Expanding on Modern Portfolio Theory

Scott Burns: Hi, there. I am Scott Burns, Morningstar's director of ETF research, coming to you live from Morningstar's premiere ETF Invest Conference.

Today, I am joined by SSgA's Dan Farley, who is the global head of investments for the Multi Asset Class Solutions team at SSgA, a team that oversees over $190 billion of institutional money worldwide. Dan, thanks for joining me.

Dan Farley: Thanks for having me.

Burns: So you have recently written some – an article about Modern Portfolio Theory and the evolution of it. As a practitioner of strategic asset allocation, I mean that does kind of fly in the face of Modern Portfolio Theory a bit. Maybe you can talk to us about what it is that you are seeing in this time, this age-old, oft-held process of MPT and mean variance, you know asset allocation, and how SSgA is really kind of differing from that road?

Farley: Sure. I think it's more of a question of evolution as opposed to a brand new path that we are looking at. What we are trying to focus on is to think about how people have used the tools of Modern Portfolio Theory and the efficient portfolio construction and to challenge a little bit to think about what was missed, how could we think about risk in perhaps an expanded light.

So traditionally, when practitioners are building portfolios, they are looking at risk, return, and correlation of various asset classes in the portfolio, and those were all good things to look at. However, oftentimes it is a snapshot in time, historical view of those things. And so, for example, correlation, there was much written and concern about the fact that in the crisis in the last couple of years, correlations went to one, and diversification doesn't work.

Burns: Right. That always happens in crisis.

Farley: That always happens in crisis. So, as a practitioner…

Burns: I have read my books.

Farley: There you go. As a practitioner to be thinking about how do we build portfolios, to take the long-term view, very important, but also to be thinking about how can we test those portfolios in those more crisis environments, such that the investors and the managers will have a better understanding about where the risks are in their portfolio.

<TRANSCRIPT>

And so to think about that in a more dynamic framework, I think, is important and I think probably gets lost in plugging in numbers into an optimization software and an answer spits out, and all of a sudden that must be the right answer, right?

So understanding the tool I think is where we go forward. And quite frankly, Markowitz had said the same thing, that the structure he put together now decades ago was simply just a structure, right?

I think the other thing that really some time needs to be spent on is the fact that standard deviation is not the sum of all fears. There are other risks that people need to be thinking about, and things about liquidity, for example, clearly lost and forgotten, and when we look at some of the lockups that happened, particularly in the alternative space, that people were shocked, and all of a sudden, they couldn't get their money out. But liquidity is a real risk.

Burns: And then there is real value to that liquidity, too, and I think that's where a lot of advisors learned from their investors. So I showed you the standard deviation and it fit your risk profile, but what we didn't have was a liquidity risk profile and that was all new information to us.

Farley: Understanding that and you know that and to think about how currency impacts the portfolio and should you be hedged or not. All these things are typically things that are not optimizable, but need to be understood and need to be brought into the portfolio construction process.

And those things that have perhaps less liquidity either need to be adjusted in the mathematics of it at all or there needs to be a part of the portfolio dedicated to that space that would be a liquidity buffer. And so that if you needed to get to it, you could exercise some cash relatively quickly.

Burns: Right. And you know that's one of the things just I am assuming, in your practice you guys are using alternatives, including hedge funds and things like that. How much of those hedge fund returns over the past 10 years would you say were really a lot of liquidity risk that was masquerading as alpha, and how does your group now kind of approach that?

Farley: Sure. I think that there has been a number of things that have happened from a hedge fund return perspective, even to the point where people say, well, a hedge fund was just a bunch of beta and I could have bought that more cheaply. It depends on the portfolio construction because, quite frankly, depending on the strategy, you want that manager to be simply moving to all the right betas at the right times, right. That's, okay. That's a good strategy. It's just a question if it's always just simply taking advantage of, say, the small-cap premium or some other risk premiums, junk bond, constantly.

Burns: Junk bond, or muni bond, illiquidity...

Farley: Exactly. I think what we have really tried to focus on is to think about how different strategies generate the returns. And so, for example, we will look and our hedge fund team will look at hedge managers and put them into camps. One that they would call convergent, which means whatever their style is, the manager has identified a fair value price for the currency, the bond, the stock, whatever it is, and looks at where the current market value is, and those two things should converge, right.

So that could be fixed-income arbitrage, equity arbitrage, you know merger arbitrage. Everything within arbitrage usually is a convergent strategy as composed to things that are divergent. Things like commodity trading accounts or global macro, sometimes tactical asset allocation depending on the process, momentum. Those things tend to do poorly when the market is going straight up. They tend to be a drag on your returns in very, very good years.

However, with some level of allocation to those kind of divergent strategies they provide you diversification or it puts the hedge in the hedge funds, so to speak, to protect you in that environment. So we have spent a lot of time thinking about how those two things behave and how you want to combine them in terms of a portfolio construction. And I think that's true whether we are talking about traditional asset classes or alternatives. It's really important to understand, not just necessarily if they are quant versus fundamental, but really what are the drivers of the returns, how are they generated, and when will they do well, and most importantly, when won't they work.

Burns: Right. And how do they all work together when you start lumping these different strategies together?

Farley: I think one of the things if you look at, particularly in the hedge fund or funds space was that the managers felt they had a lot of diversifications, they had a lot of different styles of markets that they played in, but they were all convergent styles. And if the concept of fair value goes out of the window, then none of them work, whether it was a stock fair value or a bond fair value or currency.

Burns: So you diversified your manager risk but not really your asset class, your beta risk or strategy risk for that matter?

Farley: It's a good way to think about it, yes.

Burns: All right. Well Dan, thanks again for your insights, and thanks for joining us.

Farley: My pleasure. Thank you.

Burns: And I am Scott Burns with Morningstar. Thanks for joining us.