Campbel Harvey, Professor of Finance, Fuqua School of Business, Duke University

[00:00:01] Hello, this is Dean Curnutt and welcome to the Alpha Exchange, where we explore topics in financial markets associated with managing risk, generating return, and the deployment of capital in the alternative investment industry.

[00:00:19] Best known for his seminal work on the information content of the U.S. Treasury yield curve nearly four decades ago, Campbell Harvey has produced meaningful academic research in all corners of empirical finance. In this episode of the Alpha Exchange, I caught up with Campbell, now a professor of finance at Duke and partner at Research Affiliates, on his recent work on gold, an asset near and dear to me.

[00:00:44] We discuss this piece, Is There Still a Golden Dilemma?, with Claude Erb that updates work they did back in 2013 on the yellow metal. Our conversation explores the financial properties of gold, with emphasis on its capacity to hold its purchasing power and to help defend against equity market drawdowns. On the first, Campbell makes the point that over the past two decades, gold has easily outperformed inflation.

[00:01:10] He adds, however, that gold is considerably more volatile than inflation is. Thus, there are periods when gold can also underperform inflation. On the equity drawdown front, Campbell's work shows that while not an explicit hedge like the S&P 500 put option would be, gold has proven durable to equity risk-off periods. We move to the drivers of the gold price, and here Campbell discusses the role of both ETFs and central banks.

[00:01:37] Lastly and importantly, Campbell's work shows that entry price matters. When the price of gold deviates from fair value, the forward return profile tends to be worse. Today's substantial rally may easily continue, but investors must be mindful of the risks of buying at extended levels. I hope you enjoy this episode of the Alpha Exchange, my conversation with Campbell Harvey. My guest today on the Alpha Exchange is Campbell Harvey.

[00:02:04] He is a professor of finance at Duke University's Fuqua School of Business, also a partner at Research Affiliates. Campbell, it's great to welcome you back to the Alpha Exchange. Very good to be back. Yeah, so we live in super interesting times, and one of the areas of focus that I personally had has been on trying to find diversifying assets. Assets that just don't behave like typical equities or those that have your more traditional risk-on properties.

[00:02:32] And gold has been a focus of mine, and clearly it's been an asset that you've done a lot of work on recently as well. So you have published an update for some new work on a paper you had done about a decade ago, maybe 12 years ago. And so I'd love to get into some of your recent work on gold. So we'll use that as the basis for our conversation and maybe also talk a little bit about rebalancing as well.

[00:02:57] So tell us first what got you interested in exploring gold and its risk-return properties. So it's kind of interesting. I thought of doing something on gold because many investors have gold in their portfolio, both retail and institutional investors. And I started to look around the academic literature to see what people had published on gold.

[00:03:24] And it turns out that there's hundreds of papers, but they're all dealing with the gold standard. And I couldn't find a paper that actually looked at gold in terms of its investment capabilities and how it interacts with a well-diversified portfolio. So that was kind of the start of it, my work with Claude Herb. And we did this historical study considering data over a couple of millennia. So gold's been around a long time too.

[00:03:54] So again, it's kind of surprising. And perhaps part of this was that for many, many years, there was a prohibition for U.S. investors holding gold. So that ran from 1933 effectively to 1971. So there's this gap. But yeah, I wanted to fill the gap and to investigate the properties of gold.

[00:04:18] And so one of the parts of your recent paper is just to kind of explore this idea, which is often suggested without not a lot of evidence, which is gold is an inflation hedge. But you do something really fantastic here, which is you go, as you mentioned, back hundreds and hundreds of years. And you're making a comparison to wages currently of army captains versus, as you say, Roman centurions. So take us through that exercise.

[00:04:47] This was a very fun part of the research. So it turns out the Romans have great records, like very detailed. So we could actually observe the actual wages in terms of the Roman centurions. And those coins still exist today, like in museums.

[00:05:08] So we actually check the content of the coin and come up with how much gold they were paid in terms of weight. And then what we did is we took that weight of gold to figure out, well, what is the value of that weight of gold today? And it was remarkable that it was about the amount that you would pay a U.S. Army captain. So what does this mean?

