Financial Literacy

Episode 234: Prof. Robert C. Merton: ICAPM, Retirement, and Models in Finance

Robert C. Merton is the School of Management Distinguished Professor of Finance at the MIT Sloan School of Management and John and Natty University Professor Emeritus at Harvard University. He serves as Resident Scientist at Dimensional Fund Advisors Inc.
Merton received the Alfred Nobel Memorial Prize in Economic Sciences in 1997 for a new method to determine the value of derivatives.

He is past president of the American Finance Association, a member of the National Academy of Sciences, and a Fellow of the American Academy of Arts and Sciences. He received the inaugural Financial Engineer of the Year Award from the International Association for Quantitative Finance and the 2011 CME Group Melamed-Arditti Innovation Award. He received the 2021-22 James R. Killian Jr. Faculty Achievement Award from MIT.

Merton received a BS in Engineering Mathematics from Columbia University, a MS in Applied Mathematics from California Institute of Technology, a PhD in Economics from Massachusetts Institute of Technology and numerous honorary degrees from US and foreign universities.


Few people have impacted the way the world works, and today, we have the privilege of speaking to one of them. Professor Robert C. Merton is the Distinguished Professor of Finance at The Massachusetts Institute of Technology (MIT) Sloan School of Management and Professor Emeritus at Harvard University. He has a Ph.D. in Economics from MIT and is currently the Resident Scientist at Dimensional Fund Advisors. Professor Merton was awarded the Alfred Nobel Memorial Prize in Economic Sciences in 1997 for his work establishing a new method to determine the value of derivatives. He also created the Intertemporal Capital Asset Pricing Model (ICAPM), a popular tool to help advisors make informed financial decisions and understand market trends. In our incredible conversation, we cover portfolio theory, moving from Capital Asset Pricing Model (CAPM) to the Intertemporal Capital Asset Pricing Model (ICAPM), and how financial models work. We also discuss the difference between the value of your capital and the value of the cash flow that can come from that capital, why size can't be a factor, what aspects to consider when calculating the worth of an account, and the definition of market efficiency. We also delve into retirement, how to safely invest for it, what pitfalls to avoid, and how retirement funds may change over time. He also shares his opinion about some popular financial advise and what the roles of financial advisors should be. For all this and more, tune in to hear from the man behind the model and Nobel laureate, Professor Robert C. Merton!


Key Points From This Episode:

  • We start with Professor Merton describing the concept of market efficiency. (0:04:28)

  • He explains the basics of his ICAPM asset pricing model. (0:09:10)

  • How portfolio theory changes when moving from single-period to multi-period. (0:10:46)

  • Hear a practical example of expected returns changing over time. (0:14:15)

  • Why ICAPM is dependent on the sensitivities to risk of an individual investor. (0:22:52)

  • Find out how to determine if the correct proxy has been identified for a risk. (0:25:34)

  • Learn how home country bias fits into portfolio choice from an ICAPM hedging perspective. (0:31:54)

  • The definition of a risk-free asset and how it changes with time. (0:33:24)

  • What influence the time horizon should have on the mix between stocks and bonds in an investor’s portfolio. (0:35:33)

  • His opinion about young investors using leverage to make investments. (0:41:50)

  • What people should be doing to get more accessibility to leverage. (0:47:39)

  • Professor Merton tells us who should focus on value stocks and growth stocks. (0:51:34)

  • Discover what makes retirement income a difficult problem to solve and tips to ensure your retirement. (0:56:47)

  • We discuss using Monte Carlo simulations to estimate how much people can spend in retirement. (1:09:04)

  • He provides insight into how to get more from your retirement investment. (1:13:04)

  • Whether nominal annuities are considered low-risk assets for retirees. (1:16:48)

  • An overview of the impact mathematical models have had on the finance sector. (1:20:12)

  • Areas of finance practice that are lagging behind the financial models. (1:27:35)

  • Hear what popular financial advice Professor Merton thinks is misguided. (1:33:22)

  • Ways his work on option pricing has impacted society. (1:41:26)

  • The role he sees for financial advisors. (1:45:42)

  • Why he decided to join Dimensional Fund Advisors. (1:48:54)

  • Professor Merton unpacks the definition of product design. (1:52:14)

  • Stay listening for the extended discussion. (1:57:24)


Read The Transcript:

Ben Felix: This is the Rational Reminder Podcast, a weekly reality check on sensible investing and financial decision-making from two Canadians. We are hosted by me, Benjamin Felix and Cameron Passmore, portfolio managers at PWL Capital.

Cameron Passmore: Welcome to Episode 234. Welcome to 2023. As has been the case for the past few years, we have a phenomenal guest to kick off this year. We welcome Professor Robert C. Merton to the podcast, which Ben, safe to say, this was simply an incredible, unbelievable conversation.

Ben Felix: Oh, of course. How could it not be? Absolutely incredible. Not only incredible, but he was tremendously generous with his time. This is on the longer end of the episodes that we've ever done. The time for me, at least flew, flew by.

Cameron Passmore: Absolutely flew by. Absolutely. Professor Merton is a Distinguished Professor of Finance at MIT Sloan School of Management, also Professor Emeritus at Harvard. He has a Ph.D. in Economics from MIT. He is currently the Resident Scientist at Dimensional Fund Advisors. He is probably, or possibly best known for, along with Myron Scholes, he was awarded the Alfred Nobel Memorial Prize in Economic Sciences in 1997 for a new method to determine the value of derivatives. We have another Nobel laureate on the podcast, Ben.

Ben Felix: We sure do. Yeah. We covered a ton of ground. I could say more about his bio on a second, Cameron, but it's worth just mentioning, we went through portfolio theory, which he's made huge contributions to, moving from the Capital Asset Pricing Model (CAPM) to the Intertemporal Capital Asset Pricing Model (ICAPM), which we've talked a ton about on the podcast. It was amazing to get it right from the source. He literally created the model and the thinking behind the model.

We also went into retirement, which has also made tremendous contributions to, both on the theoretical side and a practical sense in designing products. At the end of the conversation, which was just as fascinating as every other part, we talked about the relationship between theoretical finance and mathematical models and practical finance. How does this stuff actually affect the world? You got to think from the perspective of Professor Merton, who designed some of the models that have had massive impact on the world, hearing him talk about the impact of models on finance was absolutely, absolutely fascinating.

Cameron Passmore: I was honestly blown away by that. I knew it was huge. I just didn't appreciate the vastness of the impact of these models for the past 40-plus years, right? Also, his description about why size can't be a factor was really interesting. Had not heard about it, or think before. It was really interesting.

Ben Felix: We've talked about how factors are instrumental variables for finding other stuff — that proxies for other risks, something like that. Size, as he gives us an example of, can't be a risk. It just has to proxy for some other risk. Yeah.

Cameron Passmore: Also, the difference for retirement accounts. The difference between the value of your capital and the value of the cash flow that can come from that capital.

Ben Felix: That was a whole other fascinating discussion. What should you report on to show somebody what their account is worth? Should you report on its asset value, or its annuity income stream value? Fascinating part of the conversation.

Cameron Passmore: Also, make sure at the end after we thank him for joining us, we continue to chat afterwards. He allowed us to insert back some of those comments, which were also pretty interesting. Make sure you listen right to the very end.

