The founder of AQR Capital Management discusses alternative strategies, private markets, tariffs, and the role of AI in investing.
Today on the podcast we welcome back Cliff Asness. Cliff is the founder, managing principal, and chief investment officer at AQR Capital Management. Cliff writes often about investing and financial matters on AQR’s website and has been a prolific researcher throughout his career, with his contributions appearing in many of the leading scholarly journals, including the Journal of Portfolio Management, Financial Analyst’s Journal, the Journal of Finance, and the Journal of Financial Economics. Before co-founding AQR, Cliff was a managing director and director of quantitative research for the asset management division of Goldman Sachs. Cliff Asness, welcome back to The Long View.
“Cliff Asness: Value Stocks Still Look Like a Bargain,” The Long View podcast, Morningstar.com, May 31, 2022.
“2035: An Allocator Looks Back Over the Last 10 Years,” by Cliff Asness, aqr.com, Jan. 2, 2025.
“(So) What If You Miss the Market’s N Best Days?” by Cliff Asness, aqr.com, June 5, 2025.
“Why Not 100% Equities,” by Cliff Asness, aqr.com, Feb. 12, 2024.
“Exceptional Expectations: US vs. Non-US Equities,” by Antti Ilmanen and Thomas Maloney, aqr.com, Q2 2025.
“In Praise of High-Volatility Alternatives,” by Cliff Asness, aqr.com, Sept. 4. 2024.
“Should Hedge Funds Hedge?: Why Some Alts Should Have a Beta of 1.0,” by Cliff Asness, aqr.com, March 28, 2025.
“We Have ‘Surrendered More to the Machines,’ Says Quant Fund Titan Cliff Asness,” by Costas Mourselas and Amelia Pollard, ft.com, June 3, 2025.
“CIO Perspectives: An Interview With Cliff Asness,” aqr.com, Sept. 1, 2024.
“AQR Launches the AQR Fusion Mutual Fund Series,” AQR Funds News, aqr.com, June 25, 2025.
(Please stay tuned for important disclosure information at the conclusion of this episode.)
Christine Benz: Hi and welcome to The Long View. I’m Christine Benz, director of personal finance and retirement planning for Morningstar.
Dan Lefkovitz: And I’m Dan Lefkovitz, Strategist for Morningstar Indexes.
Benz: Today on the podcast we welcome back Cliff Asness. Cliff is the founder, managing principal, and chief investment officer at AQR Capital Management. Cliff writes often about investing and financial matters on AQR’s website and has been a prolific researcher throughout his career, with his contributions appearing in many of the leading scholarly journals, including the Journal of Portfolio Management, Financial Analyst’s Journal, the Journal of Finance, and the Journal of Financial Economics. Before co-founding AQR, Cliff was a managing director and director of quantitative research for the asset management division of Goldman Sachs. Cliff Asness, welcome back to The Long View.
Cliff Asness: Thank you for having me.
Benz: Well, thanks for being here. We want to start by talking about macroeconomic issues. You have been a vocal critic of the Trump tariff policy. Of course, it’s been dialed back quite a bit, but do you think it will have short- or long-term implications for economic growth, inflation, anything else?
Asness: Well, first, I am a critic. I’m a critic of anyone’s tariff policy. This just happens to be Trump right now. I was critical when Biden kept a whole bunch of the Trump tariffs from the first go-around. I am in the traditional economists. This has virtually settled economics, to steal that term from science. Tariffs are taxes. I don’t care how many people say they’re not. They’re taxes on the buyer. Just effectively think of them as a sales tax. They reduce the global pie and global prosperity. They do cause some winners and losers, and that’s a political issue. I have no great genius how to solve that, but they make us all collectively poorer. If the strategy, and some claim this, if the Trump strategy is ultimately to get lower tariffs by threatening higher tariffs, I’m not an expert on the gamesmanship there. If that worked out at the end of the day, I’d tip my cap and say, thank you for making trade freer, but it doesn’t seem like where we’re going.
You are right. They’ve cut back, but I think this is a little classic, propose something truly beyond the pale and then make rather large increases that will have an impact seem moderate is kind of where we are now. So, I guess I’d say the impact to the global pie, to prosperity around the world, will be a lot smaller than the initial off-the-charts proposals, but I think the world will be collectively somewhat poorer for these hikes, and I’m kind of against that. I should tell you, you guys know I’m a financial economist. I largely deal with, is this stock better than this stock? I make no claims to be an expert macroeconomist, though I don’t think one needs to be in this case. This is the first three weeks of macro class.
Lefkovitz: A lot of investors out there are concerned about stagflation as a result of the tariffs. Won’t ask you for macro forecasting on the likelihood.
Asness: Hey, Dan, I didn’t say I won’t pontificate about it. I just said I’m not an expert.
Lefkovitz: Feel free to pontificate, but I guess I was going to ask you from an investment standpoint if you have a view on what types of assets or what types of strategies that you’d expect to fare well in such an environment?