[00:05:36] It means that gold over millennia has held its value. So effectively, it keeps up with inflation. Another way of looking at this is that the real price of gold, so taking gold and subtracting inflation from it, is constant over a very long period of time. So the return in real terms is zero.

[00:06:04] And again, I'm qualifying this in terms of very long periods of time. We also had another example. We found out what the price of a loaf of bread was in the weight of gold in the time of Nebuchadnezzar. So this is even before the Roman centurion.

[00:06:26] And that worked out to about $7 today, which frankly, at my local bakery to get the sourdough loaf, it's $7. So again, this is consistent with gold having this long track record of preserving value.

[00:06:44] And again, I do want to emphasize that we're measuring this over centuries or millennia, not over the next year or the next five years or the next 25 years. One of the things that you do early in the paper is you decompose the time variation in gold. I think as you alluded to, you break it down into inflation component and a real component.

[00:07:07] And so perhaps just running with this idea of inflation or gold is an inflation hedge. What do you find in terms of the fluctuations in the price of gold? So again, gold's got this great reputation for the reasons I just went through in terms of its ability to hold its value over centuries or millennia. But in the short term, it's a completely different story.

[00:07:34] So many investors don't realize that the volatility of gold is approximately the same as the volatility of the S&P 500. So that means it can go up and it can go down in a big way. So gold's got drawdowns that are very similar to what you would see in the stock market.

[00:07:55] So a simple way of looking at this, if you're looking for gold to act as a hedge, inflation is not that volatile. So if you calculate the annualized volatility of the inflation rate, it's less than 2%. And you're counting on something like gold with a 15% volatility to hedge that very low volatility fluctuation in inflation.

[00:08:25] That just means that gold will be unreliable as an inflation hedge and just casual inspection of the data. So over the last 20 years, gold has outperformed inflation. So it's done its job over the last 20 years. But then you go 20 years before that, gold underperformed inflation. So it disappointed as an inflation hedge.

[00:08:50] So this is not unexpected, given that the price of gold is quite volatile and you're trying to hedge something that has much lower volatility. Almost, frankly, one-tenth the volatility. So that will just be unreliable. It seems that a lot of times when people say gold is an inflation hedge, there in some ways, I think the implication might be it's hedging erosion of the currency.

[00:09:18] That the dollar is just on a one-way path towards being worth less and less. That's what the inflation's about. And yet, if we look at, for example, at 2022, that's your most recent period of very, very high inflation in the U.S. We peaked at, I want to say, 9.8% year-over-year CPI, middle of 2022. But the dollar rose a lot, right? Because the Fed was tightening, rates were going up, and gold did very poorly during that period. Yeah.

[00:09:47] So gold failed to be the hedge at that point. If you're measuring during the inflation surge and the drawdown in the market around that. So again, you could look at various different episodes. Sometimes gold will be effective. Sometimes gold will not be effective. So the lesson here for investors is it is a good idea to have some hedging assets, both for inflation or other crises.

[00:10:16] But it shouldn't be one asset. So you need to look beyond gold. And most institutional investors have what they call a safe bucket. And this consists of various assets that they hope will hedge, for example, inflation. And it's not just gold. So gold will be there, but it could be other commodities. It could be inflation-protected bonds like tips.

[00:10:43] It could be some other real assets, but not just gold. So gold, again, is volatile. And you need some other protection. Yeah, I put it in that category of high basis hedges. And as I think you say, you want to diversify even amongst your hedges because in a lot of ways, you're not exactly sure where the center of the storm is going to be and what's going to really have kind of beta to the type of risk-off that you encounter.

[00:11:11] You did, as part of your recent piece, look at 10 different drawdowns or risk-off episodes. And you looked at the performance of gold during those periods. Why don't you run through what you found? Anything kind of jump off the page and anything surprise you from that? So this is an unpublished update of my 2019 paper.

