Using quantitative methods, enhanced-index strategies add an active component to passive investing.
This article first appeared in the February/March 2011 issue of Morningstar Advisor magazine. Get your free subscription here.
Exchange-traded funds have become popular tools for establishing the passive building blocks of a portfolio because of their low costs and tight tracking errors. One area in which they have yet to move the needle is in active management. And it's not without reason. What successful mutual-fund manager wants to disclose his or her holdings every day, as ETFs are required to do? Transparency is fine when you have 500 holdings and no fundamental opinion about them. But when you are an active manager with 30 to 40 ideas, you don't want to disclose to the market that you are buying or selling a stock.
Somewhere in the middle of the active/passive spectrum lies "enhanced" indexing. This approach relies on quantitative or rules-based methods, usually in the form of slight factor tilts to a traditional index. The result is a portfolio with characteristics that active managers usually seek, such as stocks with low price/earnings ratios or good momentum, and the mechanical and predictable attributes that indexers like. Enhanced-index fund managers usually do not mind disclosing their holdings, because their active bets are so small that the decision to sell one stock should not influence its price.
ETFs that use enhanced-indexing strategies are showing signs of success (Exhibit 1). A number of them have achieved 4 or 5-star Morningstar Ratings. And while all enhanced-index ETFs rely on quantitative methods, they vary in their degree of complexity, ranging from a simple equal-weighted and dividend-yield approaches to more-complicated multifactor strategies.
Perhaps the least complicated enhanced-index fund is Rydex S&P Equal Weight
The result is a portfolio with a return profile more like a mid-cap fund, with the resulting higher volatility. The S&P has more than half of its assets tucked in to giant caps and less than 12% in mid-caps; it's roughly the opposite in the Rydex fund, meaning that the fund is much more responsive to factors such as momentum and valuation. (Mega-cap companies tend to grow much more slowly and are less responsive than smaller companies to these factors.)
This equal-weighting approach worked great during the past 10 years, but will it work during the next 10? Small-cap stocks now trade at a valuation premium to large caps, while the opposite was true a decade ago. Large caps also have more international revenue and will be more resilient in the face of anemic economic growth. But even if small caps do not do as well, the strategy also benefits from a forced rebalancing to maintain the portfolio's equal weightings. Every quarter, the fund buys the stocks that have sold off and sells the ones that have appreciated--it is buying low and selling high.
While this methodology might work, there does not seem to be any fundamental reason why all stocks should be weighted equally. It makes intuitive sense to weight companies by some fundamental measure of worth, such as earnings. WisdomTree Earnings 500
Finding Dividend Yield
Another simple approach to quant investing is to buy stocks with strong dividends. Dividend yield has many uses as a valuation indicator. Ben Graham compared the dividend yield with bond yields as a measure of a stock's attractiveness. Dividends have historically accounted for 40% of the returns from investing in stocks, and despite conventional wisdom, high-dividend-payout companies tend to have stronger earnings growth. So the case for investing in companies that pay dividends is a strong one.
A logical way to create a high-yield index fund would be to weight stocks by the percentage of their dividend yield, so that higher-yielding stocks would make up a larger percentage of the fund than do lower-yield names. But what about junk companies that have a temporarily high dividend on the verge of a dividend cut?
One way to avoid the riskier dividend-payers is to buy high-quality firms, such as those with strong brands or dominant market positions, that consistently have raised dividends. Proctor & Gamble
SPDR S&P Dividend
WisdomTree LargeCap Dividend
The Multifactor Approach
Rob Arnott of Research Affiliates patented an approach that weights firms based on a four fundamental factors of value: dividends, book value, sales, and cash flow. Like the equal-weighting approach, Arnott's multifactor method also benefits from the forced rebalanced strategy of selling winners and buying losers. But it also results in a deeper value tilt and less small-cap exposure than equal weighting provides. In fact, by using four measures of value instead of just one or two, the method offers a more-consistent return. After all, when a fundamental analyst thinks of value, he or she takes into account a variety of factors. Thus, the multifactor approach might do a better job of recreating the process of a fundamental analyst. The strategy is available through PowerShares FTSE RAFI US 1000
First Trust Large Cap Core AlphaDEX
We can analyze the performance and risk of these funds through a returns-based attribution model (Exhibit 2). The Carhart four-factor model is an extension of the Fama-French model, adding momentum to market, value, and small-size factors. For the market factor, an exposure, or beta, of 1 indicates that it matches the market perfectly. Positive values for the other factors indicate an overweighting or risk to those factors; a negative value indicates opposite exposure. For example, a negative value factor indicates exposure to more-expensive stocks.
The Rydex S&P Equal Weight Fund has positive exposure to small-cap stocks and a negative exposure to momentum. Even though the fund does not target momentum explicitly, the process of selling winning stocks and buying losers to rebalance back to equal weightings has a natural anti-momentum effect. Weighting its holdings by earnings, the WisdomTree Earnings 500 Fund naturally produces a large- cap bias because large companies have a greater amount of earnings than do smaller companies.
Among the dividend funds, SPDR S&P Dividend has a small-cap and deep-value bias (note the relatively high value of the small-size and value factors), which results from weighting stocks by dividend yield. Smaller-cap companies that have higher risk sometimes offer high yields to attract investment. WisdomTree LargeCap Dividend's strategy of weighting by a dividend's dollar amount rather than percentage yield results in a large-cap bias and value bias. Weighting by the dividend's dollar amount gives preference to larger companies, which have the scale to support a large dollar payout. These companies also tend to be value companies with few growth opportunities. These funds have a value tilt, but Vanguard Dividend Appreciation's screening method results in a negative value bias. The fund's screen (companies with a consistent 10-year track record of increasing dividends) results in a quality tilt toward companies with strong brands that are able to maintain market share.
With the multifactor quant funds, we see that PowerShares Dynamic Large Cap is using momentum, while First Trust Large Cap Core Alpha leans toward small caps. PowerShares PowerShares FTSE RAFI US 1000 has a large exposure to value stocks. This results from its stock-weighting methodology, which gives preference to the four-value factors.
Whether enhanced-index ETFs will continue to enjoy th eir strong performance is open for debate and depends somewhat on how well small caps and value stocks perform. Regardless, we are likely to see more quantitatively run ETFs. Given their low cost, they are a good choice for investors who are looking more active exposure within an ETF than the purely passive, traditional index-based ETFs.
Michael Rawson is an ETF analyst with Morningstar.