During the five-year period ending April 30, 2014, the Guggenheim S&P 500 Equal Weight ETF (NYSE Arca: RSP) returned 21.84% on an annualized basis. Over the same period of time, the SPDR S&P 500 ETF (NYSE Arca: SPY), which holds the same stocks as RSP but weights them by market capitalization instead of equally, yielded investors a total annualized return of 18.95%, nearly 3% less per year than RSP.
The discrepancy between the five-year track records of RSP and SPY illustrates the importance of index weighting in determining how an index (and any portfolios that track it) will perform over time.
This is not meant to suggest that an equal-weighted approach is superior to a cap-weighted approach. The PowerShares QQQ ETF (NASDAQ: QQQ), which tracks an index that employs a modified cap-weighted methodology, outperformed both RSP and SPY over the five-year period ended April 30, 2014, generating a total annualized return of 24.82%. Meanwhile, its equal-weighted counterpart, the First Trust NASDAQ-100 Equal Weighted Index Fund (NASDAQ: QQEW), returned 20.9% per year over the same period.
That we are having this discussion at all is testament to how far the index industry has come since the day back in 1896 when Charles Dow first published the Dow Jones Industrial Average (DJIA) with the aim of selling more newspapers. While the newspaper industry ultimately peaked and began a stunning decline, the index business has thrived with the explosion in popularity of passive investing and ETFs.
Questions about index weighting methodologies have been with us since the beginning. Owing in part to the limited computational power afforded by the technology of his day, Dow opted to employ a price-weighted methodology with the DJIA, basing each constituent’s index weighting on the relative size of its share price. This approach produces unusual outcomes, with each constituent’s weighting based solely on its stock price per share and unrelated to its market capitalization or any other economic or financial factor.
Consider the disparate treatment accorded to Berkshire Hathaway (NYSE: BRK-A) and Microsoft (NASDAQ: MSFT) under a price-weighted approach. While the companies have roughly the same market capitalization, Berkshire Hathaway would receive a weighting about 4,775 times the size of Microsoft’s under a price-weighted approach because its stock price of $191,000 per share dwarfs the $40 per share price of Microsoft stock. Indeed, Berkshire Hathaway would constitute over 98% of the DJIA if it were included in the index, which explains why one of the nation’s largest and most respected public companies is not included in its most famous stock market index.
The advent of modern portfolio theory beginning in the 1950s, and its assumption of market efficiency, provided the theoretical foundation for market cap-weighted indexing and offered a rationale for investors to index their portfolios rather than selectively purchase individual stocks and bonds. In particular, James Tobin hypothesized in 1958 that, under certain assumptions, the optimal combination of risky securities is the same for every investor regardless of risk tolerance. Adjusting for risk tolerance under the Tobin hypothesis merely requires buying the optimal portfolio and levering up (taking on margin to buy more of it) or levering down (converting a portion of it to cash) depending upon how much risk you want to take.
If the assumptions underlying Tobin’s theory held true in the real world, there would only be need for one mutual fund: the Market Portfolio, as finance professionals call it, and it would provide market cap-weighted exposure to every risky asset in the world.
As it turns out, according to the Investment Company Institute, there were over 9,000 mutual funds, ETFs and closed-end funds available in the U.S. market for investors to purchase as of the end of 2012. Investors are clearly not buying Tobin’s theory.
Part of this is because the Market Portfolio is a theoretical concept, but in the real world not all risky assets are available for sale. Nor do investors necessarily want the Market Portfolio. In practice, investors face different tax rates and have psychological biases (for example, home country bias) that may make the theoretical Market Portfolio unappealing to them.
While investors may not be buying Tobin’s theory, the idea of weighting portfolios of securities by market capitalization has resonated with investment product sponsors and the investing public. Market cap-weighted indexes have proven difficult for active money managers to beat. Moreover, market cap-weighted indexes generate minimal portfolio turnover (and therefore less tracking error) because index weights track changes in market value by design.
Still, a question remains: what if markets aren’t efficient?
For many observers, the dot-com boom of 1997-2000 stands as definitive proof that markets can be inefficient. The fallout from the subsequent crash led to criticisms that market cap-weighted indexes have a tendency to assign high weightings to overvalued stocks and low weightings to undervalued stocks. Along similar lines, critics of market value-weighted bond indexes have argued that such indexes take on excessive risk by assigning higher weightings to more heavily indebted companies.
Index providers have responded by developing products that employ alternative constituent weighting approaches. Equal-weighted indexes led the way and have since been followed by methodologies that employ rankings based on various fundamental factors, such as dividends and profitability, to determine constituent weightings. These latter indexes generally purport to exploit historical market inefficiencies to deliver higher risk-adjusted returns than market cap-weighted indexes.
In considering alternatives to market cap-weighted indexes, investors should keep a couple of considerations in mind. First, alternative approaches can dramatically change the nature of the size, style and sector exposures in your portfolio. They should not be compared with market cap-weighted portfolios without taking into account how the alternative weighting approach changes the nature of the investment strategy.
More importantly, back-tested outperformance means nothing if it can’t be replicated in practice. Alternative portfolio weighting approaches require more frequent and extensive rebalancing in order to achieve the alternative target portfolio weights. This can drive costly portfolio turnover that eats into any purported risk-adjusted performance advantage.