Obviously, a huge ton of awards he's won over the years. Be way too many to mention here, but simply an incredible, unbelievable, and productive career in finance.

Ben Felix: Yup. Enormous contributions to the field and as we discuss with him later on in the episode, to society, through the models that he's helped create.

Cameron Passmore: All right, with that, shall we go to the episode, Ben?

Ben Felix: Let's go ahead of the episode.

Cameron Passmore: All right. Here's Professor Robert C. Merton for our first episode of 2023.

***

Professor Robert C. Merton. Welcome to the Rational Reminder Podcast.

It's my great pleasure to be here.

Professor Merton, how do you describe market efficiency?

Market efficiency is essentially about the informational structure. I would say that what it does is that market price – well, first of all, let me step back and say, if you look at the functions of markets, it has a manifest function of allowing for trading to get things done. That's what we do. It is a very important latent function, which is to provide information. Market efficiency is about information encapsulation. If we just go back to the first principles, that that's a very important function of the markets to give us information.

Now, in terms of the efficiency of these, the formal statement is there's things that say, “Well, does the market encapsulate all relevant information for investing?” That's probably too broad and too generalized. What it does capture is the fact that people, it has enormous body of information. I'd speak to that first, then I'll answer specifically about – you can ask me about what I think about how good it is at what it does. I want to point out very clearly is that the market has information, very important information that no one has. No sovereign wealth fund, no central bank. We all have pieces of it.

What the market does, it brings all of us who are coming into that market, and taking actions on the basis of things we know, things we want to achieve, all right? That's how prices are formed. The market encapsulates that aggregate information from all of us that none of us have. The first point to make is the market is a very important source of information. It has information that no one has, okay? Now, the question is, how good is that information? That has to be measured relative to the information you have.

I mean, obviously, if you had information, leave aside whether it's illegal or not, that's not my point. If you saw a contract to make a bid for a stock at $40 a share, tender offer from a credible bidder, and the stock is at $25, the market doesn't yet have that information in it. In general, it has information as a result of old transactions and the question of market efficiency really is, how much does it incorporate? What actions can you take? What kinds of information might you have that would allow you to improve over the market valuations? Importantly, always remember that the market has information you do not.

Whenever you're assessing an investment decision, you have to always ask, if the market is disagreeing with me, or if the market doesn't seem to be aligned with me, that could be that I know things the market doesn't. It also could be, the market knows things that you don't. If you look at the pie, it isn't one thing. I mean, leave aside some very specific things, like a tender offer, or something like that. It's very targeted over one company. In general, people have all pieces of the puzzle and the market is aggregating them, weighted by the amount of action taken by those who have the information.

What does that mean? It means, it's not the average belief of an investor. It's a weighted average. It's weighted towards those who control more assets because they’re just bigger, towards those who believe they have more accurate information, and therefore, the action they take is bigger. It's a weighted average of what we all seem to believe about whatever the individual securities are that we're trading in. It's a tough thing to beat, all right? That's the rationale behind strategies that say, “I believe that the market has the best information for me. How can I then devise a strategy that works when I believe that?”

I'm not saying, I advocate, that's the only thing to do. I'm saying, that's a legitimate strategy. Market efficiency, and we've been testing that formally for more than a half century, by and large, it's pretty darn hard to beat the market. Is it impossible? Of course not. Are there people in specialized areas they can? Absolutely. In general, it's very, very difficult to beat the market. It's not an easy task.

How do you describe your ICAPM asset pricing model?

Oh, well. If you all – I'm taking it’s given, we all know what the CAPM is, the capital asset pricing model. Just to reveal briefly, that came about by taking portfolio theory by Harry Markowitz, and Jim Tobin's portfolio theory, and asking the question, if people really behave this way, what would market equilibrium prices look like? That's what Bill Sharpe and others did to create the CAPM. Well, the ICAPM or the Intertemporal capital asset pricing, I developed a multi-year, or an Intertemporal dynamic model of portfolios, instead of a single period, which is what the CAPM is. I developed that and worked out the theory of how you normatively would optimally manage your portfolio and all the rest.

Then once I did that and got the results, what's a natural question for me to ask? If people behave this way, according to my model, and you then put it into a context of asking in equilibrium, what would asset prices have to look if people behave this way? It parallels the CAPM in that sense precisely, all right. Now, what happens with the ICAPM is because it's dynamic, because it's multi-period, it is a richer model, which encompasses the CAPM at very special case, but no, the one-period model, but it's an Intertemporal model. Has a richer set of results of what affect equilibrium asset prices.

From the perspective of an investor, how does portfolio theory change when we move from single period to continuous time?

Oh, well, I mean, first of all, the idea that you can revise your decisions over a certain horizon and first of all, that's reality. The question is, how does that impact what you do today? That's really the answer. How do I behave differently today than as if I was just going to live for one period, like in the CAPM? The answer comes from the fact that you look at, first of all, time horizons. We talk about time horizons now, because we have multiple periods. There are different horizons. There's the trading horizon. That's the shortest period between which you can revise your decision. If markets were open all the time, it's a continuum in time. Indeed, markets have been even way back when I first wrote my papers a long time ago, markets were open enough that that approximation was pretty good. Today, of course, with high-frequency trading in milliseconds, practices come to theory. That's about as close to continuous trading as you're going to find. It's a trading interval. That's given.

Secondly, you have the decision interval. How often do you choose to make, or revise your decision? It could be the same as the trading interval. Then, what other interval do you have? You have your planning horizon, which could be your lifetime, okay? You have all these different timescales that need to be considered. Now, once you go into this model, what you can do is then solve it, with your trading in the continuous time model, we act like you can trade all the time. That's a good approximation to when you can trade frequently. We solve, how would you behave when you can revise your decisions?

I mean, just stop for a second and think of that. Suppose you had two perfectly identical twins in every way. One is here in the markets and so forth, and the other one goes off, I don’t know, to outer space, or to the darkest areas of the country. Is completely incognito for many, many – a long period of time. Well, clearly, the decision of what they do with their portfolio today would be different. That's common sense.

If I have to lock in a decision I can't change for five years, and you can change yours in an hour once you have new information, you won't make the same decisions. That is a critical element in the dynamic model, that it captures the fact that you get to revise your decisions, even though economically, you may be looking at your entire lifetime. The frequency of which you can revise that is a critical element in the optimization. That's how you set the model, is a practical model. Then you solve it by working backwards from where you want to end up, and ask yourself, how you behave at each stage going backwards to today.

The simple bottom line is, the fact that you can change your mind affects what you do today. In order to decide what you do today, you actually have to look way out into the future and then come back again, okay? That's the critical element in having a dynamic trading model. It takes account of where you're going and what you're doing and new information, and takes account of how frequently you can trade it.

To make it really practical for listeners, I'm going to offer an example and you can tell me if it's correct. Would an example be things like, expected returns changing from period to period, which will affect –

Oh, yes. Well, yes. Absolutely. You're quite right. I understand. Now, what are the elements that are going to change? I mean, you made a decision now, you have many, many periods ahead, so you need to take account of all of the effects that you will have. If you make a decision today, where does it possibly leave me tomorrow and so forth? What are the elements that are different, let's say, that a one-period model? The one-period model is used to CAPM. All you care about is end of period wealth. In the simple version, is the mean and variance of it and you're done, and you work out the optimal combination of risky assets.