Asness: Sure. So if we get stagflation, and I won’t take a strong stance on that, I think tariffs ultimately, they haven’t shown up yet, but they almost have to add somewhat to inflation. They almost have to slow down growth somewhat. I don’t know if that’ll ever qualify for full stagflation. I’ll let other people worry about that. But you’ve basically, very early in our conversation, threw me a hanging curve ball. This is a question I can only answer in a deeply self-serving manner, because traditional assets, stocks and bonds, the one thing they all agree on is that they do not like stagflation. Stagflation is obviously bad for bonds because it’s all about the inflation part and is quite obviously bad for stocks. Picking up inflation with slowing growth is nobody’s favorite environment. I do think, in that case—again, I’m talking about my pocketbook—but I always believe in an allocation to relatively uncorrelated alternatives. I think they would particularly be important in a world where every type of traditional asset is being impacted. That doesn’t mean they make money like gangbusters.
An uncorrelated alternative is sometimes mistakenly oversold as a hedge. A hedge is something that’s short, the traditional. A diversifier is something that’s simply uncorrelated with it. So if you have a well-constructed uncorrelated source of return, you have your normal chance of doing well. If that’s two out of three years, you have your same chance, which is wonderful when everything else has no chance of doing well in that environment. But I wouldn’t want to oversell it. Maybe the one case where I would get fairly confident, never perfectly confident, but fairly confident, is in the alts world on the trend-following side. Both from the blue, stagflation that massively happened in a day would not be good for trend-following, because you have no trend to follow, but that is just not what we’ve ever seen. We’ve seen slow buildups in these things. And it’s no guarantee going forward, but trend-following has done particularly well in longer painful periods for traditional assets. So I appreciate that as the second question, because you need alts in that world, or at least alts are your one shot.
Benz: So I should note we’re taping this in mid-June. Cliff, just to follow up on that, would you say that the trend-following-type products would be, you would think, the most effective sort of product in terms of an uncorrelated asset in such an environment, a stagflationary environment?
Asness: In a long drawdown, a painful year, a 2022, even longer, yeah, I would say that. It’s not the only one I would allocate to. You get into trade-offs. Trend-following has historically had this property of doing well in very tough markets. The geeky phrase for that is “having some positive convexity to it.” But I don’t think it’s steady state risk-adjusted return. Sharpe ratio’s an imperfect measure; we use it as just a stand-in for whatever you think of return over risk. I think you could build a considerably higher Sharpe ratio portfolio of alts if you extend beyond trend-following. So I would look at a portfolio of both, but the simplest answer to your question is, yes, if we see an extended period of stagflation, the thing I’d have the most confidence in is probably trend-following.
Lefkovitz: Cliff, you recently wrote a piece on the common assertion that market-timing is a bad idea, because if you miss out on several of the markets’ good days that you erode your long-term return. What do you think are the flaws in that argument?
Asness: OK, well, let’s start off with I am not making a pro-market-timing concept. People sometimes make very bad arguments, and I’ll explain why I think this is a very bad argument. In the service of good, convincing people not to market-time is the Lord’s work, in my opinion. But this is a terrible way to do it, and it’s ubiquitous, and it’s been around forever. I didn’t recently write a piece on it. I recently revived a piece that I wrote 26 years ago and couldn’t get published. So I had some humble moments in there saying, they didn’t say this, but maybe they didn’t publish it because the writing was just terrible. I don’t know. You guys both write. You look back on something you wrote 26 years ago and you’re like, I was really that bad? But the economics, I think, were good.
So in stages, the typical thing, what if you missed the “n” best days? Let’s just say 10 days; I don’t want to keep saying “n.” If you missed the “n” best days for the market, you lose a lot of the return. OK, maybe interesting is a mathematical fact, but let me say this to you as a strategy. If 10 times over your investing lifetime, you take all your money, sell all your stocks and put them in cash to buy them back the next day, and those 10 times were the absolute worst 10 times ever to do so. Then that would be bad. It doesn’t strike me as a particularly interesting statement. That is a radically extreme strategy and to make it worse—and this is one thing my paper does. “Paper” is too exaggerated; it’s too simple to call it a “paper.” It deserves to be a blog. But run the exercise where what if I miss the 10 worst days for the market? The results are fairly symmetric. They’re actually slightly better because markets are negatively skewed.
They tend to go down a little bit more than they go up in the short term. Missing the worst—call it the same. I’m gilding the lily if I say it’s better. It has a huge impact, invest in the market the whole time, but manage to miss the 10 worst days. The returns blow up. Am I allowed to tell people you definitely should time the market? Because look at these results. If you only miss the 10 worst days, you, I don’t know, double your return. I’m making up the specific numbers. No, that would be a ridiculous argument. I think most people would see it. It’s a symmetrical argument. You should do anything in investing proportional to your belief and your skill at doing so after transactions costs, aftereffects on risk when you time the market, you tend to move to a more undiversified portfolio. If by definition your long-term strategic portfolio is your best bet at diversifying across good assets, a radical market-timing tends to make it undiversified.
