Arnott: There's Some 'Silliness' in Smart-Beta Land

Johnson: So this is a space that has I think historically been somewhat definitionally challenged even with respect to moniker, and we played a role in that as well at Morningstar--smart beta, strategic beta, factor-based investing... You would argue that there are important differences that investors should be aware of between what you deem "smart beta" and what others might think of more traditionally as factor investing. What are those differences and why are they important to understand? 

Arnott: Some folks have described me as the godfather of smart beta, I'm kind of flattered, although the mafia connotation, maybe not. But fundamental index opened the door for smart beta. Fundamental index doesn't win because it ties you into the fundamentals of a company, it wins because you're not weighting the company in proportion to the share price. Breaking that link means you're not automatically overweight the overvalued and underweight the undervalued.

So, Towers Watson coined the expression "smart beta" to encompass an array of strategies that do the same thing. You've got equal weight, you've got minimum variance and low-vol strategies, you've got fundamental index, you've got EDHEC-Risk Efficient strategy, TOBAM's Maximum Diversification strategy, every one of these weights companies in a fashion that pays no attention at all to the price or the market capitalization. Fundamental index is unique on that list in that because you're weighting companies by how big they are, you have vast capacity, you have a bias towards large, liquid companies, but not towards high-priced companies.

So, everybody wants to be smart, so everybody wants to say "We do smart beta, too." So the definition, the umbrella of smart beta has been stretched and stretched to encompass all sorts of things. If smart beta encompasses everything then the term means nothing. Factor tilts, I would argue, are not smart beta. Factor tilt strategies start with cap weight and then put on a factor tilt. So you start with a portfolio that systematically automatically overweights the overvalued and underweights the undervalued--can't break that link without breaking that link. And introducing a factor tilt, it's got to be a pretty extreme factor tilt to wind up breaking that link. We know that linking the weight to the price costs you about 2%, cap weighting performs about 2% worse than the valuation in different indexes. So you're starting out 2% behind the eight ball and you're hoping your factor tilts are big enough to get you past that hump.

So I would argue that factor tilts aren't smart beta. I'm losing that argument. The marketplace is saying smart beta compasses all of these things. Not so sure. It now encompasses a whole lot of ideas that I think aren't strictly smart beta, and it encompasses a few ideas that I think are actually not smart at all.

Johnson: But crucially important in all of this, especially as you see proliferation within the space it seems, not to put words in your mouth, but that index construction matters and that investors should understand how the indices that underlie the funds, if they're investing in an index fund or an ETF, are built and what that means for the risk and return characteristics of that portfolio.

Arnott: Right. Early days of smart beta I was proud to be opening the door for an array of other competing strategies that were also breaking the link with price and also structured where they would have high odds of adding 1% to 2% a year.

Johnson: Cool.

Arnott: With product proliferation I've become alarmed, and I'm alarmed not that smart beta is becoming stupid, but that smart beta encompasses a whole array of strategies, a few of which aren't smart, and invites people to performance-chase. "My smart beta is better than your smart beta. Look at the 10-year numbers. Oh, sure, it's only a paper portfolio, but the 10-year numbers are awesome." So people look at that and they say, "I want this one because it's got better numbers." That's not the way to choose a successful investment. Facebook has wonderful performance since its IPO, so should we create a Facebook factor, long Facebook, short everything else and declare that because the alpha is significant, that it's got an alpha going forward? Come on! And you have that kind of silliness happening in smart beta land to some extent.

Johnson: Well Rob, I want to thank you again for being here today and for a fantastic discussion. You're welcome back any time.

Arnott: Thank you very much. It's a privilege.

Johnson: With Morningstar, I'm Ben Johnson.