Protect Yourself from High Correlation

06/04/2012 7:00 am EST


Zephyr Associates’ Marc Odo discusses his firm’s asset-allocation software for advisors, and how it helps financial professionals balance risk and return for the best degree of diversification.

Kate Stalter: Today’s guest is Marc Odo, director of applied research at Zephyr Associates. Marc, I wanted to start out with the big picture today, and ask you just to describe Zephyr. What exactly is it that you do over there?

Marc Odo: Certainly. Zephyr produces a couple of pieces of software that help people with both setting the broad strategic asset allocation, as well as choosing the individual investments that they would want to use to populate that plan, once it’s in place. So if you are looking for mutual funds, ETFs, or different managed products to help fill that role in your asset allocation plan, Zephyr assists with both pieces of that process.

Kate Stalter: From what I understand, quantitative analysis factors into your methodologies. Can you say something about that?

Marc Odo: That is correct. Traditionally when people have set about setting a long-term strategic plan, they have used an asset-allocation process that was meant to diversify their assets across different asset classes.

The problem is: In the most recent credit crisis, people were kind of disappointed in the results of that type of strategic plan, and maybe didn’t dampen the risk as much as they had thought. So there certainly have been lots of new innovations, lots of new ways, people kind of revisiting those core assumptions of the whole asset-allocation process.

Kate Stalter: Now, when people think of using quantitative methods to construct portfolios, one of the things you first think about is historical data, either from the market or from individual stocks or other asset classes. Talk about some of the inputs that you use other than, or in addition to, some of this historical data.

Marc Odo: Certainly, I’d be happy to. When it comes to asset allocation and diversification, I think the overall goal is still valid. People essentially want to maximize return and minimize risk. That is still a valid objective.

The question becomes, like you mentioned there, where are these inputs coming from? Now, sometimes people just relied on the straight historical information and kind of extrapolated that into the future, saying “Well, you know, the returns over the last couple of decades have been in the double digits, so we’re going to use that going forward.”

It’s a bit unrealistic to assume that, with bond yields being as low as they are and stocks facing a challenging environment. You can’t just rely straight on historical performance. Some people say it’s like driving looking only in the rearview mirror.

So there are some more sophisticated models out there. I’m not going to go into great detail, but one of them is known as the Black-Litterman model, which kind of puts current market conditions into the context of our forward-looking assumptions. And, again, that is focused on the return element of the whole asset-allocation plan.

Kate Stalter: Let’s talk a little bit, then, about the practicalities of taking some of these different inputs, and creating these models, and how you might actually put together some portfolios given everything you are saying here, Marc, about the high correlation across different asset classes. How does this actually work when the rubber meets the road?

Marc Odo: Yes. Again, the intent is to kind of balance the return versus risk, and hope that by looking at diversified assets, you are producing uncorrelated return patterns, which essentially means if one thing is doing poorly, you have got something out to help balance it out. It is this whole basic idea of not putting all of your eggs in the same basket.

The problem is when the markets crash…like in 2008, it seemed like everything was going down all at once. So traditionally, the way that people have done that, historically, is that they have diversified across different asset classes, so maybe a little bit in bonds, a little bit in US, a little bit internationally, a little bit in commodities.

But what we have been seeing is that people these days, because that didn’t work so well in 2008, are looking at different types of products now. So they are seeking out their diversification by hiring money managers or different mutual funds or ETFs that are just doing things differently from the crowd, that are not plain vanilla-type managers.

A couple that come to mind.. there is the First Eagle Global Fund (SGENX), and that particular manager has a mandate to just kind of go where they see opportunities. Over the last couple of years, that has worked very well for them. They have got a quite a bit of assets following that kind of go-anywhere strategy.

Another one would be the Stadion Managed Portfolio (ETFFX) out of Athens, Ga., and what they do is a little bit more focused on downside protection. When the markets go up, they are probably going to trail a lot more, but they have a very market-sensitive process that tries to avoid these big downturns we saw like in 2008 and 2009.

So to kind of sum up here, what we’re seeing is that some of the people are looking for products that kind of behave differently from the herd, and seek their correlation that way, because so many of these broad asset classes are behaving too similarly these days.

Kate Stalter: One of the topics, in addition to alternatives, that I seem to be hearing about a lot these days—and you’re alluding to today—is the idea of trend following. And I know that’s kind of a layman’s way of discussing what some of these portfolio constructions might be, but in some kind of broader sense, does trend following factor into any of your methodologies in any way?

Marc Odo: Not particularly with Zephyr’s methodologies. But certainly if you are seeking to identify managers that have been successful at doing that, there are different statistical techniques that people use Zephyr’s software for, to kind of identify those managed who have been successful at being ahead of the trend, so to speak—avoiding the downturns that we saw in 2008, being successful in the upmarket cycle. The program is rather robust and can identify managers who have been successful in those types of trend-following strategies.

Kate Stalter: It sounds to me that this software that you are talking about here today, Marc that people wouldn’t necessarily be using at home. But they might want to seek out an advisor who is using this. Would that be right?

Marc Odo: That is correct, that is correct. Zephyr’s software is in the Style Advisor and its sister program is Allocation Advisor, and it was designed for the financial professional, someone who is out there dispensing advice and has a real good handle on all these kind of advanced models and statistics, someone whose business is to kind of sort through all of the noise out there and come up with a clear signal when it comes to the market.

Related Reading:

Related Articles on STRATEGIES