Full Transcript: Ben Inker on the AI Bubble and Private Equity

Earnings Bubbles, Capital Cycles, and the Case for Going Abroad

Justin: Ben, welcome back to Excess Returns. Thank you for joining us.

Ben: Very happy to be here.

Justin: We were just talking, it was 2021 the last time you were on the podcast. It’s hard to believe it was that many years ago. But we always appreciate having you or the other folks at GMO on with us and our audience. You spend your time managing portfolios and building investment strategies, but you and your team also think deeply, write extensively about markets, valuation, asset allocation, stock market history, and a number of other investing topics.

And today, what we thought we would do with you, and we’re excited to do this, is kind of look at some of the recent research that you’ve put out, the firm has put out, and sort of just work through those items with you. I think we’re gonna start with the AI sort of boom and possibly bubble that we’re in and kind of talk about how that, I guess, correlates to past market bubbles, and then we’ll get into some of the research that you’ve done on private equity and some long-term return forecasting expectations. So we’re looking forward to a good, thoughtful discussion with you today. So where I want to start with you is related to a recent piece of research that you put out, and it was sort of looking at where we are today in the market with AI and looking at it from an investment bubble perspective, where valuations are, the market, the speculation that’s sort of underneath.

And one of the things that you said is that it’s one of the easy, easy bubbles, easy ones, as you put it, for an agnostic investor to handle. So can you just explain sort of, I guess, what would be the difference between an easy bubble and a hard one to understand and kind of what you meant by that?

Ben: Yeah. So for us an investment bubble is a situation where at least some important asset out there has risen to a level where it feels pretty close to unownable. That’s not the only definition of a bubble, but from the standpoint of putting together a portfolio, that’s a useful one. And we’ve had a number of those in the last twenty-five, twenty-six years.

And what makes some of them different from others is if you believe the situation that we believed at the time, so we believed we were in a bubble in each of those times, how hard is it to put together a portfolio that can avoid the worst of the pain, and yet allow you to retain your clients?

‘Cause the problem with some bubbles is that avoiding them means running a portfolio where if the world was normal, that portfolio would make no sense. An easy bubble is one where you can take a normal amount of risk and still avoid most of the pain of the bubble. So my example of an easy bubble in the last quarter century was the internet bubble.

In the internet bubble, growth stocks around the world were horrendously overvalued. The S&P 500 was the most expensive it had ever been in history, and even today, it has still never passed those valuation points, at least on most normal valuation metrics. But it wasn’t that hard a bubble to navigate because you could actually own a normal amount of stocks.

If kind of normal is a sixty-forty portfolio, you could run a portfolio that was sixty percent in stocks and yet not have to own any of the really overvalued stocks. You just had to say, “Okay, instead of owning large cap growth, I’m gonna own small caps. I’m gonna own REITs. I’m gonna own emerging equity. I’m gonna own emerging debt.” You’re still owning risk assets, and the crucial thing about the fact that you’re still owning risk assets is let’s imagine for a second that you were wrong and the world was completely normal, right? You still own a portfolio that has a normal amount of risk assets. You still own a portfolio where if everything is normal, eh, you’ll get a normal return.

There’ll still be plenty of tracking error versus the cap-weighted portfolio, but you are not necessarily giving up long-term returns, for the sake of saving you from the pain. So that was a bubble that was relatively easy to navigate. The next bubble, in the global financial crisis, was a lot harder because it didn’t just affect large cap growth stocks, it affected all risk assets everywhere around the world.

As near as we could tell at the time, every single risk asset was substantially overpriced, and that meant in order to save yourself from the pain, you really needed to not own risk assets. That is an incredibly hard thing to do because if you’re wrong, you are giving up a tremendous amount of return for your clients.

And if you are not right quickly, your clients are gonna really lose patience with you. The next bubble, which from our perspective occurred at the end of twenty twenty-one, was an even harder one to deal with because stocks and bonds were simultaneously really overvalued.

That was a time where everything with duration was really overvalued. So avoiding losing money didn’t just mean moving to high-quality bonds from stocks, but it meant moving to cash or something cash-like. And the problem with that is everybody knew cash was gonna lose to inflation. And the other thing is, if you’ve got clients, you guys have clients, we have clients, the one thing they don’t want you to own for them is cash, right?

You do not need a professional money manager to own cash. You can do that in the bank. So moving your clients to cash, you have the shortest potential leash there in terms of time. So that’s a really difficult bubble to deal with. The AI bubble, we think that this is really a bubble in US stocks writ large.

So large caps, yes, but small caps in the US still look pretty overvalued. The good news is if you move outside of the US, you can still own plenty of risk assets, which are priced to deliver a decent return. So this is one of those times where you can put together a portfolio of stocks and bonds that avoids most of the peril from the bubble we think exists today, and yet isn’t an insane portfolio to own if you’re wrong and everything is actually normal.

Justin: And we’ll put these risk and return charts in here so people can see them. ‘Cause you have all these different asset classes, both in 2000, 2007, I think, and then you have it as of late 2025. What does the... Just talk, I mean, maybe you could just explain what we’re looking at. It’s kind of obvious based on what’s the chart, but I think it would be good to, Ben, to just walk through that real quick.

And then talk about the difference between, I guess, what is the— does the slope tell us anything? How the slope is positive on two of them, and then negative in ‘07. Can you just talk to that? Maybe that’s to your, the valuation point, but I’ll let you explain.