In a multi-period model, because you're going to play many times, you're going to care what the environment you're going to be investing in a year from now, a day from now, or a week from now, or a year from now, in a way that the CAPM doesn't take account of because when it's one period now, and the end, then you’re done, okay?

With the multi-period model, then you're taking account of the fact that today, interest rates are whatever they are. What will they be tomorrow? In general, they're uncertain, as we know. That's one of the uncertainties about tomorrow that matters to you today. Do you see what I mean? You're not myopic and just say, “I'm going to live one period and I don't care what happens afterward,” which is what's happening there.

There are many kinds of variations, uncertainties, or risks that enter into your decision. That is the investment opportunity set, which is how the interest rates are changing, I'd say, real interest rates, or whatever. How Sharpe ratios are changing, what are the investment opportunities, lots of reward per unit risk, flat, very flat? That changes sarcastically. You have the whole investment opportunity set. Obviously, that's a big first-order risk as we've all seen. I don't think that what – if you look at what's happened to interest rates alone over the last 10, or 12 years, or 15 years even, I think we're pretty clear on that, it makes a big difference what that opportunity set is.

What comes in the dynamic model, the multi-period model that doesn't show up in the one-period model are all these systematic risks beyond just market risk. When you have one period, it's whatever my portfolio is at the end, end of story. When you're going to do it again, you're going to say to yourself, “Well, what if interest rates are 200 basis points higher next period, or lower? How will that impact me? Then, what can I do about that risk?” Because certainly, that's not forecastable, with any degree of perfection, or you can make forecasts.

You have all these uncertainties that enter into the model as a result of the time multi-period feature. Those are risks that are just as systematic as market risks. With the ICAPM, when you solve that model, first for the optimal portfolio, but then when you put it in the equilibrium of the way we created the CAPM, you now get the not only is market risk still a systematic risk and important risk just as it is in the CAPM. Now, you have the changing opportunity sets, the changing in the Sharpe ratio. Is our opportunity is going to be better, or worse next time? Are interest rates going to be higher, or lower?

Those aren't the only things. For example, in models that I've developed are more complicated, you don't have one consumption good. We consume housing, we consume various things, various sorts of things. Those relative prices are uncertain. In our bundle of things that we want to do, imagine you're an advisor advising someone, and they're thinking about what they want to do with their money. Well, they may want to buy a house, they certainly have all these various consumption things. If the relative prices of those are uncertain, that's an uncertainty that enters.

A second thing that comes in obviously, is your human capital. That your human capital risk changes. You have risk in your human capital. That's an asset of yours, a very important asset of yours, in fact, not a small asset. Especially when you're younger for most people, unless they have been born with a silver spoon in their mouth, that's almost all your wealth. How that evolves and the uncertainties about that over time are a critical element of the risks that you're facing and that you take account in the optimization.

When you’ve gone and put this in the equilibrium, you get an equilibrium statement very much like the security market line in the capital asset pricing model, that the beta, and so forth. Except, that you now have multiple dimensions of risk. Not just market risk, you have interest rate risks. Sharpe ratio. I've just given you these examples. Sharpe ratio risk. Human capital risk. By the way, even at the macro level, because as you understand from the CAPM, it's just the big macro risks that affect prices, not the individual ones. The capital-labour share is so common, we talk about it all the time. How much goes to capital and how much goes to labour is a very big systematic risk that's not diversifiable. Just as the market is not diversifiable risk. That would be an example of another element of risk that comes into the model.

The CAPM gives you the right instinct, but it's just too simple. It's not a wrong model. It's just too simple, and we know it doesn't cover it. What this does, it says, it makes it a much richer model, because now there are multiple dimensions of systematic risk. For someone like an older person like me, who's maybe retiring, not yet, but will, then I may be interested in having a shorter income, standard of living and so forth. I may want to hold long bonds and so forth and lock in my consumption. For someone, young person like that, that's not where they are. They're going to invest in a different set of risks.

The interest rate risk, for example, is different depending on where I am in my life cycle. Those should be taken account. They're all systematic. In general, you know that pension funds get murdered every time that interest rates in value, they get murdered when interest rates go up. In terms of being able to provide their benefits, they get crucified when interest rates are very low, because they just can't meet their promises with a high degree of certainty. Those are the kinds of risks that enter. You have this standard capital asset pricing with beta. One risk. Market risk.

Now you have market risk, you have interest rate risk, and you have labour capital share risk. You have all these different risks. You get, instead of a security market line, you get a security market hyperplane which is nothing more than saying, the expected return on a portfolio is a weighted sum of risk premiums for those various risks I described, times the degree of sensitivity of that asset to that risk. You get something that looks like, without a blackboard, I can't show you. It looks like the expected excess return is proportional to the beta with respect to interest rates, times the risk premium on that. Plus, the beta with respect to human capital times the risk premium on that. Plus, the beta with respect to the market times the risk premium on that.

It’s a whole list of different elements, you might call them factors, which are priced risk. Meaning, they have risk premium. Those risk premiums are different from just the market risk premium. It's a natural generalization result that comes from what people have learned from the CAPM.

Okay. I just want to ask a quick follow-up question. In the CAPM theory, there's this theoretical single optimal portfolio based on mean and variance. In ICAPM, it sounds like, there's the optimal portfolio for any individual investor is going to depend on the sensitivities that that investor has to all of the risks that are out there. Does that make sense to say?

Yes. But let me just clarify, or expand on that. What we're seeing here is that you have the systematic risk, of which there are now many of them, and individual investors are affected by them. The portfolios that they want to hold, if you want to think of they're putting together their optimal portfolio. Let's say, hypothetically, I'm an advisor, I'm trying to put, hypothetically, a optimal portfolio for you. What I'm going to say is, it's not simply your exposure to the market and the risk-free asset, as in the CAPM. It's your exposure to the market, your exposure to interest rates. All these big systematic risks, okay?

Therefore, you're going to have not just two portfolios, the optimal combination of risky assets and risk-free, you can have one of your portfolios, a market portfolio, the same one you would use for the CAPM, just for a starting point. Then you say, I have all these other risks that I need a portfolio to hedge that risk. If I'm worried about interest rates going down and I'm about to retire, I want to hedge that. It's exactly, conceptually the same as the CAPM, but it expands it to multi-dimensions. The issue is, which of those provides risk premier? That would be the question for pricing. From the point of view of the divine design of portfolio, you want to know what those are, but these are the risks that you want to expose.

Now, there are individual risks, like our mortality, which are not systematic in the sense that any one person's death is pretty independent. You can buy assets, or securities to hedge that risk as you know. You can buy life insurance, or you can buy annuities that go the other way, in case you live longer than you thought. I'm not sure if I'm completely answering exactly what you're asking, but I will say is from the individual's point of view, we tailor to all those risks.

From a market pricing point of view, risk premium factors, that should have risk premier, then these systematic factors are the ones that I've described. You can have a list of them, if you like. I'm happy to provide it for you, applying it to your readers. It's in the same idea, except the world is multi-period and that there are more risks than just end of period wealth to worry about.