[00:11:34] And that paper was published in the journal Portfolio Management and is called The Best of Strategies for the Worst of Times, a play on Dickens' Tale of Two Cities. So we looked at various different equity drawdowns. So these are large drawdowns. I think the one that was the least severe was over 15% drawdown and some of them over 50.

[00:12:02] And we've updated this to include COVID and the inflation surge-related drawdowns. So what we find is that gold actually has a positive return during these equity drawdowns in seven of the 10 drawdowns. In three of the 10, it's a negative return. So you might think, well, you know, a hit rate of 70% is not that good.

[00:12:31] But if you look at the magnitude of the negative returns during these equity drawdowns, they're fairly minor. So gold does provide, at least historically, some diversification. The paper looks at many different strategies. So gold is not the only one that we look at.

[00:12:55] So we look at, for example, if you want to hedge effectively, just buy a put option. That's going to work for sure. And we look at a strategy of consistently doing put options. We look at bonds as a possibility. We look at various different factors to see how they perform.

[00:13:16] And then we also look at momentum strategies, like a one-month, six-month, 12-month momentum signal, to see how they actually perform during these different drawdowns. And it's not just equity drawdowns. We look at recessions also as kind of like bad times. Sometimes they coincide with the equity drawdowns. Sometimes they don't.

[00:13:39] So to get a view of what the possibilities are, really what you want is crisis alpha. So what that means is you're in a crisis. You've got an asset that delivers a positive return. So that's what we're really searching for. And again, this has been recently updated. When you first wrote your piece on gold in 2012, this was coming off of a real bull market in gold.

[00:14:07] I'll always remember the surge in 2011. The Eurozone crisis was intensifying. The U.S. had its debt ceiling showdown. Rates plummeted. The VIX got to 45 in a hurry. And during that summer, that late summer of 2011, I want to say, gold really, really spiked. And so you'd written your paper during a time where I'm guessing, as you call it, the real price of gold, because inflation was very low at the time, was quite high.

[00:14:37] And I think you had made essentially, I wouldn't call it a forecast, but the observation that the entry point matters, that high real prices of gold are associated on average with lower forward realized returns. Maybe you can walk us through sort of the thought process there and just what you learned and maybe what we should be thinking now with gold at $3,000. Yeah. So again, think of this idea that the real price of gold is constant.

[00:15:05] At least it has been over two millennia. So then you could look at that constant price and you see deviations from it. So you might go well above that constant price. And then you might reasonably expect some reversion to the mean. And this intuition is no different than what we do for equity. So the equivalent measure for equity is like a price earnings ratio.

[00:15:33] And when you've got super high price earnings ratios, the expected returns are going forward modest or negative. And you can see this in the data. The timing, of course, is not clear. But what we do is when the real price of gold hits an all-time high or a high watermark,

[00:15:55] when you look at the returns in the next five to 10 years, they are well below average and often negative. So again, it's kind of the same intuition that we've got with any asset. So if the price is really high, it is likely and not for sure, but likely that the expected return is low. So beware.

[00:16:20] And again, the research suggests over the next five to 10 years, the returns are very modest. So as you went back such a long time frame to look at the behavior of gold and its performance through periods long ago, of course, our markets were not nearly as developed. And over time, just about everything gets financialized, right? And so in 2004, they introduced the GLD. And you've got this product, right?

[00:16:49] That's just a flip the switch way of getting exposure to it without lugging home some gold bars. Transaction costs are really minimal. You can see it and touch it on a screen real easily, get into and out of it. And so that's a new player in the market. You've also talked about, and I'd love for you to just explore both of these, this notion of de-dollarization. Maybe China's at the forefront, but other central banks have been very actively increasing their stockpiles of gold.

[00:17:19] So how did those sources of demand fit into the equation? Yeah. So we're trying to explain what's happened to the price of gold over the last, let's say, 20 years. So number one, I address in my paper and this idea of financialization, making gold easily available to all types of investors. So it's not just the retail investor, but it is the institutional investor that doesn't want to hold like a gold bullion.