So if you think your skill overcomes all these, you should do it proportional to your belief in your skills. If anyone believes their skill is so good that they should 10 times get out of the market for a day, then I think they need a lot of help and maybe beyond investing. So I just think this is kind of a silly way to frame the argument. Again, it may do the Lord’s work if it convinces people not to time because my actual personal belief is most people should not time the market. Most professionals should not time the market. It’s a very hard thing to do. We do a little bit of it, particularly in trend-following. Trend-following does take a position, but that’s about it. I don’t think we have great insight into the short-term moves of the market. So very good cause to keep the average investor from timing. Very, very bad argument in my opinion.
Benz: Another paper that you have been critical of is the one about whether investors should own 100% equity portfolios. And I feel like I’m hearing this drumbeat of an argument that investors should forget everything else and be 100% stocks. You offer an opposing view. Can you talk about that?
Asness: Sure. Well, first, you always hear these arguments when markets have been strong for a long time. Hey, why don’t you only do this? Let’s go back to basic theory. And when I say theory, I’m being complimentary, not disparaging. This is just basic math that if you take an MBA course in investing, it’s probably week two. In week two, the professor draws what’s called an efficient frontier. He estimates the expected return on the risk. Again, you could do simple models like volatility. You can get more subtle in what you think of is risk, but expect a return in risk for a whole bunch of assets, some assumptions about how they’re correlated, and then shows you a combination of diversifying assets generally has a higher risk-adjusted return than any single one. So, 60/40 equities bonds has had a higher Sharpe ratio than 100% equities; lower return because bonds return less. They’re supposed to return less. They’re a lower-risk asset, but a higher risk-adjusted return. So basic theory—and again, this is week two—is an investor who finds, let’s pretend 60/40 is the best combination you can make. We don’t think it is, but 60/40 I think legally we have to use 60/40 as the default because everyone has for like 50 years.
So imagine 60/40 is the best Sharpe ratio. Both stocks and bonds are great, but on their own, the return per unit of risk is lower. An investor who finds 60/40 too risky should add cash to it. And an investor who says, I can tolerate more risk than that should apply leverage. And the L word is scary to some people. I’ve written on this. The amount of leverage you need to apply to 60/40 to make a kind of equal risk to stocks is like 25%. It is not a particularly scary number. In all the history of market drawdowns, crashes, there was no bankruptcy risk to this. It wasn’t even close. So, lo and behold, investors who do this—and I wrote a paper like a million years ago, this was in the ‘90s. A lot of times when we talk about my papers, I feel really old because I’m always like, yeah, I wrote this in 1994. That was 31 years ago. Showing that a realistically, including some costs, mildly leveraged 60/40 portfolio, handily beat 100% equities, even in the US, a country with the strongest equity returns. Others have extended that at a sample; it’s worked at a sample. Others have done products based on this.
I’m not supposed to plug a competitors’ products, but my friends at WisdomTree have a product that I think does exactly—or maybe, I don’t know if it’s more aggressive—but does basically that paper from 31 years ago. So, I’m not dissing stocks. Stocks are a huge part of a strategic asset allocation. But any single asset class, any even more-narrow portfolio is simply leaving the benefits of diversification absolutely on the table. With leverage, you always have to think if someone tells you this is a great strategy, but you have to lever it 41 times—you have to take your 2-basis-point edge and lever it 41 times—I think you probably should go, yeah, I’m going to pass on that one. But mildly leveraged 60/40—again, I’m assuming, please don’t take 60/40 that literally—but mildly leveraged 60/40 is exactly what they teach you second week of your MBA program. And it has worked like a charm long term. So, I don’t see why someone would leave that on the table.
Lefkovitz: Sticking with asset allocation, Cliff, you published a piece coming into 2025. It was very clever. You framed it as looking back from the year 2035, looking back over the past 10 years of returns. I wonder if you can talk a little bit about the genesis of that piece and also how you arrive at the capital markets assumptions and especially the Fartcoin allocation. I wanted to ask you about that.
Asness: Well, a few things. First of all, I’m sure you know this, but the structure of how I did the piece was a thinly veiled way to say I was right about everything 10 years from now. It’s not even veiled. It’s being too nice to the paper to say there were allocations. I only spoke in generalities. This was supposed to be kind of half fun, but also what I actually think, in particularly the Fartcoin thing. And I do find it somewhat amazing that serious people like yourselves and myself actually end up discussing something called Fartcoin on a well-respected podcast. So, in some sense the creators have succeeded in getting us to do that. But I was just having fun there. I will admit I’m a crypto cynic. I am least cynical about bitcoin. That doesn’t mean I like it. I don’t want to go that far. But the leader in the category that does have at least a cap on how much can be created, unlike many of the other meme coins. I’m still not a buyer. I don’t think we need it.
The arguments against “fiat currency”—don’t lead you to want someone playing computer games all day to generate this made-up thing. I am a cynic about the whole thing. But if anyone’s going to succeed, it’s going to be bitcoin. That was the one thing in the piece that I just flat-out lied about. Everything else was my actual opinion. My opinion that bitcoin would do terribly, but Fartcoin would hold up, that was a pure joke. So you have found the one lie in the piece. I don’t think compliance is going to let me launch a meme coin here. But I want to launch a meme coin literally called Ponzi scheme. Ponzi coin. It’s the same idea. Let’s just throw the ridiculousness in people’s faces. But I believe, I hope somebody can steal this idea. I’m putting this into the public domain. I think they will still buy it. Half the reason people buy these things is transgressive joy at doing something ridiculous. So why not go all in and call it Ponzi coin?