How do we know whether we have identified the right factors to proxy for the multiple risks that all us investors are worried about?

Very good question. The way I would answer is, it's like anything else in science, you have some theories as to what would be there. I've given you a few, that you think would likely to be candidates with risk premium. Older people like to have long bonds. They have a big demand for them. Younger people like you may not so much. We're going to have different tastes in that domain. That's what makes a market. I'm going to want the long bonds that would be in your market portfolio, and I'm going to bid them away from you, if I want to get it sorted for a lot of old people will affect that asset’s price. I've just tried to give you a concrete example.

That's how we would look at each of these risks. We've tried to say, What are potential factors? That's the first point. They exist. They will always exist, by the way. They don't go away. People talk about crowded trades, or something goes away. They never go away, because these are real risks that are that are being born. How do I actually find those factors? It's in the end, an empirical question. Because, first of all, you can't always find the ones that you like to have in theory, and we use those to start with.

Secondly, there's a whole industry out there looking for factors, okay. You've probably seen the papers. I don’t know, 400 factors, factors for everything. That is extreme. How you do it? Well, it's a mixture of empirical and trying to say, “Is this factor a structural factor, like interest rate risk? Or, is it a instrumental variable for it?” That's a fancy way of saying, it's not that risk that we are looking at, let's say, size, for example. Size is not a fundamental risk. Size is correlated, or related to some factors that appear to be there all the time. It's a surrogate factor. There's the primary factor, that's the causal factor that the ones I've been giving you examples of. Then they're the ones you find empirically, and you have to try to un-sort those and figure out which ones are real factors that you can depend on.

That's what Fama and French did decades ago, when they said, “Well, we've seemed to identify a thing called size and one thing called a value in there is a – we seem to be able to show this as systematic.” They would be the first to say that size, obviously is not a fundamental causal factor. If it were, I mean, let me just quickly show you why it can't be. If I was a corporate finance person deciding to run my firm, and if I increased the size of my portfolio, my cost of capital goes down because the expected return is smaller, holding everything else fixed. How can I make my firm more valuable overnight? Merged with another firm. Now I'm a bigger firm, same cash flows. But because I'm bigger, I have a lower discount rate, a lower cost of capital and instant value. Well, that doesn't work.

It's very tricky to try to unscramble this. I'm sorry to be a little vague about it, but I'm trying to give you the intuition of why size can't be a fundamental risk the same way interest rates are. Do you see what I mean? Because otherwise, you have a money machine. We have to find ways to identify surrogate factors that we believe are stable, or at least we can predict when they are changing, or when they may not be stable. That's the game, of doing it right in finding all these factors.

How would I, if you want a quick prescription of a by list of what I would look for, what is a good sign for a factor? One is that it's been there for a long, long time. Something that's just been there for decades is much more – it takes decades by the way, ten years is not a long time for try to estimate whether a factor is really there. If it has a long history of being there, that's one thing. Is it pervasive? Is it true in the United States? Is it true in Europe? Is it true in Asia? Is it a fundamental element that way? Or is it a surrogate for something that's local? That work has also been done.

You want to know if it's pervasive, if it's been there a long time. You want to do, let's see, what else would I put on my list of elements? Oh, it has to be continuous. I mean, if size is a real element, and we know it's a surrogate, I've already explained that to you. If size is a real element, then it should show itself as you go across different subsets of sizes, from large cap to smallest cap, okay? It shouldn't just be for the two tails, the smallest one and the largest one. Do you see what I mean? You want to see it all the way across, if it's really size, it’s mattering? If something is midsize, it should have a risk premium somewhere between small cap and large cap. That's another test you do empirically.

Maybe you can elaborate with more questions on it. The fundamental idea is that in the end, as with all science, it's got to be subjected to the data, all right? Now, some things are easier to understand and the theory is more robust. Here, you have ideas for where those factors would come from. The actual ones that you can use is an empirical one. That's life. That's how you figure them out. People who are really good at doing that and do careful jobs, great. Now, you have another factor saying, where they just put a – they just feed it into one of these machine learning things, and they come up with elements and solve there. Those are much different than mine. I mean, they may or may not give you insights as to what's going on. Those are probably not likely to pick up the same kinds of factors that I'm talking about.

You talked earlier about hedge portfolios. How does something like home country bias fit into portfolio choice from an ICAPM hedging perspective?

Yeah. Home country bias, to some extent, there's a rationale for that, in that we don't live in a world economy in the sense of what we consume and what we do and so forth. Having a home bias is not irrational. What you need to do is figure out whether or not that home bias gives you, or your client an advantage, in the sense that if there is a risk premium, or a distortion of risk premium due to homebuyers, can you take advantage of that without disadvantaging your client?

With that, I think it's really the element of understanding what your client – what you're trying to solve for your client, or for yourself, but for your client, and then asking yourself, if there seems to be a bias in terms of too much to home and so forth, can I exploit that for my client? Again, going back to the very first question you asked about market efficiency, I have a great deal of respect for markets. When you have something like home bias in there, I take it seriously. I don't assume that it's a mistake, you understand? One needs to be very careful about understanding why there’s bias there and whether or not it affects your client. I'm sorry to make it more complicated, but that's your choice.

How does the definition of the risk-free asset change with time horizon?

Oh, well, look, if you step back – go back to investments 101, the very beginning, we talk about risky assets. The only way I think you can define risky assets is defining something which is called the risk-free asset. Once you've defined the risk-free asset, then risky assets are everything that's not risk-free. That's important. You can see it. The most obvious way is if you live in, talking about home bias or whatever, if you lived in let's say, whether you're in Canada, or you're in London, or you're in New York, or another country, you have different currencies.

What might be risk-free for you in one currency is risky for someone from a different country and their currency. You understand? Or your bundles. First, you have to define what's risk-free for the actor, for the participant, for your client. What defines risk-free? Then knowing that, now you know what's risky. What can be risk-free for me can be risky for you. I mean, obviously, if I have liabilities and everything in US dollars, and you have it in euros, if you hold US dollars the way I do, you’re taking a risk that I’m not, because of our different structures, right?

That's the way you do it. Important place where this comes up, and I use it all the time is in the area of retirement. We can talk about it at some point. What's the risk-free asset in retirement? It's very different from the traditional risk-free asset in portfolio theory, which is typically a treasury bill, or a very short-term stable thing. That's a critical element because if you don't know what's risk-free, you don't know what's risky.

I want to come back at time horizon and ignore human capital for a moment.

Sure.

Ignoring human capital, what influence do you think time horizon should have on the mix between stocks and bonds in an investor's portfolio?

That's a very good question. It's a complicated question. I'll try to break it down the way I think about it. Obviously, if you have to pay your taxes in a day, or a week, or a month, with big penalties, if you don't, then you don't put that money in stocks for the week, okay? I mean, that's just simply common sense. If you're looking at long horizons, such as, as often said, that in the long run, stocks really aren't risky because is a argument, which is both mathematically and empirically incorrect, which is that if I have a very long horizon, then I'm going to earn by the law of large numbers, or some argument, or just looking at the data, I'm going to earn the expected return without taking risk.