Ben: Yeah. So the charts you were talking about are charts we have built of the forecast we were showing to clients at the time. So this is not some sort of data mine back cast, but the forecast we were showing our clients at each of those times. We’re showing them as a risk-reward scatter plot.

So the horizontal axis, as you go out farther from the origin, you’re getting to riskier and riskier assets. Vertical, the higher you are on the page, the higher the expected return. And the importance of that risk-reward line, of that regression line, is that’s basically telling you how much are you getting paid for taking risk.

The slope of that line should be positive, right? In a rational, properly functioning capital market, you should get paid for taking more risk. Now, risk isn’t exactly the same thing as volatility, but it’s close enough for this purpose that you should really see a positive slope to that line. Under normal circumstances, if everything was priced fairly, we think that that line would have a slope of about zero point seven.

So if you look at that line in 2000 and say that line has a slope of zero point four, that’s still positive. It still means you are getting paid for taking risk, somewhat less than normal, but not an utter disaster scenario. And so when we look at a risk-reward line like that, we say, “Okay, well, you probably don’t wanna own more risk assets than normal at a time like that. Maybe you wanna shade a little bit less, but this is not a time where de-risking your portfolio is a crucial thing, ‘cause you are getting paid for taking risk.” If we looked at the world in 2007, that had been completely turned on its head. You were paying for the privilege of taking risk. This was the first truly global bubble.

As Jeremy Grantham was saying at the time, every single risk asset we could find looked massively overvalued versus history, and not that mispriced relative to each other. So one other crucial thing that was going on in 2000 was the average risk asset looked much better than the cap-weighted stock market. So if you were prepared to own a risk asset portfolio that looked different from the market, you could make a lot more money that way.

In two thousand and seven, you couldn’t. That equal weighted portfolio of risk assets looked almost exactly the same as the cap-weighted portfolio, so you couldn’t diversify your way out of the pain. You needed to move up that line. You needed to move towards the origin. Now, in one sense, like an optimizer would say, “Well, that’s easy. All of the low-risk assets have higher expected returns and lower volatility? Man, this is a piece of cake. I can just put together a portfolio of just bonds and cash, and I’m fine.” But as you guys well know, justifying that to your clients is really hard. Justifying it to your optimizer is easy. But justifying it to your clients is really hard.

In twenty twenty-one, we didn’t have a situation quite like that. You were not paying for the privilege of taking risk. The slope of the risk-reward line was positive, but everything was really below zero. So across the length and breadth of assets, just about everything had a negative expected return in real terms.

And that to us seemed, well, that’s not a sustainable situation. There really needs to be a positive real return to investing over time. And so what you want is to own assets where a repricing is going to be as painless as possible, and that means you need to own stuff with really short durations. Right? A repricing of stocks is devastating if it’s a repricing downward, because stocks have a really long duration to them. Bonds have a long duration. Real estate has a long duration. Infrastructure had a long duration. Everything that has private in front of it, well, maybe not private credit, but other things have a long duration.

So what you needed to do was avoid duration. Again, an incredibly difficult thing to do. As of this fall, we thought US equities looked really overvalued. The rest of the world looked just fine. It was easy to put together a portfolio that took a normal amount of risk and looked pretty good. The bad news since then is that a portfolio like that has really done quite well, right?

The non-US risk assets have done well, and their pricing is somewhat less attractive than it was then. The slope of that risk-reward line was zero point four then. It’s zero point one now. Now, it’s still about zero point four if you excluded US equities, and for our part, we kinda want to, at least in the portfolios we’re allowed to.

So in that world, yeah, you are getting paid for taking risk, but it is getting to be a tougher environment. And again, it got there the good way, by having emerging do well, having non-US equities do well, having emerging debt do well, having almost all risk assets do pretty well.

Jack: It’s interesting. You think about what’s obvious to the optimizer and what’s obvious to the person. It’s the opposite of each other effectively. And I guess client behavior and expectations is sort of the gap between those two things. Like, even if it’s obvious for an optimizer to own the risk-free asset or something, you can’t actually do that in the real world.

Ben: Yeah, it’s very difficult to do that in the real world. And one of the things the optimizer doesn’t know, that your client cares a lot about is there is some risk you are wrong.

Jack: Or there’s a risk it’ll keep going too, even if you’re right in the long run, there’s a risk it’s gonna go on for a number of years.

Ben: Yeah. I mean, and that’s pain... If you really believe you’re going to be right in the end, you can put up with that pain. But it is so difficult for the client to have that faith. Look, I can tell you as someone who has lived through these bubbles and believed this stuff going in, as the market is moving against you, it is hard to maintain that confidence yourself.

And if this is merely a professional that you hired because, I don’t know, you liked the story they told, and you liked some of their historical track record, your level of faith is gonna be a lot lower than my level of faith in myself is. So the real problem with these situations is if you need to run a portfolio that is gonna look very stupid if the world fails to fall apart quickly, your client is going to fire you.

That is bad for you. It is also bad for your client, because your client is overwhelmingly likely to fire you and hire the person who was doing the exact opposite of you. And so we saw this in the internet bubble. Clients would fire us and hire the aggressive growth managers. And so they made less money on the way up, and they lost more on the way down.

Jack: Have you learned anything about how to manage that? I mean, Jeremy told some great stories when he was on recently about this, the late ‘90s and clients firing you guys and things like that. Like, have you learned anything about how to maybe make clients stick with a strategy better? I mean, part of this is human nature that you probably can’t adjust, but I mean, have you learned anything this time relative to what you went through that time?

