Tue, 23 Sep 2014
Low-vol strategies may limit investors' upside in a bull market, but their truncated downside risk may be worth the trade-off in the long run, says S&P Dow Jones Indices' Craig Lazzara.
Mike Rawson: Hi, I'm Mike Rawson with Morningstar. I'm here today at the Morningstar ETF Conference, and I am joined by Craig Lazzara. Craig is the global head of index investment strategy with S&P Dow Jones Indices.
Craig, thanks for joining us.
Craig Lazzara: Thank you. Happy to be here.
Rawson: Craig, ETFs have gained popularity, in part, because passive investing is becoming more popular but also because investors are interested in nontraditional types of indexes. Can you talk about some of these nontraditional indexes and what investors are using them for?
Lazzara: I think the way to think about that question, Mike--and it's a great question--is to frame it in terms of the way the index business has evolved. In the beginning, which means 30 or 40 years ago, the index world was limited to what I would classify as first-generation indices--or things like the S&P 500, the Russell 1000 in the U.S., IFA internationally--all of which are cap weighted indices designed to be a proxy for an asset class. (In this case, it would be the equity assets.) And they were, as you said, a very good way for people to gain exposure to the market. Reliably, over time, most active managers underperform them; so, they are a viable strategy. And that's where indexing became popular years ago.
And then, as that popularity grew, there was a second-generation of indices; I'll call them the extensions and subdivisions of the first: Sectors of the S&P 500, or individual countries, or growth and value divisions are some that are popular. But again, they are all cap weighted and designed to represent a proxy instead of the subset of an asset class.
You come to third-generation--what we like to call factor indices. I know you at Morningstar like to say strategic beta, which is a perfectly good term. Think of factor indices as vehicles that are designed to give you exposure to a pattern of returns that you as an investor might find congenial.
Another way to say that--and it's not a very succinct way--is to say that factor indices or strategic beta gives you access to patterns of return that in former years you would have had to pay an active manager active fees to get. And now, you can "indicize" those strategies in order to have access, more efficiently, to patterns of return or factors of return that an investor finds useful.
Rawson: So, you're getting exposure to factors that have been associated with excess return historically in the past, and some of these factors that have become popular recently. Indexes have been developed and funds have been developed based on these factors such as low volatility. What kind of exposure are you getting with low volatility? How are these low-volatility funds intended to be used?
Lazzara: Low volatility is a very interesting factor. The academic research goes back to at least the early '70s because one of the things that the capital asset pricing model predicts is that the return of a stock should be proportionate to its beta, its systematic risk. And it turns out that's not a very good prediction for high levels of beta.
So, the initial work, I think, was done by Fischer Black and some colleagues who were concerned about this challenge to the new capital asset pricing model. And they had some explanations for why this effect took place--where lower-volatility or lower-beta stocks seem to do relatively better than you would predict compared with higher-volatility or higher-beta stocks. And there are some other explanations other than the one they came up with.
But what we do in our low-volatility indices is to own--for example, in the case of S&P 500 Low Volatility--the 100 least volatile stocks in the S&P 500 and weight them inverse to volatility. So, a stock with a vol of 20 gets half the weight of a stock of a vol of 10. And hold that for a quarter and then rebalance. There are two things that are worth mentioning about that particular approach to the low-vol factor. One is that volatility tends to persist over time. So, stocks that had low volatility for the last year, typically have low volatility going forward. So, if you want low vol going forward, buy whatever was low vol in the past; that's not a bad way to do it.
More importantly, what low volatility in the future means is that, in a declining market or a weak market, you are likely not to decline as much. So, you get protection. It's not absolute protection, certainly; it's not portfolio insurance in any sense. But you get protection from declining markets, and at the same time, you get participation in rising markets. Again, that is not full participation--you typically underperform. We've analyzed this quite systematically, if you look at the history of our S&P 500 low vol, for example. Look at the negative months--the months where the market was down--and split those in half. So, that you have the most negative and then the smaller negatives.
What you find is in the worst bucket--the biggest negative months--roughly 85% to 90% of the time low vol outperforms the S&P 500. In the smaller negatives about 70% of the time. Small positives about 50-50. But then when you get to larger positives, low vol lags.
Now, what we say about that to potential investors who use low vol, to put it bluntly, is--and I say this to financial advisors all the time--if you have a client who when the market is up 30% one year and you're only up 20% and that client's going to be mad at you and fire you, don't do this because that's what low vol is designed to do. The reason it produces excess returns over time--or one of the reasons--is it truncates the downside. The price for truncating the downside is truncating the upside.
