of the same sponsors tend to change components at the same time. The difference is that the equal-weight indexes must be rebalanced back to the target weightings periodically, while the market cap weighted indexes are not rebalanced to correct for market price changes. Rebalancing generates a few basis points per year of cost drag in ETFs that track equal-weight indexes.

Fundamental Indexing Enters the Debate
Based on research conducted by Rob Arnott, chairman of Research Associates, a new concept in indexing was introduced in 2005 as a joint venture between the FTSE Group and Research Associates. The new broad-based U.S. equity index, FTSE RAFI 1000, selects 1,000 components based on a rules-based model that includes sales, cash flow, book value and dividends. A fundamental index attempts to go beyond the concept of mirroring the experience of an "average investor" by selecting and weighting component stocks based on current and quantitative ranking of company data. (For more insight, read Fundamentally-Weighted Index Investing.)

When the FTSE RAFI 1000 was paired with a PowerShares ETF to form the PowerShares FTSE US 1000 (PSE:PRF), it proved so popular with investors that PowerShares has since rolled out a large menu of fundamental ETFs. WidsomTree Investments and Claymore Securities also promoted the concept by launching ETFs tied to fundamental indexes.

The performance of the FTSE RAFI 1000 has provided an early test of how well fundamental indexes can perform vs. traditional market cap weighted benchmarks. In the table below, actual performance (since 2005) is supplemented with hypothetical backtest data showing how the index hypothetically would have performed over the past 10 years.

Performance Comparison For Periods Ending 2/28/08

Index	1 Year	3 Years Annualized	5 Years Annualized	10 Years Annualized
FTSE RAFI US 1000	-8.66%	6.20%	13.96%	7.10%
S&P 500	-5.08%	5.84%	11.32%	3.50%
Russell 1000	-5.40%	6.19%	11.86%	3.83%

Source: PowerShares. Periodic updates of this information can be found at Invescopowershares.com.

Based on this data, it appears that fundamental indexes occupy a middle ground between market cap weighted and equal-weight indexes. They tend to track somewhat closer to market cap weighted index performance over time, while avoiding those indexes' large-cap bias and vulnerability to momentum-driven markets.

Research Associates contends that the benefit of fundamental indexing is in tempering portfolio risk, especially during momentum-driven markets. By selecting and weighting components based on company data and periodically rebalancing the index to reflect new data, the fundamental index will tend to have a lower price/earnings ratio (P/E ratio) and less volatility than market cap weighted indexes. Performance data also suggests that fundamental indexes may tend to overweight value stocks and underweight growth stocks relative to comparable market cap weighted indexes. (For more, read A Guide To Portfolio Construction.)

Over long periods of time, Research Associates claims that fundamental indexing should produce a small but meaningful performance edge over market cap weighted indexes. The firm's analysis shows that on a backtested basis over 45 years, the FTSE RAFI 1000 hypothetically would have returned an annualized 12.5%, compared to 10.3% for the S&P 500's actual historical return.

Conclusion
For investors, perhaps the best way to learn more about any type of index is to allocate a portion of a portfolio to it via ETFs or index mutual funds and then track it against alternatives over time.
 

by Rich White, (Contact Author | Biography)

Rich White has been a freelance financial writer since 1981. He is the former editor of Financial Planning magazine and author of several books including "Twelve Steps to Your Personal Success in the 401(k) and Small Plan Market" and "The Complete Rollover Guide for Financial Professionals".