Ben: So we’ve learned a few things, I think, I hope. One of them is about managing the portfolio. And so in the 2000 event, we were running more de-risked portfolios than we needed to. And it worked out just fine ‘cause bonds did well and the path for stocks was a little bit bumpy, even those stocks that went up.

But in retrospect, we didn’t need to be as underweight stocks as we were. And that is certainly something we are keeping in mind today. We are trying to make sure we don’t own any assets that we don’t think are giving a decent risk-reward ratio. But we are trying to hold onto those risk assets that we think we can afford to hold.

Because what have we learned living through four bubbles in 26 years is, man, timing this stuff is tough. And they can go on longer than you ever thought possible. And having a portfolio that only makes sense if the bubble bursts quickly, it’s a tough way to make a living. So we’ve tried to do some stuff in terms of making Bayesian adjustments to our forecasts and trying to build portfolios that are robust to that.

Another thing we’re doing is being very open about what we’re doing, why we’re doing it, and what the likely consequences are to the portfolios. But man, if there was a way to do this without the career risk, without the client firing you risk, this would be a much easier business.

Jack: How would you contrast the 2000 period to today in terms of one of the questions a lot of people have? I mean, many people will agree the AI-type stocks are in a bubble-like environment. But some will argue now the more large cap stocks that represent most of the index are at least maybe expensive, but at least at non-bubble-like valuations, and they’ll contrast that with 2000 where they’ll say those types of stocks were at bubble-like valuations.

I mean, do you think that’s fair?

Ben: Yeah, it is certainly the case. If you look at the giant stocks, most of them are not trading at crazy valuations. So let’s hold SpaceX aside, let’s hold Tesla aside, let’s hold Palantir aside. But if you look at Microsoft, hey, Microsoft is trading at much lower valuations than it was six or nine months ago.

And the valuations don’t seem crazy relative to earnings. A tricky thing about today is we worry that this may be an earnings bubble. 2000 was a little bit of that, but mostly a valuation bubble. We saw earnings bubbles in Europe in 2007, 2008, where the earnings had just gone up 100% over four years, and even today they are struggling to make it back up on an index basis to those levels.

We saw the same thing in EM in kinda 2012. Earnings bubbles can exist. One of the things that can drive them is rapid increases in investment. ‘Cause the thing about investment, in the long run, what matters with investment is what is the return on investment of that, right? What’s the ROIC? In the near term, if you think about investment as it’s done after the income statement is over, right?

I am Microsoft. I am spending $200 billion on data centers. Well, none of that comes out of my income, but all of that spending winds up somebody else’s revenue. Eventually, I’m gonna have to depreciate it, but I depreciate it over time. So as investment is going up, that kind of mechanically makes earnings higher than they would otherwise be.

And actually right now, we’re in one of these situations where if it takes three years to build the data center, until that data center is complete, there is no depreciation. And right now, given how rapidly the investment has ramped up, a ton of that investment is not just only depreciating one-sixth or one-seventh of the spending, but hasn’t started depreciating yet at all.

So this is a very good time for corporate profits. We think probably an unsustainably good time for corporate profits. But relative to 2000, a smaller portion of the market is trading at insane valuations.

Jack: Yeah, that earnings bubble thing is so important because people don’t think... People always think about price bubbles, but they never ever think about earnings bubbles. And if you do get into a situation where earnings are just unsustainably high, then you can seem to have reasonable valuations but still be in a bubble, which I think is tough for people to understand.

Ben: Yeah. I mean, I think people are... I mean, not that anybody cares about this anymore, but let’s say you’re looking at resource companies, right? Everybody knows you don’t buy resource companies when their P/Es are really low. Their P/Es are really low because you are in at kind of a peak earning situation, and the capital cycle is about to come around and bite you on the ass.

You buy resource companies when the P/E is high ‘cause the earnings have collapsed. We are in the midst of an amazingly large capital cycle associated with AI. And I think we’re going to be in a situation where before the end, right, the SK Hynix’s and Microns of this world will look really cheap on a trailing P/E basis, but turn out to be lousy investments.

Jack: Yeah, I was thinking about the semis is exactly what I was thinking about when you said that because semis have traditionally been cyclical businesses. But the argument now, which is probably a bubble-like argument, is they’re not anymore. You know, we have this new thing, AI, driving everything, and this is no longer a cyclical business.

Ben: Yeah, and of course it has to be a cyclical business. I mean, one of the things about the memory makers is it’s not exactly a commodity, but ooh, it has had all of the characteristics of a commodity business. And so the commodity cycle and the capital cycle tends to be really painful. You can get situations where that cycle turns into a super cycle, right?

We had this iron ore super cycle when from 2005 to 2012, companies were investing to be able to deliver more iron ore, but it still kept working out because China’s growth in demand was just higher than anybody ever expected. Eventually those companies got smashed, but it took a lot longer than anybody expected.

What we don’t know right now is how long it is gonna take for supply to keep up, to catch up to demand. If it doesn’t, then this will be like this weird failure of capitalism that we’ve never seen before. Could happen, but that would truly be a this time is different situation.