Benz: We’ll see if someone picks up on that, Cliff. Want to follow up on our last conversation with you, which was three years ago at the Morningstar Investment Conference—was 2022 bad year for stocks, bad year for bonds. On the equity side, value did hold up a little better than growth in a tough equity market. I’m curious what’s your perspective is on what we’ve seen from value stocks since then.
Asness: OK, I love this. I appreciate it. This is the second time you’re giving me a little license to brag. Value as a concept we all understand. Pay a low price compared with fundamentals. I think the way quants use the term value is just price to fundamentals. The way active managers—Graham and Dodd type managers use the word value—it is a more holistic concept. They’ll look at how profitable, how risky it is. Quants do the same thing. They just call those other factors. I actually think quants misnamed the value factor like 30 years ago. It should have been called the low price to fundamentals factor. It’s not as piffy, but it’s more accurate. But I am a believer on average over the long term, low price to fundamentals—largely I think because of behavioral reasons. People go too far. The low price to fundamental stocks deserves to be low, but not as low as they are, and the high ones deserve to be high. So I am a believer in that general strategy, even though that is not all I would do by any means.
Now, in 2022, every form of value worked. In other years, it gets much more subtle. For instance, the traditional value indexes are cap-weighted. And we, of course, focus mostly on the United States. Both of those things have led to extreme outperformance of growth. The Magnificent Seven, if you will, they’ve had tough periods, but have trounced the rest of the market. And the traditional kind of indexes way to do it has a huge bet on that. If you are a quant geek forming a long-short portfolio with value as one factor, again, not the whole thing by any means, I won’t speak for every quant. But I think I’m speaking of most of them. You’re going to have something much closer to a thousand—not equal weight, but let’s make it simple, call them equal weight—a thousand stocks equally weighted long, not just in the US, where this phenomenon of the Magnificent Seven has been dominant.
You’re going to do this globally. And this may be more specific to some quants than others, but we have always favored not taking an industry bet when it comes to value. We found historically, and we wrote a paper on this. I’m going back even further. This is 1995—only 30 years ago—we wrote a paper showing that value and most other quant strategies outside of momentum don’t do a particularly good job at the industry decision; do better if you take that out. You do all those things. It still hasn’t been a banner period for values since 2022, but it’s actually held its own. If you do the traditional ways and the index ways, and you actually went short the Magnificent Seven, you have not had a very good time as a value manager of any kind since then. In normal times, these are very different strategies. This construction matters. I will say during wild blow off, plus or minus six months around a bubble peak, how you construct these things matter a lot less.
When value is kind of destruction, peaked or … I don’t know how you want to view that, but when the value destruction hit maximum in late 2020, it didn’t matter. Well-constructed—I can tell you I have scars on my back—well-constructed quant versions that hedge a whole bunch of things out. Don’t take a huge bet on any seven stocks or don’t take a huge industry bet, were in pain right along with the more traditional indexes. But long term, we think the risk-adjusted return is better to do it in the more in the more quantitative way. And in any period that’s not near a panic or a bubble, how you do it starts to matter a lot more.
Lefkovitz: Cliff, do you think that non-US equities are an indirect value play?
Asness: Yeah, the rest of the world is cheap compared with the US. Here, I’m using cheap like a quant, just considerably lower multiples. Obviously, there’s a level of US superiority in growth that can justify a higher multiple. But the US has definitely outperformed the world for a long, long time. Let’s call it a quarter of a century. I like saying that instead of 25 years, I think it has more gravitas. But I’ve written on this, Antti Ilmanen at AQR has written on this. You can do it different ways and get slightly different answers, but call it 80%, 85% of the US’ victory has come from multiple expansions. So pick your favorite one. Let’s use the Schiller Cape. I’m not claiming that’s the be-all, end-all method evaluation. And don’t hold me to the specific numbers, but if 25 years ago, the US was considerably cheaper than global stocks, it is now considerably more expensive on this measure. That has driven again, the lion’s share of US outperformance, not all of it.
Again, that 15% to 20% that’s unaccounted for, that is the US actually being exceptional over this period, actually growing earnings, whatever cash flow, whatever measure you want to look at. Better than the rest of the world. So the US’ victory is not 100% from revaluation, but it is 80% from revaluation. And now the US is considerably higher priced. So at the very least, I prefer a diversified portfolio around the world. Looking at it and saying I assume the US will win by the same margin in the next 25 years as it did in the last, is basically saying we’re going to see extreme multiple expansion on a relative basis—US against the world—from here again, which would take us to stratospheric differences that we’ve never ever seen. And I think any level of growth differential would find it very hard to justify. If one is more of a mean-reversion believer, maybe you want a little more global—that is a pretty low Sharpe ratio bet. Pure value to do country selection is not how I’d want to make most of my living. But it is something I would take a very small amount of risk on.