If I can wait 200 years, I'm going to get by the law of large numbers. That's a fallacious argument mathematically. Empirically, it's refutable. What I first want to say is, what's not true is that in long run, stocks are not risky. They are. In fact, how much money you will have at some future date gets larger if you're taking risks, the range, the farther out you go. I just wanted to, first of all, be very clear about that.

By the way, I wish it were otherwise because then we could solve all our problems in the United States. We could just tell Congress, borrow a billion dollars, which they do all the time in the United States. Put it in whatever portfolio that some pension advisor say, we'll earn in the long run 6%, 8%, your choice. Then you'll put it in there and you're eventually going to surely got to have a very big payoff far enough in the future.

I mean, let me just use, I don't know, 8%, or something. Some number like that, that they sometimes use for portfolios. If you could get 8%, what's the time horizon for a country? Let's say, the United States or Canada. I hope, at least a couple of 100 years. What is a billion dollars? If I borrow a billion dollars at 3% and I invest it at 8%. I borrowed the government rate and I invested in this magic portfolio, how much will I have at the end of 200 years? That's a math problem. I can give you a hint. 4.3 quadrillion dollars. You've heard of trillions, now we throw those around. A quadrillion is a 1,000 times bigger than a trillion, so 4.3 quadrillion dollars from borrowing 1 billion dollars today and investing it in that magic portfolio. That's a long horizon for a person. We hope, it's not a long horizon for the country.

If I have 4 point something quadrillion in the bank, the government bank, like anyone who has a lot of assets, safe assets, you can go and borrow against it. You see where I'm going. We've got quadrillions for sure. Therefore, we don't have to have any taxes. We'll just borrow for it and then pay it off at the end. We can spend as much more or less we want. By the way, if you're worried about 4.3 quadrillion not being enough, borrow 2 billion. Now, you're up to 8.6. You see my point?

The very argument used, you've all heard it, particularly pension funds that in the long run, it's really not risky. It'll all work out through some argument. If that were really true, the country has a much longer horizon than any pension fund. It can really do the job. I’m taking you through this example because I hope it will make people say, “That doesn't make any sense. It can't be such a free lunch,” and there isn't. Whatever you think is there is that way.

Time horizon is important. Obviously, you're planning a horizon and so forth. That factors into it when you solve the problem optimization. The idea that somehow, if you have a long horizon, you don't have to think about it that in the long run, you're going to get what the expected return or some other simplifying argument is fallacious. It's dangerously fallacious in that if people really behave that way, or we run our government that way, we have a serious problem. I tried to explain it by showing this little – what Jon Swift would call a modest proposal, that creates something that's completely ridiculous to say, if you don't see the mathematics of it, or if you don't see the wherever you can see there's something wrong with it. It gives too much of a free lunch for all of us forever.

We see that one popping up, by the way, as an aside from time to time with in the political framework where they say, you can borrow forever and it'll all be – work out. I don't know if that's responsive enough. I think, I'm trying to give you a sense that what is not right. Now with horizon, it's not simple. I may have a long-planning horizon, but the horizon might not have a very big impact on what I hold my portfolio each time, other than my risk-free asset. Obviously, the risk-free asset is going to pay me for sure when I want to be paid. If I have a very long horizon, the risk-free asset is going to be a very long, some bond, whether it's inflation protected, or in dollars, or euros, or whatever, okay.

Other than that, I see the time horizon as important and there are some characteristics such as mean reversion and so forth when you get into technical stuff, where if you have a long enough horizon, that you may get more benefit from it. I just view that element as just standard portfolio optimization.

As a follow up to that, Professor, I must ask you, what do you think about young investors using leverage to invest?

Well, if you work it out, I mean, I've been working on retirement solutions for a long time. If you work it out, most young people – leave aside people who have inherited, or something, but who, mostly, your biggest asset is your human capital. That's relatively safe compared to the stock market. I say, not this short, but it's relatively safe than the stock market.

Your human capital is there. Then, in terms of your horizon, I would say that it's a relatively safe asset. Part of that, by the way, I will remind you, has to do with life cycles, meaning that our tastes is to try – if we could, most people would like to have a smoother consumption over their lifetime than something that goes bouncing around. We tend to want to have a smooth consumption. That seems to be a very strict behaviour. Franco Modigliani and Milton Friedman got Nobel Prizes in part for that notion of lifecycle.

With those people, what they consume is basically, pretty low risk in their human capital. It's not risk-free, of course, but it's low-risk. It's going to determine what they want. In other words, if you get used to living one lifestyle, you're going to want to continue at least living with that lifestyle. You may want to improve it. I can promise you, you're not going to like reducing it. What that means is that being able to have that particular lifestyle is the safe thing. Well, what description was going to make for your lifestyle? If you're a typical person, it's what you've earned with your human capital. Your standard of living is going to be what you've ended up being able to earn.

In terms of planning horizons and so forth, that's where you would to try to be, for example, in retirement. We'd like to be able to sustain a standard of living you became and used to in the latter part of your work life. That element of it is there. Now coming back, if you look at someone starting out they have a large amount of human capital, which is relatively low risk, compared to the standard of living that they're going to have. I mean, some of us have higher human capital than others. The point of the matter is, that that's much safer in terms of the variation in value than stocks.

What does that mean? It means that they probably should have a larger position in taking risks, because right now with a human capital, they have it all in a relatively low risk asset. Like anything else, if you want to try to do better and you want to take on, so the answer is yes, I think it's rational and it should be put into designs to allow, say, particularly in retirement accounts, where people can have a riskier position in a retirement account because of the human capital. After all, the human capital is going to produce the contributions to your retirement plan. I mean, and so you're putting in.

You have this safe thing, relatively safe thing. You should take some risk typically, not because you have a long horizon, just because most people are willing to take amount of risk to try to get a little better outcome. Not everybody wants completely risk-free. Leveraging will be fine. Now, that said, it's very important in how it's done. You do not want people buying on margin and getting margin calls and going not bankrupt, necessarily, but could be. There are ways to create leverage for the portfolio, without putting them at risk of bankruptcy, or the extremes.

You can either insure them a portfolio, put a floor on it. Or, the way you run the portfolio has limited liability and you just do it at risk. I don't think the wise thing would have people opening margin accounts in their 401Ks. That wouldn't be the way to do it. The way you can do it is to say, when you're young, all your, let’s say, retirement, let's just be more specific, all your retirement assets are where? In your human capital, future contributions. That's relatively safe. It's okay to take some risk to try to improve what you get at the end of retirement, by essentially leveraging against your human capital.

You can't trade your human capital. You can't. You really can't pledge very much of it. You can't leverage it in the usual fashion. You can't sell it off. It’s illegal. Leveraging. I would say yes, that's a good product, or idea to have for younger people but only if your advisor or the system is taking account of your human capital when it makes these decisions. I built retirement systems, where we extracted what the person's income was, without asking them. I mean, so they don't have to do it, where sometimes we can infer just from their contributions from their plans. So, we have an idea of what their human capital is now. “Then we use that and say, what's the capitalized value of that human capital?” This is any other asset. Use that as a basis to decide how much leverage that they should take.

Can you expand more on the practicalities of that? It's not margin? What should people be doing to get exposure to leverage?