Jack: I want to ask you about AI CapEx in general, because I know you guys have studied history a lot, and all of us are trying to figure out what to make of this whole thing with this amount of spending. I’m just wondering, in a historical context, like how do you think about this? I mean, if you look at capital cycles in the past, this doesn’t necessarily bode well for the people that are doing the building of this in terms of how it’s gonna work out. But then people argue this is intelligence, this is different, this is something that’s gonna replace human intelligence, so we can’t really analyze it in that way. Like how do you think about that?

Ben: Yeah. So for one thing, in terms of scale, right? This is a big CapEx cycle, but it is certainly not an unprecedented one. I think if you look at the data center spending this year, it’s kind of forecast to be, I don’t know, $700 billion, I think is the most recent number I saw, and that’s like two point two percent of US GDP.

So it’s a big number. A little bit bigger than the fiber optic spend in the late 1990s. About half this size relative to GDP as the railroad spend was in kind of the middle and second half of the 1800s. And probably pretty similar to the electricity build-out kind of near the turn of the century.

So it’s big. And there aren’t a lot of things that have been that big. But there have been things that, relative to the economy, have been of reasonably equivalent size. Historically, they haven’t worked out well. And that’s for a couple of reasons. One is, if you’ve got a transformational technology, it is actually not hard for people to understand, man, this is gonna change the world.

And whether it was canals or railroads or electrification or the internet or automobiles, people were right. It really did change the world. The tricky thing is, just because something changes the world doesn’t necessarily mean the profits accrue to the people who built it, right? Railroads changed the world.

Railroads didn’t change the world because being a railroad was awesome. They changed the world because they completely collapsed the cost of long-distance transportation, which meant you could have much more specialization in manufacturing, you could have much more specialization in agriculture, and you could move people across the country, right?

California couldn’t really have existed before the railroads. But running a railroad has never been an amazing ROI activity, and when you get a ton of investment, that tends to depress the ROI in that area to begin with, right? The UK did this before the US. In the UK the first company that built a railway line between London and Manchester, if it had stopped there, they would’ve made a lot of money.

Unfortunately, they were followed by five other companies, and once you have six railway lines going from London to Manchester with huge overcapacity, it’s not just that the marginal dollar invested is gonna have a lousy return, you will destroy the return of all of that CapEx. So what has almost always happened, when you have a situation where you’ve got this transformational technology and really good early ROI, is you get way too much investment and you destroy that ROI.

It doesn’t mean that the technology doesn’t change the world. Railroads did, electricity did, autos did, the internet did. That fiber investment was a lousy ROI, but it did enable the world we live in today. The tricky problem is unless you are stopped from doing it, the natural response by capitalists is to throw enough money at the opportunity that you are going to destroy the return on investment.

Jack: Yeah, when you were saying that, I was thinking about how this has been funded. And one of the differences here between fiber and what’s going on now is, at least at the beginning, this was funded through cash flow as opposed to debt. But now you’re seeing that debt get added on. And so I’m wondering if that plays into your idea of like they’re just gonna keep throwing money at it.

Even if it might have been initially funded by cash flow, they’re still gonna pile the debt on, and they’re still gonna keep throwing money at it to the point that it doesn’t make sense.

Ben: Yeah. I think that is going to happen, unless it’s stopped by something, right? If we are fundamentally incapable of providing enough electricity to these things, and so as much as people would wanna build an infinite number of data centers, they’re stopped, well, maybe they can kind of be saved from themselves.

But otherwise, capital will flow. And one of the things my former colleague, Ed Chancellor, who’s kind of one of the premier economic historians of bubbles, points out exists in bubbles is the financing gets to be really sketchy. You get Ponzi finance, you get circular finance, and my God, this cycle we have seen some incredible deals from the standpoint of strange ways to finance purchases.

What they have in common is they are structured in a way to look as profitable as possible initially. So OpenAI makes a deal to buy tens of billions of dollars worth of AMD GPUs. Now, if you look at the way that deal is structured, OpenAI is being given warrants on AMD stock, which is worth half of that.

So effectively, they are buying these GPUs half from shares of AMD that they are being granted. AMD could have given them a fifty percent discount instead, right? That would have been a more straightforward way to do the deal. But this one’s gonna make AMD look a lot more profitable, and still deal with the fact that OpenAI simply did not have the money to pay full price for all of these chips, even if they had wanted to.

The recent deal with Anthropic, offering to lease $36 billion worth of Alphabet TPUs, where Broadcom is stepping in and promising to buy back the TPUs in the event that Anthropic defaults in order to give this an investment-grade rating. Well, it’s structured in a way that this does not show up entirely as a liability on Broadcom’s balance sheet ‘cause it’s viewed as a low probability, but it does benefit from Broadcom’s credit rating.

And you can be sure that in this deal, Broadcom is being paid for the fact that they are making the deal possible.

Jack: It’s interesting, for the average investor it is very hard to have any idea what’s going on with any of this stuff. Like, the circular nature of the circular deal with this company and this company, and then this company’s over to this company. It’s like, unless you’re in the weeds of this stuff, this is very, very hard to figure out what’s going on.

Ben: Yeah, and some of it really relies on the fact that the accounting isn’t gonna keep up. And we saw circular financing in the internet bubble, but a lot of that was, I don’t know, tamer. Right? It was two dot-com firms putting ads on each other’s sites and declaring that as revenue. This actually involves lots, lots more money, and creates more risk of problems when things go wrong.

One of the nice things about the internet bubble is the internet part was funded basically entirely through equity, so there were no systemic problems when it went wrong, right? The housing bubble was funded through debt, and there’s lots of systemic problems when debt goes bad. This time around, it’s a combination.

