Mark Bennett: This has clearly been a challenging period for active managers, especially if you’re underweight or don't own any of the Mag 7. Over the past five years the S&P 500—a passive index—has performed in the top quartile of an actively managed universe of U.S. large cap stocks. Using an industry standard global benchmark, the MSCI ACWI, the U.S. weight is north of 65 percent. The next largest country in the benchmark is Japan at 5 percent. Clearly, this is an issue affecting all active managers, both U.S. and global.
John, let me start with you because you've done some really interesting work disaggregating this index’s construction. Weights are as high as they've ever been. But you've looked at another measure, what’s termed “volatility share,” to describe the issue.
John Thorndike: While the index is quite concentrated, to us it felt like it was even more concentrated than it looked. What we wanted to find out is how much of the index’s volatility is driven by the mega-caps. To do this, you have to take into account a stock's weight, the stock's volatility and its correlation to the index. And when you do that math, you end up with each stock's contribution to the overall index volatility. If you divide that by the index volatility, you get their volatility share. So, correlation times volatility divided by market volatility is what we all call beta. Your volatility share is your weight multiplied by your beta. As we did that work, we went back and looked at the concentration of the index over the last 50 years because that's the period in which investors have been able to index, index funds having started in the early to mid-1970s.
So, this concept of wanting to buy the market, not doing any work on individual companies, just getting broadly diversified exposure was generally acceptable. But the index was at 35 percent in the top 10 names, which is like a 99th percentile reading. As we looked at these stocks’ contribution to volatility, it wasn't 35 percent, it was 45 percent. So, you had almost half of the market's volatility being driven by just a few names. Is that giving you the diversification that you want?
One of the takeaways from this analysis was that there's only been one time in history when mega-caps have been as volatile and as correlated with one another and that was the tech bubble. But there are differences. You can have stocks that by themselves are highly volatile. You really got that during the tech bubble. Or you can have these stocks all behave much like one another, so as not get any diversification within that group. They can be volatile and closely correlated. Today what you get is a mix that doesn’t look extreme on either metric, but the combination is quite extreme. You have these stocks that are pretty volatile, and because they share this one common theme of artificial intelligence, so they behave a lot like each other.
‘Extension strategies are a great way to diversify your active risk … what fiduciaries need to ask is, how much do I only care about tracking error and how much do I care about the absolute risk in my portfolio?’
— John Thorndike, GMO
Bennett: So, the conclusion is that the performance of the benchmark that is basically the industry standard for most U.S. managers is not behaving like a diversified market proxy the way it should. And probably even more importantly from a passive allocation standpoint, this is not the market portfolio that others might suggest it should be or has been historically. So, Jeff, as someone who runs a traditional long-only portfolio, how has the weight of these securities individually or as a group impacted your decision making?
Jeff Blumberg: We try to invest in companies without a benchmark in mind. And for a large period of our history, we were almost blissfully unaware of the index. I say that a bit tongue in cheek because, of course, we knew what the big index constituents were. But when we had a strong year, there would invariably be stocks in the market that we missed that were up 100 percent. Yet we were still able to have very good years and not think too much about what we didn't own. In the last couple of years that's changed given the likes of NVIDIA and others.
We've had to be more concerned about what we don't own, but we still try to adhere to our core principle, which is investing in what we like and not because of what it represents in the marketplace. In other words, we want to have multiple return drivers at any one point in our portfolio. We don't want to have more than a 5 percent – 6 percent weight in any single name. Yes, we've owned some names like Meta or Google, but we've owned a lot of other things too.
Bennett: Peter, share your thoughts on the difficulty with portfolio construction when 30 percent of your risk budget is taken up with these large index positions.
Peter Stournaras: Amid all the industry discussion about narrowness, try to retain these two facts. If you didn't own the five biggest contributors to the benchmark from 2002 to 2018 it cost you, on average, about 150 to 200 basis points annually. You can make that up. In the last five years, it's been more like 500 basis points, and in four of the five years it has been over 700 basis points. That's a lot harder to make up. If you missed out on the five biggest contributors to the benchmark over the last five years, you're sometimes starting out 850 basis points behind. When I started in this industry, you were competing against your peer group. Now, we’re competing against benchmarks. We’ve been using a long-short approach to portfolio construction: making bets where we feel we have an information advantage, minimizing bets elsewhere. We think that’s how to approach portfolio construction.
‘We want to focus where we feel we have an information advantage. When the concentration gets extreme, there needs to be a recognition that active share is going to decline.’
— Peter Stournaras, T. Rowe Price
Some clients have been getting behind active share1. How does that work against a concentrated benchmark? The only opportunity the managers have is to underweight the mega-caps, which is a tough approach. We want to focus where we feel we have an information advantage. When the concentration gets extreme, there needs to be a recognition that active share is going to decline, or you have to look at alternatives like active extension.
Bennett: Before we get into active extension, let me ask: Can you have an information advantage in Mag 7 names? Or is it better to treat them as a bucket?
Blumberg: With large- and mega-cap names we don’t necessarily feel you can have an information advantage. We’re relying more on experience and good judgment. In fact, with big companies today there is a deluge of information. What makes an investor successful is parsing that information to find the one or two things that really make a difference.
Bennett: John, curious to know if you have a view on this question.