Well, first of all, when you say people, I like to think of advisors, or designed products. I don't think most of us want to be trying to do, creating leveraged positions for ourselves. If those who want to do it themselves, fine. What I would say is, what sometimes done is people put them into more volatile stocks. That is, generally, if you believe in the CAPM as a first approximation, that's probably not a very efficient way. You're much better off to have an efficient portfolio, like the CAPM and leverage it up.

Now, if you leverage it up by literally borrowing, you run the risk, and in fact, it may not even be legal to do it in a mutual fund, or in some of these things. I'm not a lawyer, so don't hold me on that. You can do it, for example, by using things like options. Options on big indices, where you can create as much leverage as you ever want to have by choosing the right mix. Then, did you have limited liability? You could hold bonds.

I'll just giving you an example. You're going to hold bonds and options, call options. Depending on what call options you pick, and how many, you can get almost any kind of leverage on you want. That would be one way. Another way can be more efficiently designed portfolios, all of them designed to have higher return producing, what I mean by that is you will be compensated. There are risks that have risk premium. You don't want to put them in a casino. You'll get a lot of volatility, but there’s not a very good return. You see what I'm getting at?

It's just applying portfolio theory as we would show it, by recognizing that what you want to look at always in investing is the total assets and liabilities of the person. Not just what they have in their retirement account, or just what they have in liquid cash. I mentioned human capital several times. You absolutely have to consider that as someone, let's say, as an advisor. You just can't, in my view, do a good job if you don't consider all the assets that are available. If I'm doing a retirement account, maybe all the assets that are available for retirement. Once you do that, then you construct the financial part of the portfolio, the stocks and bonds, stocks and so forth as part of that, but taking account of all the assets.

The mistake is to think of that all of the assets are what is in my 401K and act like that's the only thing we have. The leverage part and everything comes out of a holistic look at that, which can be done. I mean, I've done it. We've designed products and services to do this thing. It's definitely doable. I mean, this is not something that is so complicated, it can't be done. You can do it in scale, by cleverly and properly inferring things about what most people's earnings are, and so forth, which allows you to infer about their income, if you don't really know it and do it from there. That's the way I would recommend approaching it, where you take all the assets that you can into account, that is feasible, and then look at that as a whole portfolio. Not just what's being held in one or another account. I mean, we know that's the right way to do it. I'm saying today, that's doable.

You briefly touched on the idea of tilting toward riskier stocks as a form of leverage, I think, as part of what you're just saying?

Yeah.

Can you expand on who? It's probably an extension of what you're just saying. Who should tilt toward value stocks? Also, who should tilt toward growth stocks in their portfolio?

That's a very good question. I would say, as they're defined, that I don't see any particular reason, from a preference point of view, to tilt a particular more to value versus size, or something of any of those factors by themselves. I just look at those as sources of – I look at them from the equilibrium point of view. These are identified sources of what appear to be risk premium. Now, if you don't believe that, then end of story. If you believe that these are identifiable risk premium, then it's a portfolio problem, where you say, value has this risk premium. It has this correlation with other things. Do you see what I mean? Because I think it has a risk premium into it, I want to hold some of that.

I don't view the tilting of value, or growth by itself as something that's dependent on the individual. I view of it more like, getting the optimal combination of risky assets. Value is a category – see what I'm trying to say to you? It's a piece of the whole portfolio. It isn't something that I should at my age doing this. Now, what is true, is looking at where my human capital comes from. If my human capital is, I work for Amazon, or I work for, well, the tech firms, probably, I don't want to be doubling up in my portfolio on those because when you look at it holistically as your human capital, the human capital does have risk to the extent that it has risk because you're in this industry.

We all understand that if you put all your money in your company stock, that's a pretty concentrated risk and very unlikely to be a wise one from that point of view. The standard rules that we apply, what I'm trying to do is not to keep running inceptions. I'd like to be able to answer your question by saying, the way to approach things is to look at the risk. Look what it's there, say, think holistically that you're looking beyond the portfolio, looking at all the assets that you can identify that are important and looking at their risk.

You then from the point of view, or value growth that you asked about, I view those as sources of return that we believe in, or assuming you believe in them, and then asking with those sources of return, how much should I be holding of each of those? That's not really very much related because it's a little bit – It's a little more complicated in the multi-period model. Let me just make it, go back to CAPM. What matters in the CAPM, we know from portfolio theory, is that you have the optimal combination of risky assets. It's not that you're holding individual pieces. Value and size, or those elements seem to add risk premium to the basic model. You're going to want to add those. But when you're done, you want to end up with the optimal combination of risky assets. Then, what you want to do is look at that along with all your other assets to decide how much of that you want to hold and the correlations.

Then finally, you want to do – if you're going to do it, you then go say, “Look, this is the best portfolio I can put together for every period, one period at a time.” When I do the dynamic optimization over the rest of your life, I solve your lifecycle, I solve your lifetime portfolio problem. I have papers on that. I work backwards and I say, “Okay, here's what you should be holding. I'll be mixing those assets together, but I can change your portfolio tomorrow.” I’m not saying that it's a wise thing to do a lot of trading, do you understand? You have to be very careful about time horizons and everything here. There's a time horizon between which you can make changes, which is very important.

There's the planning horizon, which is your whole life. How are you thinking about this? Which is very important. Those need to be solved separately. I think you will find, if you have the papers, that the portfolios that you can hold are ones that you then revise through time to get to the global optimum. But the component parts, like value, or others are all there, just as they are now. It's all doable, but it should be broken into pieces. I don't think I would say, well, I'm in a growth area, or I'm young, so I should have growth stocks. I find that we're way beyond that in my judgment, in terms of how we create solutions. That's a long time ago, way of looking at things. If you’re going to be around a long time, so buy growth stocks. We're far beyond that, I hope, in implementation. Certainly, anyone with doing advice, or designing financial products for people to use.

Let's shift a bit to your work on retirement solutions. What makes retirement income a difficult problem to solve?

Well, first, recognizing that it's income is what you want, is not a small step. In fact, let me just point out to you. At the moment, although that's going to change, I hope, what do you see when you go to your retirement account and you see how you're doing? You see the net asset value, don’t you? That's a wealth number. Your goal, the purpose of your retirement account is to fund your retirement. What is that? That's, I would say, is some kind of standard of living and standard of living is always defined by a stream of sustainable income. Jane Austen, when she talked about Mr. Darcy is a catch, he said, “He's not each and say, he's worth a 100,000 pounds. He's worth 10,000 a year.” Described it as income flow.

If I come to your city and say to you, “Can I live like you here?” You would say to me, “No. You need $1,632,450 in the bank.” You’d say, “Well, to live like that, you're going to have to be earning here in my city so much a year.” In other words, whatever type of standard of living, always we talk about it as a flow. That feeds back into the answer to your retirement question because the goal for retirement, I would say, is a good retirement, or so forth, is an income goal. The risk is your income in retirement, not the value of the account.

If you don't measure risk right, you can't possibly manage it right. Order zero on retirement, is to recognize that retirement is about providing a sustainable income in retirement, a standard of living, not a pot of money, okay. Once you accept that, now you have what's risk-free, what matters. It's not having a million dollars. A million dollars in a 10-year bond, three years ago, 60 basis points. It's $6,000 a year, 12, 15 years ago, it was, I don’t know, 5%. It was 50,000 a year. Now, it's about 30,000 a year.