Plenty of it used to be funded with equity, but increasing amounts are being funded with debt. I think the hyperscalers have doubled their debt ratios in the last nine months. I mean, they are taking out huge amounts of debt. They’re still very high-quality companies, and I’m not, I’m not saying they’re going to default on this debt, but they’re taking on debt really rapidly.

Jack: Just one more quick question on this before we go to your return forecasts. The idea of all this issuance coming out now, you’ve got these big IPOs, you’ve got companies that were buying back shares are now issuing shares. Like, how do you think about that? How do you put that in context in terms of like how you think about that and what it means for the market?

Ben: Yeah, I think that’s a really big deal. And one that I hadn’t understood how important it was, until I started to a couple of years ago. There was this good paper, I can’t remember who wrote it anymore, but the inelastic markets hypothesis that showed that the way traditional finance treats the financial markets is that they really should be very elastic.

That if a little bit of demand comes in or a little bit of supply comes in, that should have a very small impact on prices. And what this paper purported to show was, no, actually supply and demand really matter a lot. And so one of my colleagues, John Pease, presented that paper internally and then tried to replicate their findings and found, yeah, it’s really true.

The supply and demand matters a lot. There are natural experiments you can look at where the supply or demand are there, but we know there is no information flow associated with them, and it still has an impact on prices. And right now, we are likely to be in a situation where over the next twelve months, we see more supply come into the US stock market than has been the case in living memory. If we just get SpaceX and OpenAI and Anthropic and maybe a few other of the companies that in dollar terms don’t matter that much, we could be talking about five or six percent of aggregate US market cap coming in as supply.

Now, the thing is, it doesn’t come in instantly, right? SpaceX went public a little while ago now. But they only sold $75 billion worth of shares, and they sold that at a $1.8 trillion market cap. So there wasn’t very much additional supply that came online. As more of the holders become freed up from their lockups, a lot more will be able to change hands and a lot more effective supply comes in.

So if you look at the impact on the stock market, it’s not when an IPO occurs, it’s kind of in the twelve months afterwards, because it takes a while for those lockups to expire and the stock to change hands. But kind of the rule of thumb, if you looked at history, is that a one percent increase in supply is associated with a seven and a half percent worse return over the subsequent year.

And if that held up in a linear fashion, and we’re talking about five or six percent, well, that would be a pretty ugly return. I don’t know that it’s going to be that extreme. At that level, it is hard to disentangle from the fact that the two highest points of supply that we’ve seen in the US market in the last fifty years were two thousand and two thousand and twenty-one.

And both of those were also times when the markets were kinda crazy frothy. So pulling apart how much of this was driven by the supply and how much of this was driven by bubbles burst, I’m not sure. But right now we’re getting behavior that looks very bubbly. We are seeing a lot of supply, or we are going to see a lot of supply.

And that may be a real challenge to the market.

Jack: It’s interesting, like that inelastic markets hypothesis paper, it’s come up on the podcast before, but it came up with Mike Green, talking about the flows from 401s, like into the market and the impact that some of that can have that maybe is a lot more than people expect.

Ben: Yeah. I mean it is something you really have to, as an investor, continually keep in mind. Like, where is the supply coming from? Where is the demand coming from? How price sensitive is that demand? One of the things when the money is flowing in from 401ks, and let’s say they’re buying the S&P 500.

Well, the S&P 500 is trying to buy the same percentage of Nvidia as it does GM. But if the holders of Nvidia are less price sensitive than the holders of GM, it’s still gonna have a bigger impact on Nvidia. And when you think about growth stocks, the holders of growth stocks are, for perfectly reasonable reasons, less price sensitive than the holders of value stocks.

So even if that supply is coming into the market in a completely passive fashion with no views associated with it, it will have different impacts on different kinds of stocks.

Jack: I wanna shift to your asset class return forecast here. And before we... I’ll put the chart up right now, but before we get into the actual forecast, I wanted to maybe have you talk about how you do this. So these returns that we’re seeing in front of us right now, what is the process that goes into calculating these?

Ben: Yeah, let me talk about the basic idea first. At heart, all we’re really saying is that things will eventually trade at fair value. We use a seven-year forecast because historically seven years has been a decent estimate of how long that takes on average, even though things can be much faster or shorter than that.

So the way we calculate the forecast, the way we build it, is by looking at the income it’s reasonable to expect from an asset, the growth that you can expect from that asset, and then adding an additional term for the gain or loss associated with the valuation shift that is required for that asset to come back to fair value.

And we kind of assume it’s gonna come back one-seventh of the way each year. Now, there’s plenty that can go wrong with that, as in any kind of forecasting. But one of the things I like about it is it sort of is assuming that capitalism is going to work in the long run. You know, it would be weird if we lived in a world where the return on capital and the cost of capital did not come into alignment with each other.

That’s what capitalism is supposed to do. And if capitalism is going to do that, you’re gonna get this kind of gradual mean reversion in valuations.

Jack: You chose to break out low interest rate environment from a normal interest rate environment in this. Why did you do that?

Ben: So I said, we think things revert to fair value. What I glossed over was, okay, well, what is fair value? And the reality is it is much easier to come up with a fair value for assets relative to each other than in some absolute sense. So, and again, even just take a step further back, fair value to us is a valuation level from which you can get a fair return to that asset without having to have a valuation shift. So what’s a fair return?

