Markets: News  

Chuck Jaffe

These ETFs get an 'F' for fiction

7/25/12 3:52 PM ET (MarketWatch)Print

BOSTON (MarketWatch) -- Some 1,400 exchange-traded funds are available currently, with more being created practically every day.

New research from the Vanguard Group confirms what many observers already know: New ETFs based on young, back-tested indexes are more about what an investment firm can sell than what it actually believes.

Accordingly, new ETFs based on new indexes with no extensive track record are the Stupid Investment of the Week.

Stupid Investment of the Week highlights conditions and characteristics that make a security less than ideal for average investors. It's written in the hope that spotlighting danger in one situation will make it easier to avoid trouble elsewhere. The column is not meant as an automatic sell signal; in this case, it is strictly a warning against buying new ETFs based on unproven indexes.

The danger here is not "new ETFs" so much as "new indexes."

The Vanguard report, "Joined at the Hip: ETF and Index Development," found that the 1,400 U.S. listed ETFs track more than 1,000 different indexes. But more than half of these benchmarks had existed for less than six months before an ETF came along to track it.

In addition, Vanguard research shows that more than half of the indexes used "back-tested" performance data, showing "results" based on hypothetical assumptions.

Back-testing strain

Another definition of back-testing: Fiction.

"Most investors believe in past performance histories being a predictor of the future, even when the research shows it really doesn't work that way," said Joel Dickson, one of the study's authors and a principal in Vanguard's Investment Strategy Group. "But how much do you want to believe that hypothetical performance is a reliable predictor of the future? You're not even looking at live performance, just what someone decided to study."

Actually, it's worse than that, because index providers go fishing to find the "right" strategy to market.

In a grossly oversimplified example, say some firm has the idea for indexes based on the letters of the alphabet in the name of the company. It plans to create a "Z Fund," for example, that only has stocks with the letter Z in their name.

Thus, it has 26 potential indexes, one for each letter. It back-tests the portfolios to see how each "alphabet strategy" would have done over the last 10 or 20 years. Say the letters C, J, X and Z had superior performance, while A, S, Q and Y are dreadful, and the remaining letters are average.

In ETF World, the firm opens the four funds based on the top indexes and ignores the mediocre and underperforming alphabet strategies.

The theory, of course, is investment hogwash, but the firm now has an index, back-tested performance data, a marketing concept and something to sell.

Caveat empty

ETF firms aren't going quite that far, but Dickson noted that investors must rethink how they view indexes, especially when applied to new ETFs.

Said Dickson: "Indexes are not just unbiased, objective measures of a market segment or industry or investment style. They are marketing tools for ETFs to be launched in a way that allows the fund sponsor to show 'historical performance' that an investment manager never could show in a traditional fund."

In short, ETF creation is done through the rear-view mirror, looking back to find stories about how a strategy would have been viable, thus making it worthwhile to pursue now.

"The relevance of past performance when selecting ETFs is quite limited," said Todd Rosenbluth, ETF analyst with S&P Capital IQ. "The market rotates quite regularly, and something that was back-tested to work in 2011 -- when consumer staples and utilities were the best places to be -- would have failed miserably in 2012, when those were the two worst performing sectors."

Added Tom Lydon, editor of ETF Trends: "Right now, all of the interest is on fixed-income funds and we're seeing a lot of new fixed-income indexes. Well, those were back-tested against a 30-year history of falling interest rates to historic lows, and if we assume that environment won't continue forever, you really are looking at funds where you have no idea what they could do if and when conditions change."

Of course, everything depends on the ETF being created. There's a difference between a new ETF based on the Standard & Poor's 500-stock index and one you have never heard of from a firm with a "proprietary method" for choosing stocks from a large universe.

Likewise, a manager such as Bill Gross bringing his track record to an ETF cloned from his Pimco Total Return Fund is not doing anything goofy to find a niche.

But for the vast majority of new funds, average investors should wait and watch. These funds often fold if they fail to attract a following, and their efforts to attract attention may involve using strategies that blow up if the market turns.

"You've got about 1,400 ETFs today, and about 95% of the money invested is in the largest 10% of those funds," Lydon said. "You have plenty of choice, and you're not really missing out on anything if you don't buy the newest ETF out there. … If you see some creative new index with an ETF, watch it for awhile. Maybe it turns out to be something you want to own, but you don't need to take the risk and jump in right away, when they're saying the strategy is proven but you know it's really not."