So sorry, long-winded answer to a short question. I do see it as a value bet. But I think there’s some interesting aspects. I think people maybe overextrapolate what we’ve seen the US do in the last 25 years, not realizing how much of it has come from people just willing to pay more and more for the same fundamentals in the US compared with other places.
Lefkovitz: Well, going back to that market outlook piece, the 2035 allocator looks back, it sounds like you are expecting international equities to outperform over the next 10 years.
Asness: Well, yeah. Again, that piece had no subtlety to it. I didn’t talk about how much risk I would take in such a position. But because of what we’ve just talked about, I am a little bit at least of a mean-reversion guy. So I probably have a little bit more over the next 10 years than I normally would, whatever your base case is. And again, I was getting to write it. So of course, I ended up being correct.
Benz: We wanted to switch over to discuss what you’re up to at your firm, AQR Capital Management. Before we get into some specific questions about various strategies that you run, can you talk about the firm generally and maybe talk about the ratio of AQR’s assets in retail mutual funds versus hedge funds that are accessible only by higher-net-worth investors?
Asness: Sure. Just looking at hedge funds, which are typically an LP structure, and mutual funds leave out a fair amount of our assets that are separately managed accounts. But if you’re just going to do that breakdown, we’re about 70% mutual funds, 30% LPs. If you want to look more broadly, and just talk about strategy types, mutual funds are about a third of the rest of AUM. So, they breakdown; they’re both super important. A third versus two-thirds, I wouldn’t call one—let’s round it and call them equal. One thing we do, which I’m proud of, is the mutual fund versions of our strategies use the exact models that the LP versions use. In terms of pretax alpha. So, to me, it doesn’t even matter where they come from. But if that made any sense, call it fairly balanced among the two.
Benz: OK, that’s helpful. And then in terms of the strategies that you run, AQR is primarily known for its alternative strategies, which we’ve discussed a little bit here, but you do also have long-only strategies. How would you envision the long-only products being used in a portfolio?
Asness: Well, first, we have long-only largely because we, there are many investors who simply can’t, can’t or won’t, allow shorting in their portfolio. We think our process is frankly better if you allow us to go both long and short. The alpha on the short side … Imagine you have alpha. Let’s start with that. We think it’s fairly symmetric. We think our ability to pick stocks that will underperform is roughly similar to outperforming. And if so, in a long-only beat-the-benchmark portfolio where you’re owning just different stocks than the benchmark, there is often a limited amount you can actually move the dial with negative use. You could only not own something. And unless something is a giant stock, a lot of stocks are small in the index and not owning them is fairly trivial. I would never short a ton of a smaller stock, but to move the dial, you might want to have a bigger bet than simply not owning. And we do think the risk-adjusted return, if you’re allowed to short, is higher.
But simply put, we think the long-only beat-the-benchmark more constrained versions of the process still have a positive expected outperformance. We wouldn’t do it otherwise. We charge a lot less for it than we do for long short. So the net returns are closer than the gross return difference. The gross returns, again, should be better if we’re allowed to short. So we do it because it’s a marketplace that exists. It’s a very big marketplace, and we think we can add value. So the way an investor just thinks of that is just traditional active management. Instead of equities, will these long-only equities on average outperform? So we don’t think we could do quite as much as we can do if an investor allows us to go short, but we still are quite happy with it.
Lefkovitz: AQR is launching a series of what you call fusion strategies. Maybe you can explain what those are, what the philosophy behind them is, and what you expect from them.
Asness: Sure. There are really two concepts here. And I’ll say in advance, both go to something I’ll explain called capital efficiency. One is, imagine you’re running an uncorrelated alternative. I mean, liquid, true uncorrelated, you’re short and long in equal amount in terms of dollars, beta. You’ve really thought about this, and nothing ever comes in perfectly uncorrelated. Of course, real life will always throw you curveballs, but imagine you think you’ve created an uncorrelated source of alpha. Step one is, in general, it is more useful to investors if you run this at a fairly aggressive level. In the long-short role, that tends to be, again, the “L word”: leverage. But there are two choices: I can run this at a 5% annual volatility, call that about a third of equity volatility, or a 10% annual volatility. Quite simply, investors have to give half the money at 10% vol to get the same bang as they would at 5%. That is capital efficiency. They can do something else with the rest of their money.
They can move the dial more. It is the exact same product. And in fact, this can cause sticker shock at times, but it should be exactly double the fee. Or if I want to be more positive about AQR, the lower vol should be exactly half the fee. You can’t cheat in either direction, but that’s a form of capital efficiency. The second is, all right, so you’ve created something uncorrelated. I’ve spent much of my career complaining, whining, writing about so-called alternatives that are really getting a lot of their return from market beta, from being long. We wrote a piece on hedge funds in 2001, showing that on average, their correlation to long-only equities was over .8. It had healthy positive betas. I have a feeling you’ll ask me about this, but we think much of the return to private investing is really just market beta as markets go up over time. So this is a little bit of an odd position for me, because I’ve spent a lot of a career saying, be careful about beta being snuck into your alternatives.