Thorndike: I think you can certainly exhibit a behavioral advantage when it comes to these companies. Days of extreme volatility should dispel the notion that just because they’re big they're efficiently priced. We believe we can get analytical and behavioral advantages using quantitative tools because they bring discipline to what we do.
Stournaras: We treat the Mag 7 as a sector. In a similar way we treat AI, bitcoin and quantum computing as themes or sectors to make sure that we're not making a huge bet.
CAPSULE SUMMARIES OF PARTICIPATING MANAGERSEgerton Capital is a London-based private partnership managing a long-short fund and a long-only fund with long positions identical in both. Egerton invests globally without a particular style bias but describes growth at a reasonable price as an appropriate description of its approach. The firm has $16 billion in assets under management.GMO is a Boston-based asset manager with some $65 billion in AUM across equities, fixed income and liquid alternatives. The firm primarily serves institutional and private wealth management clients and describes its approach as valuation and quality oriented.T. Rowe Price is a publicly owned, Baltimore-based firm with a total of $1.6 trillion in AUM across a wide range of asset classes and strategies, including equities, fixed income, multi-asset funds and alternatives. The firm serves both individual and institutional clients.
Bennett: Let’s get back to structures and strategies that can help mitigate some of the concentration. For a long while Commonfund has been a big believer in active extension strategies2. TRowe runs extension strategies as does GMO. Peter, can you describe the mechanics of extension and what the benefits are. Risk bucketing would be relevant as well.
Stournaras: With the constraint of investing long only, managers are handcuffed against benchmarks that are more concentrated. They’re not able to benefit from the active share that they could have based on information advantage or whatever edge their process may offer. So, an extension strategy like a 130/30, where you're 130 percent long and 30 percent short, allows you to, first, take advantage of information where you have negative views on stocks that have a smaller weight in the benchmark. And in the current environment, almost everything has a smaller benchmark weight because of the mega-caps. Second, it allows you to increase your active share. You're not captive to the benchmark where so much of your capital is not captured by benchmark weight. So, if I am only going to be plus or minus 100 basis points in NVIDIA and I’m running a large-cap growth portfolio, that means 8 percent of my capital is tied to NVIDIA regardless. The active extension allows you to raise your active share, take advantage of your skill and increase the breadth of your skill leading to a more efficient portfolio.
Bennett: At the same time, you have enough of a risk budget to properly run a diversified portfolio with tracking error. In fact, one of our managers in effect views the Mag 7 as a risk bucket running an extension usually around 125/25. You can just index that 34 percent of the index accounted for by the Mag 7 and then, with extension, have enough of a risk budget to generate a diversified return stream and the potential for alpha.
Thorndike: I completely agree that active extension strategies are a great way to diversify your active risk. It's a better mouse trap by far than a long-only constraint. You still take as a given the risk of your benchmark as something that you want. The question I think fiduciaries need to ask is, how much do I only care about tracking error and how much do I care about the absolute risk in my portfolio? If what you care about entirely is tracking error, going to active extension is a better mousetrap. If you care about the absolute risk of your portfolio, then you have to look at the risk of the benchmark that you're starting with and ask if it helps you achieve your objective.
Bennett: John, you’ve done some interesting work from a from a market expectation standpoint in terms of what the markets are implying these Mag 7 names need to do from an outperformance standpoint to justify their weights in the index.
‘With large- and mega-cap names we don’t necessarily feel you can have an information advantage. What makes an investor successful is parsing that information to find the one or two things that really make a difference.’
— Jeff Blumberg, Egerton Capital
Thorndike: When you look at the market and you think it's doing something weird, a good question to ask is, what do I have to conclude that the market is acting rationally? So, we modeled the index as a two-asset portfolio: the Mag 7 and the S&P 493. If the weights in the market at any given point were actually optimal—say, from a mean variance perspective—what does the market have to believe the expected return spread is for the Mag 7 relative to the rest of the market? How much return advantage do those mega-cap stocks have to deliver relative to the rest of the market to justify from a mean variance perspective their weight in the market at any given point in time?
Historically, for mega-cap stocks, that spread has been negative much of the time. You could have believed that the mega-caps were going to underperform the market. They were more stable. They're big companies. Maybe you give up a little bit of return in order to reduce risk. About 25 years ago, you had to believe that the mega-caps stocks—the biggest stocks in the tech bubble—were going to outperform the market by 800 basis points to justify their weight. When we did the work again late last year, that number was close to 600. Basically, these were the only two times when the market was saying these guys are going to crush the rest of the world and, therefore, we're willing to allocate a ton of risk to them. If you believe they're going to outperform by 600 basis points from here that’s fine but, recognize that you have a point of view.
In other words, you're not passively allocating and saying, I don't know anything, I'm just taking what I get, everybody's about the same. What you’re doing is investing in the market with a very strong view that these stocks are going to continue to outperform despite doing something that’s very different today from what they’ve done in the past. And that difference is investing real money, billions of dollars, in capex. And for me, that just raises a ton of questions about future returns.
Bennett: To stay on that for a second … to be fair, that measure has been elevated for a couple of years now. And they've delivered.
Thorndike: They have delivered. But the capex has been creeping up.
Bennett: There seems to be a change in the business model. The Mag 7 have been great free cash flow generators and they’ve been light on capital intensity. But that’s changing—Microsoft gives guidance and they talk about investing $85 billion. So, the Mag 7 have outperformed, but the question has to be asked: Are we at an inflection point or not? Time will tell … history will tell.