I mean, the point I'm trying to get at is that the income you can generate is a function of not just your wealth, but what that wealth can buy. That's the first thing. If you don't measure it in the right units, the units that matter for retirement are the same ones that everybody understands very well or when they're told with Social Security that when you retire you'll get this much. Everybody understands, because you can't get by what it means to get a flow of income and they can even imagine how they could live on that income if it's in real terms. You understand what I mean? If I told you, you're going to get a $100,000 in real-term, and in retirement, you can imagine what that lifestyle is like, right?

It's very natural for people, but it's not the way we do things in retirement. We always talk about the pot of money. That is converting a pot of money into what you can have in retirement is a very complex calculation. The first and foremost thing on the retirement is that you have a goal, the goal should be – a goal could be – it may not be a certain amount of money. It may be, I want to be able to sustain the standard of living I enjoyed in the latter part of my work life when I retire. Then you keep changing as your lifestyle changes, you change your goal and you keep targeting to that goal. That's a perfectly well-defined stretch.

The goal is income and it's real income, obviously, because you’re talking about standard of living. I don't want to turn this into a class. I'm just simply saying, it's important to know first, what you're trying to achieve and then to measure it, and then to say, “How do I achieve it?” Now, once we understand what we're trying to do, now, the risks that we're really worried about is not the risk of the volatility of our portfolio and value, it is the risk of our retirement well. By the way, if you think that's a hypothetical, I'll just remind you, that corporations, for example, with defined benefit pension plans, meaning they are promised to give retirement benefits, they don't measure it by what the value of the count is, they measure by the funded ratio. How much income could we buy for our constituency to match what we promised them? That's an income number. That's not a wealth number.

I mentioned that just to say, this is not some radical academic point of view. This is the way in the real world, that professional, who take on the liabilities of paying people retirement measure risk, okay, so people can understand it. That's the element that would do it. Therefore, what kinds of strategies would you have? The risk-free asset is going to be something like tips, okay? It's going to be adjusted to the duration of when you're planning to retire, not a treasury bill, which is probably one of the riskiest assets you could hold in a retirement plan. I hope everybody would agree that if you don't measure it right, it's pretty – the risk right, pretty hard to manage it well.

That's something that I’m doing in practice for, I first started getting into that in 2002, 2003, and have built things to do that, where the goals are income goals. Then I will mention one other thing in the United States and actually, implemented in Mexico, too, those are two that I know about, is the Secure Act in United States is going to require that providers show the amount of income that the account could produce in retirement, and they're going to use how they do that's being worked out. In Mexico, for example, they use actual annuities. They mark them to that. When you look at your account, instead of saying that you have this many dollars, or euros, you say, “I have so much per year income in retirement that I've accumulated.”

You can add that to social security, you can add it to your DB, because all those numbers are in the same unit. People don't have to go through enormous calculations. By the way, they understand that better. People understand if you tell them that they have an income of so much, how they can live. Much better than if you tell them they have a $167,000. This is really important. You're not going to have to train people how to understand the income number. They understand it very well. What they don't understand right now is looking at the amount that's in their account and trying to figure out how they're doing. That's a very complicated calculation.

I just put that out there. I hope other entities will follow suit. I'm hopeful in the US, we'll get it going well enough. We look at the right risk. We will be showing them when they say, read in green, the right number. Sorry for that little speech. I think it's so important that if you've got to look at the right thing, and this is a practical way to get it done, by having government put that in as a requirement to express it and do it in units that everybody's – that are consistent.

Do you think the US will follow Mexico and use –

Yeah, they've already – oh, use annuities you mean?

Yeah, to show how much income?

That, I'm not sure. You can look that up. It's a work in progress. The original was passed in 2019. The same as Mexico. The final rulings are still being written as to how to implement it. At the very least, they're going to use long-term bonds and try to match it to a income. I suspect that they – I think they may well look at annuities as some kind of guide. Then my guess is that the rate that they put into the annuity, you take an annuity, which is there, and then the rate that they put in will maybe come from the government bond market. You see what I mean? It'll be a hybrid. You won't be looking at just the annuity. You're looking at an annuity structure, but you're not going to put in what some insurance company thinks you can earn and offer.

What you're going to do is take that model, whatever rate it has, and then have a rule that you apply. The 10-year. If it's nominal, the 10-year treasury, if it's real, I don't know, 10-year tips. There'll be some rule that I suspect will link it to the government bond market to make it systematic. Because you don't want people calling up shopping different insurance companies for this purpose. You want a number that most importantly, gives people an indication in a meaningful way of what they have, telling someone that they have $800,000, if they're fortunate enough, and asking them how are they doing for retirement is not a thing that they can typically answer.

Is there a risk of showing people an income number that's too low? Because it's based on annuity pricing, whereas they may be investing in stocks for a retirement, partially?

No, but that's the future. Look, you wouldn't let me come to you and say, “I've got a great investment for you. I think it has a 30% return, and the stock is selling for a 100. Let's put it on the book at a 130, because that's what it's going to expect you to earn next year.” No, no. That's slope you wouldn't let me do. You don't let me get a bonus for what I expect to happen. I get the bonus rewards realized.

The answer is, you show them the same way you wouldn't allow me to change your – you go crazy with it. Most people would. Instead of showing you the actual value of your account on a mark basis when you go to your retirement 401K, I put in what I think it's worth. I think it's going to be worth twice that because I think stocks are going to do very well. You can't do that. I mean, that's crazy land, all right. There is a systematic way to give them better information. I believe, over time, if this is put in properly, everyone will look at that number in the same way, and with a better understanding of what they have, than they have now from looking at the AUM, or the NAV. They can't make that conversion.

As I said, there's precedent for this in the whole DB sector of retirement. That's how professional entities that are responsible for retirement view things. They view it as how much income can we buy for their constituents with what we have now? Not what's the account worth. I'm optimistic, and I think it'll be an improvement because people will understand better what they have. I hope, eventually, will understand that when they see the green and the red, that they are actually getting the right signal. As you know it is now, that if I hold the bond portfolio and rates fall, I see green because the value of it went up. I'm in big trouble, because I'm short duration because I'm going to make a bunch of more contributions in the future at a lower rate. That's actually bad for me, and we're showing them a green and saying, “Hey, they're rich.”

Whatever the precise way, the way we do it now is not very conducive to finding information, useful information to our constituents. Why I'm optimistic is the fact that the Secure Act, both here and then in US and then in Mexico, at least are moving in the right direction. I think longer term, it's going to make it easier for advisors, by the way, because then people will be able to understand the numbers they're showing them better. That's the optimist in me.

That's why I asked the question about stocks earlier because you may have said that this is crazy to do. I don't know. We use Monte Carlo simulation from a portfolio of risky assets to estimate how much people can spend in retirement, as opposed to showing them what they could get from annuity.

Well, the reason is not that I expected – let's be very clear. It's not what I expect that they will do. I'm not saying they should buy an annuity. I'm showing you value in income. If there's an annuity I can buy, not hypothetical, that I can buy, or something that's very close to that, that tells me how much income if I would do it now I bought a deferred annuity, how much income I could lock in in retirement? That's a concrete number. That's what we want to show. You don't want to show that, “Hey, I did a bunch of Monte Carlos. With my thing, I can get you to a much bigger number than historical in the future.”