Well, what’s a fair return to stocks? I don’t know. Tell me what the alternatives to stocks give. The way we tend to think about it is let’s start from cash. Let’s start from kind of the ultimate risk-free asset and say, “Well, how much more should I get paid for owning this asset because of its greater risk?”

So bonds are somewhat more risky than cash. They should probably give a return associated with that. We assume you’re gonna get somewhere around 100 basis points of term premium for owning a bond instead of cash. Now, stocks are a lot riskier than that, and in particular, they tend to lose you money at really unpleasant times.

They tend to lose you money when the economy is going badly and your income from other sources is likely to be lousy. So you should get paid a good deal more for owning stocks than bonds. We cuff that as around 4.5% for stocks relative to cash, so maybe 3.5% relative to high quality bonds. The thing none of that can tell you is what should the return on holding cash be?

And as near as we can tell, there is not an easy conceptual answer to the platonically correct return on cash. So what we do is we come up with a couple of different scenarios in which all of the asset classes are priced reasonably one to another, but where we can say, “You know what? Either of these scenarios could make sense as a long-term equilibrium for the interest rate environment.”

Today, our best guess is that we are in that low interest rate environment, so the return to cash in the long run will be somewhere around zero in real terms, maybe a little bit positive, but around zero in real terms, so after inflation. Stocks need to deliver four and a half points better than that. In order to do that, they need to be trading at a normalized P/E of somewhere around 21 times.

If we are in a world where cash is gonna return more than that, and it used to return more than that, then stocks need to return more than four and a half. And if we push everything up by one to 1.5% in terms of required return, well, then equities can’t trade at 21 times normalized earnings.

They need to trade at 16 times normalized earnings. So that’s why the interest rate environment matters. It changes the fair value for everything. And man, life would be convenient if I knew which of those scenarios we were in, but I don’t. So we publish both of them so that the clients understand some of what can drive different returns from assets in the long run.

Justin: What is interesting about the chart is clearly it shows that international over US and I think some of the highest conviction sort of ideas here or where the possible best returns are sort of coming from this value and deep value cohort of the equity market.

Ben: Yeah. So today we see a big gap between the US and the rest of the world. Now, it hasn’t always been the case that the rest of the world has looked more attractive than the US. It got there the hard way by underperforming for a long period of time. But US stocks have outperformed over the last, whatever, ten to fifteen years.

Some of that has been truly deserved because of better fundamental growth, but it’s also been associated with rising relative valuations. This idea that the US should trade at a valuation premium to the rest of the world, it didn’t exist fifteen years ago. There was no history of the US trading at a premium to the rest of the world as of two thousand and ten.

Today, everybody assumes it is the way the world should be. The other thing that has happened that has been important is a rise in the value of the dollar. As of two thousand and twelve, the dollar was really undervalued. Today, it looks substantially overvalued. So when we’re building our portfolios, we want to take into account the fact that non-US equities trade significantly cheaper than the US.

That’s going to be a help. Owning non-US dollar currencies for US-based investors is going to be a tailwind, because of the overvaluation of the dollar. And then within that, from a style perspective, today, small looks cheaper than large pretty much everywhere, and value looks somewhere between quite cheap and extraordinarily cheap, depending on the market you’re looking at, around the world.

Justin: A lot of times with these expected return projections, it’s hard because you see something like the large cap growth and how low or negative those returns are. But it’s not like the annualized return is likely to be in that range year over year.

A lot of times you can be in this environment and then you get like a bear market or big drawdown and then sort of that’s how those returns actually come to fruition. Do you see what I’m saying?

Ben: Yeah. Certainly we are not expecting that the market is going to go in a straight line from here to fair value over seven years. The way these things have tended to play out is, yes, you will have a really bad year, right? 2022, the stock market went down 17%. That’s really bad, but inflation was also really high.

So in real terms, it fell by a lot more, and that made it a lot cheaper. It didn’t make it cheap in the case of the US. In some of the other markets, it really did. But yes, a sharp bear market changes things quite quickly. And if that bear market happens to occur at the same time that we are experiencing higher than normal inflation, you can have a profound change in valuations pretty quickly.

Justin: What’s the idea or the concept behind the benchmark free portfolio? I’m guessing it’s like you don’t have to necessarily have a benchmark like the S&P. You can maybe have your own benchmark. But I’ll let you kind of explain what you’re trying to get at with that.

Ben: Yeah. So people tend to put together their portfolios for a couple of different reasons which are not fully compatible with each other. The sixty-forty portfolio kind of took form as some sort of happy medium because it was a nice blend of risk and return. It was a level of risk that, in general, people could live with.

And because stocks and bonds under a lot of circumstances can be natural complements, it was a reasonably sweet spot in the risk-reward trade-off. And so when we were building portfolios for our clients back in the late nineties, sixty-forty was kind of a traditional way they would want us to build multi-asset portfolios, but they would give us a benchmark.

And the strange problem we were faced with in 1999 was we were running these portfolios for a bunch of clients. All of them were unhappy with us because we were underperforming, and your clients will always be unhappy with you when you are underperforming. But we got two very different complaints from the clients.

One set of clients was saying, “Okay, I’m looking at your portfolio. I’m looking at my benchmark. My benchmark is fifty percent S&P. You own twenty-five points in US large cap stocks. So I understand the fact that you don’t like the US stock market, but that single bet is overwhelming everything else. How can it possibly make sense for you to spend that much of your risk budget on this one bet? That’s stupid.” We had another set of clients in the same portfolio that was looking at the portfolio and saying, “Wait a minute. You’ve got a negative expected return for the next ten years for US large cap stocks. Why are you wasting twenty-five percent of my portfolio in an asset that is risky and has a negative real return?”

