Which ETF Strategies Are Poised to Outperform?

Alex Bryan: For Morningstar, I'm Alex Bryan. We're at the 2017 Morningstar ETF Conference. I'm joined today by Chris Brightman, who is the chief investment officer at Research Affiliates, who will be talking about the pitfalls of investing in some smart beta products later today.

Chris, thank you for joining me.

Chris Brightman: You're welcome.

Bryan: Chris, Research Affiliates has put out some research recently that suggests that a lot of smart beta strategies, or strategic beta as we call them here at Morningstar, their back-tested performance has been quite good. But then when you go and look at the live record of a lot of these products, that outperformance goes away. Why do you think that is?

Brightman: Allow me to broaden the observation for just a moment. By no means are our findings specific to smart beta strategies. If we look across the full range of ETF strategies that come to market, active and passive, what we find is that the performance leading up to the launch of the ETF is fantastic. It's beating the market by 2, 3, 4, 5 percentage points per annum for multiple years. And then, after the ETF is registered with the SEC and then launched, the performance seems to flatline. Now, of course, there's considerable dispersion, but the average performance across all of the ETFs is approximately marketlike. And it almost looks like the graph suggests there's something about the launch of the ETF that changes the performance. But of course, that's not what's happening.

What's happening is that the great average of all these different strategies that you might invest in, tend toward marketlike returns. But the providers giving the marketplace what it wants, what investors demand, is picking the things that in the past had the best returns. And other research that we've done has shown that if you were going to choose an investment strategy based upon past returns, especially annual returns, three-year returns or five-year returns, you'd actually be better off choosing things that had the worst three- and five-year past returns rather than the best past three- and five-year returns because of course there's mean reversion at the frequency of years.

Bryan: Is it your view then that providers are choosing strategies that are, they are choosing to launch strategies that are expensive? So, they are launching these strategies at precisely the wrong time? Is that the main thing that's going on here?

Brightman: That is consistent with the evidence that we find. And here I wouldn't want to cast a blame on the providers because the providers are responding to what the marketplace wants. Nobody launches a new strategy because their marketing team tells them that everybody is scared of this asset and doesn't want to invest in it. They launch what the clients say, 'I wish there was an ETF that attract this segment of the market or that country or this style.' And so, it's really the providers catering to investor demand.

Bryan: Are you concerned that providers are going out and overfitting the past data to make something that looks good in a back-test that may not have any merit outside of the sample period?

Brightman: Well, of course, that's a concern and many people talk about that. But I don't believe that the problem of overfitting is necessary to explain the results that we found in the study you referred to. There I think it's just selecting the segment of the market that's had the best three- and five-year performance because that's what people want to invest in. Now, of course, the more complicated a quantitative investment strategy becomes, you are quite right, the more concerned we should be about overfitting.

Bryan: How should investors adjust their expectations? It feels safe to jump into the funds that have the best past returns. But as you alluded to, that's not necessarily a recipe for winning performance results. How should investors think about adjusting their expectations?

Brightman: If you were going to base your investment decisions upon past performance, you'd be better off choosing the segments of the market that had lower returns. But that's very unrealistic. Instead, you can actually use some fairly intuitive tools to forecast the future returns. Now, that sounds very difficult. How am I going to know what the future is? But there are a couple of empirical observations. So, it's sort of patterns in historical data that help us forecast future returns, value and momentum.

What we really want is strategies that are cheap, likely because they have underperformed by a significant amount over a prolonged multiyear period, but have recently turned up. And we want to avoid strategies that have become expensive likely because they have been performing extraordinarily well over a multiyear horizon and particularly, right after they have turned down. So, if you combine long-run mean reversion or valuation measures with short-term momentum, you can actually forecast which strategy is going to provide superior returns looking through the windshield instead of the rearview mirror.

Bryan: If we were to do that right now and look at where valuations stand in relative momentum, where should investors be thinking about increasing their allocations? What type of smart beta strategies are priced to give you the best performance going forward do you think?

Brightman: Let me answer that in two dimensions; one in terms of style and the other geography. In terms of style, the one strong result that I would call attention to today is the expensive pricing of low volatility strategies, and they have negative momentum. So, this is probably, the most significant warning is watch out for low volatility.

Now, if you have a long-term strategic allocation and you intend to stick with it and you're not looking to do timing of your styles, there is no reason to abandon your low-volatility portfolios. But don't jump into a low-volatility strategy thinking that you are going to receive in the future the sort of outperformance relative to the market that you had in the past, because that was driven mostly by rising valuations--not sustainable, more likely will get some pullback there. That's in terms of style.

In terms of geography, here we have even greater confidence that the U.S. is an extraordinarily expensive market, and investors would do well to allocate, whether through smart beta, strategic beta or any other means of managing an equity portfolio, to develop ex-U.S. and emerging market equities. Again, always hold a diversified portfolio, but now is a wonderful time to diversify away from the U.S. equity market increasing exposure to EM and developed ex-U.S.

Bryan: All right. Always great insights. Thank you so much for being here to share them with us.

Brightman: You're welcome.

Bryan: For Morningstar, I'm Alex Bryan.