Buy the Unloved With ETFs

At Morningstar we've been writing for more than a decade about a strategy that invests in asset classes that investors are selling and sells asset classes that investors are buying. We call this strategy, "Buy the Unloved." Today, exchange-traded funds offer a new way to put this strategy to work.

The Original Strategy
The premise behind buying the unloved is simple: Fund flows chase returns, and the combination of flows and strong past returns are good indicators that an asset class is overvalued. Internet funds in 1999 or real estate funds in 2007 are examples of why you wouldn't want to buy the most popular fund categories. On the flip side, you would have done well buying small-value funds in 2000, when other investors were not.

Here's how the strategy works. Invest in mutual funds from the three equity categories that received the greatest redemptions in the prior year. (Bond funds and asset-allocation funds are excluded.) Then do the same thing the next year and the next year. After three years (or four or five--those time periods work just as well), you start rolling over that original group of funds in the next group of unloved.

Over the years, we've found this contrarian strategy works pretty well. In addition, we've found that applying a similar strategy to funds from the three most popular categories (the loved) each year has done quite poorly. Through October, the top three unloved categories were large growth, large value, and world stock. The most loved were diversified emerging markets, commodities broad basket, and foreign large blend.

The strategy is definitely a niche one. You wouldn't want to touch the core of your portfolio with it.

New Opportunities for "Buy the Unloved"
The growth of ETFs presents some intriguing new opportunities for this strategy. Because the idea is to find undervalued assets rather than great managers, ETFs make a lot of sense as the way to execute the strategy. If it's an asset play, why not invest in the cheapest way to gain exposure? The Buy the Unloved strategy often leads into sector funds or regional funds, yet there are very few open-end index funds covering those areas. Now that ETFs have covered darn near every niche imaginable, we have a new opportunity for a less costly way to buy the unloved.

In addition, because you can short ETFs, it's now possible to Short the Loved, or run a long-short strategy employing both.

We set out to test whether these concepts would actually produce good results. Because the ETF is still a fairly new vehicle, we used indexes as proxies, which enabled us to go back to 1996 for our tests rather than, say, 2006, when ETFs hit critical mass. We then compared the results with broad market benchmarks to compare performance, risk, and the risk/reward trade-off.

Our tests don't completely reflect the real-world experience. For example, our tests don't reflect commissions or ETF expense ratios. But commissions are small, and ETF expense ratios would actually help for shorts because you'd benefit from the bite they take out of returns. Besides, the benchmarks we compared the strategy with also overstate the real-world results of passively holding broad benchmarks, as they did not have expenses deducted. Thus, the differences in performance should be fairly small.

Results: Buy the Unloved Long-Only
Using returns from 1996 through July 2010, the long-only Buy the Unloved strategy generated respectable performance. (Click here to see how it performed.) The unloved funds returned an annualized 7.7% compared with 5.4% for the MSCI AC World Index, a global all-country cap-weighted benchmark over that time period. The unloved also beat the S&P 500 and Morningstar US Market indexes, which returned around 6%. However, given that foreign and regional funds were included in our Buy the Unloved strategy, global indexes make better benchmarks.

The unloved strategy outpaced the benchmarks in the three-, five-, and 10-year periods as well as from the 1996 start date. Buy the Unloved also clocked in with a slightly lower standard deviation since 1996, as well as over most trailing periods.

The Buy the Unloved strategy indexes produced better long-term risk-adjusted results (using the Sharpe ratio) than the benchmarks for the 10-year and since-inception periods.

Results: Buy the Unloved 100% Long/25% Short the Loved
Of the two versions of the long-short strategy that we tested, adding a 25% short position in loved categories to 100% long the unloved proved to be the most impressive mix. In this case, the return since inception was 7.8% annualized. That's just a hair above the long-only on the Buy the Unloved. It also beat the three benchmarks over the trailing three-, five-, and 10-year periods.

But what is more impressive is what happened to standard deviation. Having 25% Short the Loved took standard deviation since inception from about 15 for the unloved-only portfolio to 11 for the long/short mix. That compares with roughly 16.5 for the US Market and MSCI AC World and 16 for the S&P 500 during the same time frame. This suggests a very welcome reduction in volatility with an increased return at the same time. Intriguingly, its Sharpe ratio topped that of all the other strategies and benchmarks. That indicates it had the best risk/reward profile.

Why is 25% the best short proportion? The small hedge seemed to help damp down volatility without surrendering much in the way of returns to leverage and opportunity costs the way you would see with a market-neutral position. It's possible that the optimal spot could be on either side of that 25% short position, but it is likely in that ballpark.

Results: Buy the Unloved 100%/100% Short the Loved
If a little shorting the loved was good, then a lot should be awesome, right? Well, that depends on where you come down in valuing stability. In fact, returns came down to an annualized 6% since inception, but standard deviation came down to 8.5. That 6% figure beat the long-short fund category average of 4.6%. However, its standard deviation of 8.9 was well above the long-short average of 4.9.

Conclusion
We certainly can't predict the performance of these strategies over the next 10 years, but these results are intriguing. The Buy the Unloved 100%/25% Short the Loved portfolio stands out as having the most promising risk/reward payoff and merits further study. ETFs present an easier and cheaper way to play the Buy the Unloved strategy.

This article originally appeared inMorningstar FundInvestor.

		FundInvestor Newsletter
	Morningstar FundInvestor is a 48-page newsletter dedicated to helping investors pick great mutual funds, build winning portfolios, and monitor their funds for		greater gains. This one-year subscription consists of 12 monthly issues, a Reader's Guide, and four free investing reports. Learn more.
	$129.00 for 12 Print Issues		$119.00 for 12 PDF Issues