And so we realized we were running one portfolio, but our clients were thinking about it two different ways. For the clients who were really concerned with tracking error, we were taking way too much tracking error on a single bet. For the clients who were really saying, “Okay, sixty, forty kind of makes sense from a risk framework, but what I really care about is absolute risk and absolute return,” we were wasting a piece of their portfolio, right?

Why do we own twenty-five points in US large caps? It was out of fear that they might do well, despite the fact that we hated them. So that was a piece of the portfolio that was not doing any good from an absolute risk and absolute return perspective. From a tracking error perspective, it was. But if you cared about absolute risk and absolute return, you could put together a much better portfolio.

And so what we started showing to the clients in the fall of nineteen ninety-nine was, “Hey, this is what your portfolio would look like if we didn’t have a benchmark.” And it was the fall of nineteen ninety-nine, and nobody was really interested. Even the people who hadn’t fired us yet were not saying, “Yeah, what we wanna do is get even more of what you guys are doing.”

But nineteen ninety-nine turned into two thousand, and two thousand turned into two thousand and one. And by two thousand and one the first couple of clients started saying, “Hey this benchmark free thing, that makes some sense.” So we launched the strategy in two thousand and one.

It’s designed to take similar risk to a sixty, forty portfolio. But what we promise is we’re not gonna waste any of the portfolio in assets that the reason why they’re there is fear they might go up despite the fact that we don’t like them. Everything in that portfolio has to make sense on its own without worrying about tracking error.

And so it’s a portfolio that isn’t gonna take an insane amount of absolute risk, but it feels no obligation to look like a traditional sixty, forty portfolio. We feel no obligation to own US stocks if they’re not attractively priced. We feel no obligation to own bonds if they’re not attractively priced.

We’re gonna own whatever is out there that makes sense from a risk-reward trade-off.

Justin: And would that portfolio kind of generally follow how your expected return, that chart that we were looking at, kind of looked like?

Ben: Yeah. It’s going to move somewhat more slowly. One of the things about our forecasts is they are value-driven, and one thing a value manager can tell you is, man, value gets you into and out of everything too early. So we rather intentionally move by a slower moving average of those forecasts. And there are some things that we will invest in that we simply cannot forecast.

So for example, within liquid alt space, one of the things we like today is merger arbitrage. Now, I can’t come up with a seven-year forecast for merger arbitrage. These deals are all either going to complete or fall apart within the next three to twelve months. It doesn’t make sense to think about a seven-year forecast for an activity like that.

But it does make sense to ask the question, are you getting paid adequately for taking the risk of merger deals blowing up? If the answer is yes, this is an interesting asset, and we will put it in the portfolio.

Justin: In the last few minutes here while we have you, I wanted to spend some time on some of the research that you have done on private equity. So I know from one of your papers, I think you looked at close to or maybe even over 700 leveraged buyouts going back to 1981. And what did that research project — what did that actually show that the typical private equity portfolio sort of looked like?

Ben: Yeah. So for one thing, that analysis was one of those examples of things artificial intelligence is quite useful for. And we do believe artificial intelligence is useful. One of the things we think about, we talk about super analyst. We talk about, okay, well, what would you do if you had an unlimited number of investment interns?

And one of the things you could do is, all right, look through the last fifty years at all the LBOs that have ever occurred. A useful thing about that for us is we look at the characteristics of publicly traded companies. We don’t have that kind of detailed information on the privately held companies.

But the nice thing about LBOs is they’re all companies that were once public. And so we can look at their characteristics when they were bought or just prior to them being bought. And so we thought that that was a useful thing to do because lots of our clients have very significant allocations to private equity.

They, in general, understand the fact that that private qualifier aside, this stuff is equity. And so it must embody economic risk. They think they should get paid an equity-like return from it, hopefully more than an equity-like return, but they get the fact that this is equity-like risk.

What they tend not to do is dig further in and say, “Okay, well, beyond the fact that this is equities, what kind of equities is it?” And some things are pretty obvious even without doing the work. The one group of companies that you could not LBO is giant companies, right? Private equity firms can buy lots of things.

They are not gonna buy a trillion-dollar company. You just can’t do it. So we knew before doing this analysis that it was gonna skew small. We didn’t quite understand how small it skewed. In the history of LBOs in the US, there was one LBO that occurred in a company that could have been called Mega Cap at the time, which was the RJR Nabisco deal.

Otherwise, it’s all been mid caps or smaller and hugely skewing smaller. So you kinda know that, but when you think about that from a risk and profitability standpoint, one thing that has occurred in the US over the last forty years is the return on capital of large cap stocks has been on this really nice upward trend.

The return on capital of small cap stocks has been a straight line. Now, a straight line with a lot of volatility, but there has been this increasing wedge built between the profitability of very large companies and the profitability of small cap companies. And a lot of this has occurred, we think, because of increasing amounts of monopoly power, and just market power, among large cap companies.

Well, if you can’t buy large cap companies, you’re not gonna benefit from that. So one thing about small caps over the last forty years, they have gradually had decreasing relative quality versus the market ‘cause their profitability has been coming down. Their leverage has also been coming up. One thing that was a little bit surprising to us when we did look at this data, not utterly shocking, but a little bit surprising, is that these companies when they were bought were less profitable than average. And they had relatively high amounts of debt.