[00:17:48] It's a real pain to do that. It's expensive to do that. So the financialization is very important. And we argued that that potentially was a structural shift in the demand for gold that could lead to a deviation from this real price of gold being constant through time. So kind of a step up in the price, given that all of this new demand comes in. So that's like number one point.

[00:18:16] And in the background, it's important to understand that the supply of gold is limited. So in other situations, when there's a surge in demand, supply increases and that moderates the price increase. But with gold, we produce maybe 3,300 metric tons. That was last year of gold. And it's very difficult to ramp up production.

[00:18:44] So to do a new mine or even to open a closed mine, that takes a long time. And even doing that, the amount of supply that comes in is going to be limited. This means that shifts in demand lead to sharp increases in the price. And some of the volatility of gold is exactly a result of that. So that's number one factor.

[00:19:08] The number two factor you mentioned is de-globalization and de-dollarization. So think about Russia invades Ukraine. The U.S. responds with sanctions. And many of the sanctions focused on the U.S. dollar. So, for example, restricting swift access of Russia.

[00:19:32] So, effectively, the dollar was weaponized to punish Russia. And it's pretty obvious that this caught the attention of China. So, China also is heavily reliant upon the U.S. dollar. And they strategically would like to de-dollarize. So, to reduce the dependence on the dollar.

[00:19:58] So, they're not held hostage to U.S. policy on the dollar and potential sanctions in the future. So, what does that mean? Well, we need a credible alternative to the dollar. And for China to have their currency as a credible alternative, they need to stock up on trusted collateral. And gold is a trusted collateral.

[00:20:27] So, China has been a buyer of gold. So, they've increased their stock of gold over the last two years by 15%. And this is the official holdings. And there's a strong reason to suspect that they've got much more than the official holdings. Again, this is a strategy to reduce the dependence on the U.S. dollar.

[00:20:52] It's a risk management strategy that I think everybody in finance would understand. And the important point here is if China is a buyer and we have documentation that they are, and given the limited supply of gold, that will push the price up also. China's not the only one buying also. So, India is another big buyer. Again, with the limited supply, that pushes the price up. So, that's number two factor.

[00:21:21] So, number one is financialization. Number two is de-dollarization. And then the number three factor is playing out today. And that is with heightened uncertainty, people are looking for a safe haven. Yeah. And it's trying to get the right balance. So, you point to factors that clearly underpin the story that this could be early innings.

[00:21:47] For all we know, it's just hard to know when you have behemoths like China structurally accumulating an asset and not really price sensitive, right? They're doing it with a long-term objective. And then, again, the story around uncertainty, I think, is another one. And set against your research, which suggests that entry point does matter as well. So, how do you balance, not asking you for any predictions, but how do you balance that, yeah, there's some pretty good structural tailwinds here,

[00:22:16] but the price is something you've got to pay attention to, at least from a long-term return objective standpoint. So, one thing that does worry me is that given that gold is perceived as a safe haven, that given the heightened uncertainty that people are adding gold to their portfolio, institutional investors are adding gold to their portfolio. And this creates a momentum effect.

[00:22:45] So, people see that the price of gold is going up. They buy more gold. And the price of gold continues to go up as a result of their buying. So, their thesis that this is an asset that's going up in price is kind of validated. And this works great until it doesn't. So, at some point, it stops and there's a correction.

[00:23:12] So, it's inevitable that there will be corrections, whether it's in the price of gold or the price of the S&P 500. There will be corrections. Now, as for the expected returns, we need to be careful on this because gold does have low correlation with equity. So, that means it is a hedging asset. And a hedging asset, you need to pay for. And that means that expected returns are lower.

[00:23:40] So, just because the expected return is low, that doesn't mean you exclude the asset from your portfolio. It is serving a purpose. It is serving a diversification purpose. Indeed, the ultimate hedging asset for an equity portfolio is a put option. And again, in my paper, we look at various different strategies of buying put options at the money consistently every month.

[00:24:09] And that works great. And you get every single drawdown covered by this strategy. But the problem with this strategy is in the non-drawdown periods, it's super expensive. So, overall, if you look at the average return for kind of a consistent put buying strategy, it's, in my paper, like negative 7.5% a year. So, it's hugely expensive.