But I think I actually, I even lost sight of this to some extent. The problem was never beta. The problem was paying alpha fees for beta. The problem is if you’re charging a 20% performance fee, and that is on both your alpha and a beta-one exposure to the market. Well, you can get that beta-one exposure to the market a heck of a lot cheaper than hedge fund fees. So I think even I lost sight of that. So imagine you did this: You have your higher volatility alternative, which already we think is more capital efficient. You have to put fewer dollars in to move the dial. But imagine it’s still a lower expected return than equities. So equities you expect to make 10% a year, maybe on the alt you expect to make 8% a year uncorrelated. That’s wonderful, that improves your Sharpe ratio. If you allocate out of equities, it does not improve your total return. You’re still allocating from a higher expected return asset into a lower one. And the old saw, we’ve all heard it a million times—you can’t eat risk-adjusted returns—would be true in this case.
Now earlier we talked about levering a 60/40 portfolio. If you apply a little leverage to something you often can eat risk-adjusted returns. But if all you do is allocate from the 10% to the 8%, you’re going to have a lower return—won’t be quite 2% lower because you get a little benefit from diversification on your compound return, but it’s going to go down. Now imagine the alt simply adds a position, say using an S&P futures. $1 in the alt gets you the 10% vol alt that’s uncorrelated, plus $1 in the S&P future. You can’t charge alt prices for that $1 in the S&P future—that’s the sin. But just adding it can be quite helpful to an investor constructing a portfolio. Because now if they allocate $1 out of equities to you, you are replacing that dollar of equity exposure and adding the alt as a pure additional exposure. And if you’re good at creating uncorrelated exposure that has a trivial effect at real-life sizes—5% allocation to the alt barely changes the volatility or even more general measures of risk of the overall portfolio.
So again, it is just a capital efficient way to do it. It’s a way one criticism of some alts that has some bite is, hey, great risk-adjusted returns, but they don’t move the dial enough. Both of these things, more aggressive on the pure uncorrelated alpha side, and coming with a package deal—some financial engineering actually, just adding the equity exposure—as long as you don’t charge for it, can make that alt far more efficient and make it so if you’re right about the alt—you still have to be right—it does move the dial. And that is the fusion concept. I don’t know where we came up with the name fusion. That’s one of these that we probably had focus groups for six months. No, we don’t do that, but somebody came up with it. I kind of like the name.
Benz: Wanted to follow up on the idea of moving the dial because you have written about position sizes with alternative assets that, as you said, it’s often too small to really make a difference. And it sounds like these fusion strategies aim to address that. But can you talk about how advisors should approach that decision about how much to allocate to alts in portfolios?
Asness: Sure they should allocate 100% to AQR alts.
Benz: OK, but the real answer is …
Asness: No, my compliance area just heard that and just had a myocardial infarction. I was joking. There’s no perfect answer to this. You come up with a set of assumptions. You say, can I live with this? You try to do it in a real-world way. Some of the practicalities of if you lose unconventionally, it is harder to stick with it. If you lose conventionally, even if that has annoyed me throughout my career. I don’t deny the truth of it during some of our toughest periods. They’ve been very positive periods for markets, and we’ve been proven right at the end of them. But it was still a rough ride because it was more of an unconventional loss than a conventional. I’m going to disappoint you if you were looking for an actual number. But it’s a process of looking at where your assumptions are. And we just think the fusion concept helps you get there better. If your goal is to raise the top line, there is nothing wrong with adding an alt that slightly decreases the top line but has a decent impact on the risk. Risk-adjusted returns can be a great thing. Maybe you like that trade-off. But the fusion concept lets us offer something for that investor who is looking to move the top line, not just the risk-adjusted return.
But the specific numbers. My joke before—I don’t think AQR is the only firm that’s good at this by any means. So it’s not just what you think of us. It’s what you think of what you can find out there in general. And then how does that compare to your long-term assumptions for your traditional assets? So we help investors—we give them our bias view on what we think these are, but every investor has to make that call on their own.
Lefkovitz: I wanted to ask about AI and how AQR is using it. You’ve been quoted recently as saying that “AI is annoyingly better than me.” Maybe you could explain what you meant by that.
Asness: Yeah, one way to view what a quant does, and this is probably an oversimplification, is come up with rule-based things that say this type of stock—and by the way, I keep using stocks as an example. We do this in the macro world. It’s not just individual stocks. But I think that’s the cleanest example. Come up with a rule-based approach that says stocks with these characteristics tend to, on average, beat stocks with these characteristics. Some of the famous ones in the factor-investing world are again the value factor, the profitability factor, the low-volatility factor, the momentum factor. There are more these days and there are many different ways to measure these things. How to measure momentum, how to measure value—God doesn’t come down and tell us. We have a composite of many different ways we find reasonable. It has always been the job of the senior researchers—myself included, but not just me, of course—to say at the end of the day, how should we weight these different factors? We’ve never been short-term timers on it. So this is a long-term decision that we occasionally revisit.
