From the March 2011 issue of Investment Advisor • Subscribe!

Quantitative Methodology ETFs Search for Better Performance

ETFs owe much of their success to providing investors with exposure to asset classes that have been difficult for them to access in the past. ETFs that offer a way to invest in international and emerging markets have proved particularly popular, because those markets have not been available to investors without the help of a professional fund manager.

As innovative and exciting as these ETFs are, they are typically based on indices designed to reflect the investable universe of a particular market, sector, or investment style. These indices tend to change very little year after year.

A new breed of ETF has emerged, however, based on indices that are designed not to reflect the investable universe of a particular market niche, but to pick the best stocks within a specific sector or style. These indices use a series of screens or follow a set of rules to manage the portfolio in hopes of outperforming the broader market.

There are dozens of ETFs based on indices that use quantitative methodology to select their components. In terms of assets under management, the largest of these is the First Trust Materials AlphaDEX ETF, which has about $490 million in assets, according to company data. This ETF is one of about 15 such ETFs First Trust operates that target other sectors, as well as style/size categories. These ETFs use the proprietary “AlphaDEX” methodology, which screens stocks from the Russell 1000 into three groups: growth, core/blend and value. It ranks stocks in each group according to a scoring method that uses various financial metrics such as price/book ratio or one-year sales growth, and then eliminates the bottom 25%. Of the remaining stocks, the top 20% get a 33% weighting, while the bottom 20% get only a 6.7% weighting. The index is reconstituted and rebalanced quarterly.

All well and good, but the proof is in the performance. Since its inception in May 2007, the materials fund posted an average annualized gain of 6.22% as measured by its net asset value (which includes the effects of fees), compared with 1.48% for the materials sector of the S&P 500 and -2.36% for the Russell 1000 (which do not include fees). Over the three-year period ending Dec. 31, 2010, the fund returned an average of 4.32% annually at net asset value, compared with negative returns for the S&P 500 materials sector or the Russell 1000.

Several other ETF sponsors operate quantitative methodology ETFs. Invesco/PowerShares probably has the most with about 30 ETFs that use its “dynamic” stock selection methodology targeting sector, sub-sector, market cap and broad market strategies. Starting with a universe of about 2,000 mostly domestic stocks, 25 different financial metrics are used to assign each stock a score, and the top 200 are then selected.

This process is used to create an index on which the PowerShares Dynamic MagnaQuant Portfolio ETF is based. This ETF, which opened in October 2006, has not delivered as impressive a performance as the First Trust Materials fund; its average annual return based on net asset value (which includes fees) was -0.24% from inception to year-end 2010, compared with 3.74% for the S&P 500 Equal Weight Index, 1.74% for the Russell 3000 Index, and 0.75% for the S&P 500, its self-declared benchmarks.

The Guggenheim Insider Sentiment ETF uses data regarding purchases of stock by company insiders and also incorporates data from changes in analyst earnings estimates to identify stocks that the fund sponsors hope will outperform in the future. It was launched in September 2006 and has about $151 million in assets. Net asset value returned an average 5.39% over the past three years, and 26.39% over the past 12 months through year-end 2010, compared with -2.85% and 10.76%, respectively, for the S&P 500.

S&P Senior Financial Writer Vaughan Scully can be reached at Vaughan_Scully@standardandpoors.com. Send him your ideas for ETF story topics.

Reprints Discuss this story
This is where the comments go.