[00:24:39] So, it's a negative expected return. But it delivers exactly the hedge that you want. There's other assets or strategies that you could use that are not as expensive, like buying gold. But they're not as reliable. So, it's a trade-off. So, if you want 100% protection, it's going to cost you, again, 750 basis points a year. And it's pretty steep.

[00:25:07] Also, my paper, just to preempt one of your listeners' questions. So, why do you look at money? Why not go out of the money? That would reduce the cost. Well, we actually did that. So, we looked at various out-of-the-money puts. And yes, it is cheaper, but it's not as effective. So, there's a trade-off. And indeed, the at-the-money look to be the most effective. But again, this is kind of the bound. If you want 100% protection, that's what you need to pay.

[00:25:37] If you want partial protection, you pay less. But again, the expected return is going to be lower. With any hedging asset, that's how it works. Yeah, if you look at 2022, it was a really interesting year. Because if you bought even a 5% or 7% out-of-the-money put, you actually lost money on a year when the S&P drew down 20-odd percent. That's the slow-motion decline.

[00:26:01] And it just shows you buying out of the money, unless you get a 2020-type epic drawdown or a GFC, it's tough. I mean, if the market doesn't move fast enough. So, one just last comment I wanted to make and then just solicit some of your thoughts on rebalancing. I found some of that work really interesting. But the thing that got me so interested in gold is, of course, it's got this defensive property. You make great points that you can't really count on it. You know, it's not true insurance. It can be. It can work that way.

[00:26:31] But you really can't truly count on it. And you have an asset that, by all counts, could be highly overvalued. We just don't know. But its implied volatility is actually on the very reasonable side. You mentioned the realized volatility of about 15. That's what the implied is, too. And so, to me, you have an asset that could go up a lot and could conceivably come down a lot, too. And you can buy options to express the view that it goes up a lot for a very low amount of premium.

[00:27:00] So I think that's a really interesting way of getting long gold but being able to walk away, you know, should things not go your way. On rebalancing. So we are five years from March of 2020. And that month obviously brought tremendous destruction to the S&P. I think it was down 12.5%. And the bond market was the flip side of that. That was TLT, just to use one proxy, was up 6.5%.

[00:27:26] So a almost 20% outperformance of stocks versus bonds in that month. And so you're coming to, let's say, April of 2020. And these pension funds want to get back to neutral. And you've argued that these mechanical, predictable flows, you even quantified it to the U.S. taxpayer. I believe it was $200 per person. Yeah, to the pensioner. Yeah, why don't you walk through your work there, what you found?

[00:27:56] Are there any kind of reasonable ways to do something different so you don't leave this money on the table? What did you find? Yeah, so let me go through the basic idea. So there, within the U.S., about $20 trillion of pensions and target date funds that do regular rebalancing. And what I mean by that is there are two flavors. One is called calendar.

[00:28:23] So you rebalance at the end of the month or the end of the quarter. Or threshold, which means if you deviate from a certain percentage from your target, then you start the rebalancing. So we consider both of these. Both of these are mechanical rebalancing. So it's just a rule. And the problem with a rule like this is that it is kind of like announcing your trade in advance.

[00:28:52] And if it's a large trade, that is a horrible idea. So smart traders know that these pensions need to rebalance. So if it's the case that we go from a 60-40 equity fixed income and the equity market drops and all of a sudden it's 55-45, they know the pensions are going to be buyers. And they will front run.

[00:29:16] So our study measures the amount of impact that the front runners actually have. And it turns out to be eight basis points a year. And there's reasons to believe that this number is a conservative number. But let's say it's eight basis points. So that means that the pensions would like to buy, let's say, equity at 100. Well, they're going to pay 108 cents.

[00:29:46] So eight basis points might not seem like a big deal. Let's put this in context. Eight basis points of $20 trillion is $16 billion a year. It's one way to look at it. And you mentioned the average pensioner. The average pensioner is about $400 a month. And $200 covers a lot of them.