It’s not like I come in in the morning and go, I feel really good about the low-vol factor today. But in setting these long-term weights, there was some judgment applied. We would pride ourselves on trying to be roughly 50/50. There’s no exact way to say what 50/50 even means, but 50/50 based on empirical results, both in sample kind of backtests and even better out-of-sample results since you’ve been trading the factor, but also call it 50% on, does the factor make sense? Making sense could be a deep theoretical model. It can just be common sense. And the reason you have the second part is to avoid overfitting to avoid data mining. You can find plenty of factors if you just let the data go crazy that have worked long term that you just look at and go, yeah, no way. You guys know the famous examples for market-timing by when the team from the NFC wins the Super Bowl sell when a team from the AFC wins the Super Bowl. I don’t care how well you show me that factor has worked. I’m not betting on it. What it turns out is if you start with a set of factors that you already believe in and then use modern techniques. Actually, they don’t even have to be that modern. The first way we did this, and still a big part of what we do, are simple what are called Bayesian techniques. It was just a mechanical way of updating prior beliefs based on new information, which is largely the return on the factors. Now, we’re in the midst of adding a lot of machine-learning methods that are very similar in spirit.
We have these historical results. We believe in these things. How should we wait then, let’s update with new data. I often joke that Bayesian techniques are machine-learning circa 1670 AD. I’m probably off on my years but the famous Reverend Bayes with his formula was pure math back then. That was machine learning if you think about it; just really simple. But by and large, we do think these more mechanical ways, if you start with good factors to begin with. If you let a mechanical method make up any factor at once, it’s going to gigantically overfit. It’s going to find crazy, silly things that make absolute no sense. But it turns out that if you find a set of factors you believe in, which we’ve spent pretty much 30 years creating. Then these methods seem to be both in backtests and so far in our real-life results somewhat better than us making a judgment call at the top. And I believe going forward will be somewhat less streaky. Obviously, we’ve been pretty happy with our now 27-year experience at AQR, but it has been streakier than I’d like with some very good and very bad periods.
Luckily, the good periods have outweighed the bad in magnitude and length, but we’d still prefer to be somewhat less streaky. And I do believe the mechanical methods will be somewhat less streaky, largely because no matter how good your intentions are, when you are people applying your own common sense to the allocation, you almost cannot help imposing a little bit of a philosophy, a little bit of a view. The machines are coldhearted. You’re seeding it with things you believe in so there’s still some judgment involved. But after that, the machine seemed to do a better job. So, the old joke about a recession is when your neighbor loses his job, a depression is when you lose yours. I worry about machine learning is when other people are losing their jobs. I lost a little bit of mine to the machines. Again, lots of room for human judgment and what goes in. I still think even for people who use ML—and we are extensive users these days—I think we are probably decently more toward still imposing some economic intuition. There are people in the ML world that are more pure-math people. But a little more data than intuition than we used to be because the techniques have gotten better.
Benz: And ML of course is machine learning there.
Benz: We wanted to switch over and discuss privates because you have been very vocal in addressing the uptake of private security.
Asness: What am I not very vocal about? Let’s narrow this down.
Benz: Well, that’s true. Before we delve into privates, I was hoping you could get into nomenclature a little bit because you do believe that private equity and credit are equity and credit, respectively, and not alt. So maybe you can talk about that.
Asness: Again, I find it hard to even engage in this argument because I find this so staggeringly obvious. Private equity, the canonical private equity—different firms won’t be exactly this, of course—are mildly leveraged long-only equities that are simply not publicly traded. The notion that these would be uncorrelated assets is fairly insane. I have a story for you. I keep dating myself with these stories. This is 1997. In 1997 we had a little thing called the Asian debt crisis. There’re like three people listening to your podcast who ever heard of this one. But …
Benz: I remember.
Asness: I wasn’t going to say anything. But I was still at Goldman Sachs. We were thrilled that our fund that attempted to be market-neutral was our guess was we’re flat that day. I say our guess—our P&L system said we were way up. But that’s because we were short the US against Europe. We’re not always short the US against Europe, by the way—it sounds like we are, but we were back then. And one great way to look like a genius is to be short a crashing market. The S&P was down about 7% the day of this crisis and long a closed market. Short crashing, long closed—your P&L system says you’re minting money. We said no, the Europe will probably fold a little bit more than the US tomorrow, and we’ll be flat. It did turn out to be true. So the head of Goldman Sachs—I will keep the name to myself to protect the guilty (no, he was actually a really smart person but frustrating conversation)—comes over and says, “How you guys doing?”
And I don’t let him look at the screen because I don’t want him to see it’s up a lot because it’s not true. I’m like, “I think when it all settles, we’re going to be flat.” And he looks at me and goes, “That’s great. Us too.” And I’m like, “Wait, you are leveraged long equities, right?” He goes, “Yeah.” I go, “If you had to sell them today—I don’t mean a fire sale, obviously, that’s silly for privates—but if you had to sell them today versus yesterday, they’d be down at least what the market would be down, no?” And to his credit he said, “At least.” And I said, “So what?” He goes, “But we don’t have to sell them.” To which any public manager can say, I don’t have to either. There’s really no difference if you’re in the private world. These people are geniuses at valuing a company. If the market crashed 20% and you said value my portfolio today, they could give you a damn good estimate of what it’s worth today. I don’t get why one set of investors—the private versus public—gets to mark their portfolio to what they think it’s worth. While the other set has to mark it to where the market will pay today.