So, low vols are a good example of what I meant earlier when I said "finding patterns of return that an investor finds congenial." Because there are many investors for whom that pattern of protection or participation doesn't matter. If you're, say, a university endowment with a potentially infinite time horizon--or at least a very long time horizon--the market's down 20% next year? Big deal. You'll reinvest it at lower prices; you'll make it up over time. That's not a problem for you.
If you're planning to retire in five years and the market goes down 20% next year, that may be an issue. Some investors, particularly individual investors (because they are mortal beings with mortal, finite time horizons), find the truncation of the tails of the return distributions that low vol gives you a very congenial pattern of returns.
Rawson: Craig, your group puts out a report called on the SPIVA report; it's a very popular report. It's called S&P Indices Versus Active. Can you talk about the refinings of that report and what it sets out to do?
Lazzara: It's a report we've done for about a dozen years now in the U.S.; we've expanded it internationally as well. What we do is take the CRSP [Center for Research in Security Prices] database of survivorship, bias-adjusted mutual funds. We look at all mutual funds in the U.S., and then we split them down into equity and bonds. We split the equities into large cap, mid-cap, small cap, growth, value; there are lots of different ways in which this can be done. Then we'll examine, for example, for the case of U.S. large-cap managers: What percentage outperforms the S&P 500? What percentage underperforms? Mid-cap S&P 400? One of our international benchmarks? Fixed income against the bond index and so forth. Growth against growth. Value against value.
So, you're comparing each category of mutual funds to an appropriate benchmark for that style. There are some exceptions; but very reliably, year after year, the majority of active managers underperform a benchmark that is relevant for their style. And that's true whether the market is going up or down. That's true for large, mid, and small. That's true for growth and value. It's true for international versus U.S. So, it's a very reliable finding.
Now, it's not surprising if you think about the fact that all investors in the world--individual and institutional--all the investors own all of the assets. So, the aggregate return of all the investors has to be the aggregate return of all the assets. That is to say all investors as a group collectively earn the market return--which means the only way for me to be above average is for you to be below average. And the only way for a certain group to be above average is for another group to be below average. And the amount by which the winners are above average is exactly equal--before costs--to the amount by which the losers are below average. But since the costs of active management are so much higher than the costs of passive management, that means that the average actively managed dollar must underperform the average passively managed dollar.
Bill Sharpe wrote an article 23 years ago called the "The Arithmetic of Active Management," and that was basically the conclusion. The average actively managed dollar must underperform, and that's not a function of market efficiency. It's not a function of active managers being crooked or lazy or venal; it's just arithmetic. Lake Wobegon is a fantasy world. We can't all be above average.
Rawson: Absolutely, Craig. We're here at an ETF Conference, and we know ETFs in general are passive vehicles at low cost. Now, we're starting to see more and more ETFs that are adopting strategies. And, as we talked about, a lot of these strategies are at lower price points than traditional active management was. So, it will be interesting to see the SPIVA study as we extend it out and compare these strategy indexes to purely passive, plain-vanilla indexes, [keeping in mind] the fact that these strategy index tends to be lower cost.
Lazzara: What you would find, I think, over time is that [the conclusions of] "The Arithmetic of Active Management" applies to index designers as well as active managers. So, what that means is, if you think you have a strategy index or a strategic-beta index that outperforms over time, you may well be right. But somebody else has to underperform for that to happen. Low volatility is a great example because if it is true that there is a low-volatility anomaly, as the professors call it--it seems like it shouldn't exist but it does--the reason the anomaly exists is because high-volatility, high-beta stocks underperform your expectations given their level of risk. And it's avoiding those stocks that makes low volatility an excess return generator over time.
Well, if that's what makes low volatility work, the same thing makes high beta not work--in the sense of not generating excess return over time. Now, that doesn't mean you can't have a year, but 2013 was a great example of a year when high beta did much better--because high beta, unlike low vol, is designed to accentuate the moves of the market. So, if the market is up 30%, high bet is up 40%. But if you had to pick a holding for a long period time, I wouldn't pick high beta.
Rawson: Craig, thanks for joining me.
Lazzara: Pleasure to be here. Thank you, Mike.
Rawson: For Morningstar, I'm Mike Rawson.