[00:30:11] And another way to think about it, if it is $200, that is like two years of contributions over, let's say, a 24-year span. So this is notable. It's also notable in that just the regular transactions costs for a large pension are probably in that range of eight basis points a year. So this is very substantial. It is an effect that has strengthened through time.

[00:30:41] So often when you measure some effect, it might be strong early in the data and then weaker later in the data. It's actually the opposite. And it makes economic sense because the size of the rebalancing complex has grown dramatically. So it's more likely that this rebalancing has market impact. So our paper is about kind of measuring the problem.

[00:31:08] And the paper is called The Unintended Consequences of Rebalancing. It's available for download on SSRN for free. So we don't talk about solutions in the paper. That's like another paper. But it's kind of obvious. What you need to do is to be less predictable. So this idea of rebalancing at the end of the quarter or the end of the month, that should be just taken off the table.

[00:31:37] It just doesn't make any sense to do that. So a simple way to mitigate these impact costs is to make the rebalancing less predictable. Our paper doesn't say you should not rebalance. Obviously, rebalancing is important. If you don't rebalance a 60-40, it soon becomes like 90-10 and eventually becomes 99-1. So to get diversification, there needs to be some rebalancing.

[00:32:06] But it shouldn't be as predictable. One thing that is kind of a trait of some of my research is before I go public with a paper, I do want to try to validate whether the economic effect that I detail is real. And often what I do is to talk to practitioners.

[00:32:31] And this paper, when we had very early results, I was pretty skeptical about the results. And we convened a roundtable of pension managers and represented about $2 trillion. And we presented the results. And I was super curious as to what they would say. And people were looking at each other.

[00:32:58] And then somebody said, we know about this. So immediately, my finding was validated. And then I started asking them, well, if you know about it, you're being front-run. Why don't you do something about it? And they said, it's really hard. So at some of these giant pensions, there's a lot of bureaucracy. Any change in the rebalancing policy has to go through the investment committee.

[00:33:26] And it just seems very conservative and almost lethargic in terms of the ability to change policies like this. And this would be a substantial change. So nothing really happens. Then one of the pensions said something very interesting and surprising to me. They said, yeah, we know about this. And we agree that it's really hard to change rebalancing policy.

[00:33:54] So we send the signal to our alpha desk. And I'm thinking, okay, are you saying that your alpha desk, their job is to front-run your rebalancing and potentially other pensions rebalancing? I said, yes. It's one way to get it back. Yeah. And indeed, it's exactly what the hedge funds and other smart traders are doing. So this is not just hedge funds. It's sophisticated pensions see this.

[00:34:23] And they're literally front-running other pensions rebalancing. Yeah. So the paper has created a bit of a stir on X. Some people are very critical. And it's interesting. You look, they're from hedge funds. And I totally get it. They don't like this paper at all. Because I'm revealing one of their profitable trading strategies.

[00:35:14] And you know what happens. There's a gigantic split spread collar that gets rolled in the S&P every quarter. And everybody knows next week that trade's coming. And so there's a lot of argument that there's a lot of prepositioning in the options market around it. The leveraged ETF complex, another fascinating one, where they're really doing what the mandate is, deliver twice the return of NVIDIA or three times the return of the triple Q. And that causes end-of-day rebalancing that you can see coming a mile away.

[00:35:44] It's pretty interesting stuff. Yeah. No, I totally agree. So anything mechanical, you need to be careful. And rebalancing is just one example of that. Well, Campbell, thank you so much for your time today. It's great to learn about the recent work on gold. We'll see what happens. It's a very interesting backdrop. And I'll be fascinated to watch what the next year brings. Very good. Thanks again, Campbell. Thank you for having me on the show. You've been listening to the Alpha Exchange. If you've enjoyed the show, please do tell a friend.

[00:36:13] And before we leave, I wanted to invite you to drop us some feedback. As we aim to utilize these conversations to contribute to the investment community's understanding of risk, your input is valuable and provides direction on where we should focus. Please email us at feedback at alphaexchangepodcast.com. Thanks again, and catch you next time.