I will tell you in some of our tougher periods, I was more than capable of marking our portfolio to what I think it was worth. And I would have turned out to be right had I done so. When this valuation bubble corrects, we’re going to be way up—would have turned out to be right. I don’t get to do that. So at its simplest, they’re long equities. They know, and we all know, that they move in value every day. I don’t even see why it’s any different. That’s why I say it may or may not be a good investment. I’m not saying it’s a bet on the skill of the private manager. I’m not so cynical. They can do some things we can never do like make a company better. A quant who’s long a thousand stocks and short a thousand stocks ain’t making their companies better. They’re just they’re just making bets. So there’s a source of potential alpha that I can only dream about. They charge a heck of a lot for it. I’ll let other people argue about where the net comes to. When I use the word alternative, I mean a true diversifier, not an accounting diversifier. And I think when it comes to that they’re an accounting diversifier, not a true diversifier.
Lefkovitz: Why do you think there’s so much zeal toward democratizing access and getting credit and private equity into smaller investors’ hands or portfolios?
Asness: You’re just trying to get me to be mean at the end of the podcast, aren’t you? It’s actually a high Sharpe ratio strategy for you. I’m more than willing to oblige. You know, to be honest, they’re businesspeople. They’re always seeking to expand their business. I think their traditional institutional market is full up to the gills on this stuff. So where else are you going to go? I’ll be frank with you: I think it’s generally a bad idea. I think a lot of the structures proposed are some weird hybrids where they’re somewhat liquid, but privates in there. I don’t even know how it’s going to work. I think they actually risk ruining the magic. A lot of the magic is essentially investors let them get away with saying these things have very low volatility. They let them go with the reported numbers because it makes everyone’s life easier.
And once you move toward liquid pricing, toward mark to market, even a small way, they may risk popping their own balloon there and having people start to focus on the actual economic movements. But I think on a podcast about economics and investing, I can tell you the reason they’re doing it is the same reason, to be honest, that we do new products, is to make money. I know that’s a bold statement. I will go out on a limb and say I don’t think it’s a great idea for traditional investors. I don’t think they’re getting this diversification they think they’re getting after fees. I don’t think they’re getting a particularly large excess return. So I’m a seller of this movement, but I don’t get a vote.
Benz: For our last question, Cliff, we wanted to ask, who do you rely on to give you knew ideas to inspire different ways of thinking about the economy and the market? Who are your must-reads or must-listen-tos?
Asness: I mostly read and listen to the podcast of other quants. Like anyone in this business, I read a lot of macro stuff, too. But that’s just because I should know what’s going on in the world, not because, like I said before, I’m a macro economist who’s going to add a lot of value. So I read competitors’ stuff. We have some amazingly good competitors. I’m not afraid to compliment them. I think some nice people are too scared to do that. I feel bad because this will be nonexhaustive. Can’t think of everyone on the fly. But a team at Robeco has for a really long time put out really good quant research. We don’t always agree. We differ on a few things, but I always learn something. Wes Gray, Corey Hoffstein, Owen Lamont. These might not be super well-known names to the world, but I think they should be. And you don’t have to be a quant to read any of this stuff. A lot of them make comments that are just generally great and useful. And even if I’ve been doing this for more than 30 years, I learn something from them and others who I’m probably forgetting.
Benz: Well, Cliff, thank you so much for taking time out of your schedule to be with us. We really appreciate it.
Asness: Oh, it was a lot of fun. Thank you, guys. Thanks for having me.
Lefkovitz: Thank you, Cliff.
Benz: Thank you for joining us on The Long View. If you could, please take a moment to subscribe to and rate the podcast on Apple, Spotify, or wherever you get your podcasts. You can follow me on social media @Christine_Benz on X or at Christine Benz on LinkedIn.
And at Dan Lefkovitz on LinkedIn.
George Castady is our engineer for the podcast and Kari Greczek produces the show notes each week. Finally, we’d love to get your feedback. If you have a comment or a guest idea, please email us at thelongview@morningstar.com. Until next time, thanks for joining us.
(Disclaimer: This recording is for informational purposes only and should not be considered investment advice. Opinions expressed are as of the date of recording and are subject to change without notice. The views and opinions of guests on this program are not necessarily those of Morningstar, Inc. and its affiliates, which together we refer to as Morningstar. Morningstar is not affiliated with guests or their business affiliates, unless otherwise stated. Morningstar does not guarantee the accuracy, or the completeness of the data presented herein. This recording is for informational purposes only and the information, data, analysis, or opinions it includes, or their use should not be considered investment or tax advice and therefore is not an offer to buy or sell a security. Morningstar shall not be responsible for any trading decisions, damages, or other losses resulting from or related to the information, data, analysis, or opinions or their use. Past performance is not a guarantee of future results. All investments are subject to investment risk, including possible loss of principal. Individuals should seriously consider if an investment is suitable for them by referencing their own financial position, investment objectives and risk profile before making any investment decision. Please consult a tax and/or financial professional for advice specific to your individual circumstances.)