So these companies, before they got LBO’d, were kind of junky. Now, on the one hand, maybe that’s not a complete shock because private equity says, “Hey, we can run these companies better.” And what kind of company should you be able to generically run better than it was run as a public company? A badly run company.

So maybe it’s not a shock that these companies were not particularly profitable and were reasonably low quality when they got taken out. But it’s still pretty striking. And given if for kind of the average endowment or foundation in the US, maybe half of their equity exposure is coming from privates in one form or another, that means they have this monstrously large bet in favor of small and in favor of junk, within equities.

And you probably don’t want to have a bias in favor of small. Maybe you could justify it. On the back of 100 years, you could say small has outperformed, but on the back of 40, you can’t.

Justin: Right.

Ben: And junky companies have never outperformed. I mean, they’re more volatile. They are higher beta, so they sometimes outperform, but they’ve been a horrible group to own.

So it is not a group you would want to preferentially own, and lots of people, kind of by mistake, have done it. And they can’t get out of it now. So the question we are asking, or we are suggesting to investors that they might want to ask is, what should they do in their public portfolio to compensate for the biases they’re getting in their private portfolio?

Justin: And how are you answering that? How are you addressing it?

Ben: Well, I mean, you can do a few things kind of in the pure passive situation. You could go long S&P 500 or S&P 100 and short the Russell 2000. That gives you both the size and the quality bias you want, particularly the S&P 100. The issue is it doesn’t necessarily have a particularly positive expected return associated with it, and it is gonna take up a bunch of capital.

We do think that high-quality companies, particularly if bought with some eye to their valuation, do perform surprisingly well. So we would argue biasing your portfolio towards quality is a good idea. And on the negative side, junky stocks, particularly expensive junky stocks, really do amazingly badly in the long run.

So if you’ve got a bias towards small cap junky stocks in your privates portfolio, I’d argue privates is the place you would want to put those stocks because maybe you can make them better managed. In the public portfolio, they’re a disaster. So kind of going long quality and short expensive junk is a nice way to balance some of the risk that is embedded in your private equity portfolio while generating, in the long run at least, a positive expected return.

Justin: One of the themes of your work is understanding why you should be getting paid on the given asset that you own. And we were kinda talking about that a little bit before we even started this discussion today. But how would you take that idea and put that to work in today’s market?

Like, how would you sort of marry those two things right now?

Ben: Yeah. So I do think understanding why you’re getting paid for the activity you are doing is really crucial. To be clear, it is not crucial because that is the way that is gonna get you rich. The way you get rich, the way you massively outperform, is by predicting the future better than somebody else or everybody else, or just getting lucky.

And this doesn’t really help you with that. What I think it really can help you with is avoiding doing stupid things. If you’ve got an activity which you have been sold as an amazing risk-reward trade-off. So let’s say, for example, you’ve got somebody who is saying, “Man, I can sell you a tail risk hedging portfolio that has a cash-like return.”

You say, “Hmm, well, I’d rather not lose money in bad times, so I’d love to have something that gives me a cash-like return and has that nice kind of correlational problem.” That sounds cool. And plenty of people have done it, and the vast majority of people who have done it have wound up really disappointed in it.

And why? Well, if you stop and think, somebody on the other side has to be willing to put up with this horribly nasty correlated return. I am going to lose a ton of money when the world is falling apart, and in return I’m only gonna get a cash return. It does not make sense for the person on the other side to continually do that.

So you should be really skeptical of someone who is promising you that you can get that lovely return. Same thing, somebody who says, “Man, what you really need to do is just buy call options on the market because you’ve got the limited downside and you’ve got unlimited upside.” Well, somebody’s got to be selling you those call options.

And how does it make sense for them to want to sell, to put up with that set of returns without getting paid something in return? And so it makes you think, “Huh, well maybe actually the return to being long call options isn’t gonna be that great in the long run. Let me look at that.” And when you look at it you say, “Oh yeah, if you’re long call options, you basically get a cash return even though you’re getting all that upside because you’re just paying enough each time you buy that option to suck out all of the goodness of the stock market.”

So it helps you avoid expensive mistakes. And where you are saying something that kind of requires investors to be a little bit odd, right? I’m saying low-quality companies underperform in the long run. Well, that’s a little weird. They have more downside at the bad times, right? Highly volatile companies that have levered balance sheets are more likely to go bust in the bad time.

So you should get paid for that. So if I’m going to say I think this group of stocks underperforms in the long run, there’s two things I need to do. One is I need to see, yes, they are priced in such a way that they are going to underperform, that the return from their fundamentals is going to be disappointing.

And I need to be able to look at any given time, are they still priced that way, right? Because any horrible group of stocks, if priced cheaply enough, becomes a wonderful group of stocks. And it helps you understand, like value, I love value from a conceptual basis, but value stocks can become overvalued, and there is nothing more supremely useless than overvalued value stocks.

They are something that has no virtue in your portfolio whatsoever. So by kinda asking these questions, I think you can avoid doing some very damaging things. It isn’t the secret to success, but I do think it can keep you out of a lot of painful kinds of failure.

Justin: Excellent, excellent discussion, Ben. Thank you very much for joining us. We really appreciate it.

Ben: Yeah, it was a lot of fun. Thank you.