You wouldn’t let me increase your AUM and say, “Your account says 1 million dollars, but it's actually 1.3 million.” Because I’m banking something that I think I’ll get. It's the same principle. When you're showing people numbers, I believe you need to show them numbers that are comparable, no matter who's showing them. If I get it from Fidelity, or Vanguard or from Timbuktu, those numbers have the same meaning, which they won't, if the numbers are projections. You can explain all you want to them, but that's a real number.

Getting people to understand what they really have and understand what's risky and so forth, to me, the more and more as we are going around the world to DC plans, where people are taking responsibility for their risk of their retirement is critical. It's doable. It's amazing how people understand very well, even with a sixth-grade education. They don't have to be financially literate. You say to them, what are you living on now? Okay, that you have a standard of living. Here's what you can get with what you have. From there, you go on – you see what I'm saying? To make decisions.

Those are concrete numbers, and they should be more or less the same number, whoever is providing it. You cannot put a number in which is a hope and a dream that you think you can make for them and lock that in. That's the same mistake that I talked about when we have people saying that in the long run, pensions are going to be risk-free. If you leave it out in there long enough in stocks, you'll get the expected return almost surely. That's a very, very slippery slope to do it.

Having been at this and having written about this and having been in practice with it, and done it, I just don't see any other way, except to do a mark-to-mark basis and tell them what they could get is best you can get for now, risk-free. Now, if you can improve on that, do so. Or you can say to them, “Look, the value of what you have now is worth so much income. If we take some risk in income, we can try to get you to a higher level.” That's perfectly fine.

To put in the expected number, or to put in things that are at risk, I'd tell people, that's the amount of income they have, I think is extremely risky. It makes it almost impossible as a practical matter to compare across providers. It's not practical, I don't think. An advisor, one on one, can have that conversation. They say, “Look, here's your income.” They understand. “We're going to try to get you to this and that,” and fine. Playing around with reporting numbers based on hopes and dreams, or what you think something's going to be worth, I think is an extremely slippery slope and one we don't really want to be in.

Yup, it makes sense. You mentioned a couple of things. One of them being that you don't actually expect people to annuitize immediately their retirement assets. What else do you think people can be doing to get a little bit more juice out of their retirement nest eggs?

Okay. I would say, let me answer that two parts. First, the annuity gives you risk-free. That's what you can buy. That's like knowing how much money you have in your bank account. Then in terms of aspirationally, then you can talk about all the rest. Now, in terms of getting the most out of it, what I would say is, you want to – it's one of the principles of retirement. You want to get the most you can from the assets you have in retirement. What that will involve is one way is annuities, right? Because an annuity will pay you more than the interest rate. It'll pay you more than the interest rate because it's promised. It's a payment. It's the whole payment, and for the rest of your life, whether you live to 120, you never have to worry about it.

The annuity is like social security. People like it. It makes sense. Whether you should annuitize everything? No, not necessarily. Unless, because you're going to need some liquidity, you may want to have – you have give gifts, or request something maybe. But in terms of where you get the juice from, you have to look at the assets. For the prototypical person with a pension 401K type person, not wealthy, wealthy people, the biggest asset they have is their house. In retirement, they need a house. In your retirement, the house is part of the retirement asset. Maybe a different house. You may downsize and stuff like that. The house is an important part of your income in retirement.

The house also will have value when you no longer need it. When you go somewhere, where you don't need money, and you don't need a house, and we hope a better place. You also have in your asset, the biggest asset that most typical families have is their house. It's often bigger than their pension accumulation, depending on where they are in the spectrum. The house has to be considered as an asset, the part of the house that's not needed for retirement. You're going to live in the house for retirement, so that's an annuity, and it's the right house for you.

When you no longer need the house, it still has value. Being able to tap that value and convert it into more income while you need it, is an important thing. That has to be done efficiently, of course. I'm not going to get into a long discussion about how a redesign, reverse mortgages and things. This is not the point. The concept is, that typically, for most people, the biggest asset that they have. In fact, it's almost all the saving that most people do. They don't do a lot of other, until they get up to be more affluent.

At the most important level, you need to look at the house is an annuity you're going to live in. It's the house you're going to live in. But also, that it has residual value and that, can you – you tap that value to increase your benefits. Optimize your assets. An annuity does that for you, because it gives you income for as long as you need it. You never outlive it. When you don't need money, you give it up. That seems like a pretty good deal to me that I have money for as long as I need it, no matter how long I live. That's a whole another story as to how to market that. I think that has to be marketed. People love social security and that's nothing more than an annuity, a real annuity in most places. Anyway, I’ve run on too long about this, I'm sure.

Most annuities that are available for purchase, Professor, are nominal, correct? Are these nominal annuities still low-risk assets for retirees?

Well, that's a very good question. I'm sad to see, there was a time not that long ago when I was actually when I first got engaged in building solutions, where a number of providers provided actually real annuities. The last one that I'm aware of, at least in the United States, stopped doing it in recent years. The problem is there are not real annuities I'm aware of that can be purchased. On the other hand, I also know that the world changes. We're in a period here, not just in the United States, of some pretty substantial inflation relative to what people have lived with. That's a wake-up call, I hope, or will be to people that inflation is not a hypothetical. That we've gotten so many years, where we had very modest inflation and so forth and we didn't have much discussion from it.

I'm thinking that perhaps, out of what we lived through here, we may have that. If that happens, then I think the insurance companies will start writing real annuities again. They stopped writing them when nobody would – they couldn't sell them. They couldn't sell them because we had decades of very modest of any inflation. It just didn't work. My answer to you is, it can be done. It can be resurrected. Is it available today? No. What you can do is at least for the accumulation period, the risk-free part of your portfolio can be in tips, or I bonds. If you do I bonds, they're great. They're just limited in how much you can buy. This Is US state. Obviously, different in different countries, okay?

You make sure that the income, or risk-free income of the risk-free part of your portfolio is risk-free in terms of inflation. That you can do without an annuity. I only use annuity to tell you a valuation of what you have. It doesn't mean you have to take the annuity. I hope that they bring them back. If they don't, then there are less effective means of doing it. Lots of insurance companies provide so-called graded annuities, where they try to guess what inflation will be. My answer is I'm an engineer by mindset. I have Nirvana what I'd like to have, but then I believe you actually have to have a solution. I'll do the best I can, at least on my part, to get you that real thing. I'll figure out a way. If there's enough people that worried about it, I have confidence that we've done it. We can bring it back.

Today, no. What I want to avoid is people say, buy stocks because they'll protect you against inflation, or buy a bunch of other surrogates. We've seen the history of that. The tracking error is horrendous on those things. What's worse than being uninsured is thinking you're insured when you're not. I've got you covered for inflation, and then you find out you're not because I didn't write that guarantee. I think that's really important in retirement because it’s a serious part for people. We have to take that very seriously. Sorry to run on about that.

You do not need to apologize for that. It's fascinating.

Not at all.

I want to move on to your – some of your perspectives just on the practice of finance more generally. How significant, throughout your career, has the impact of mathematical models been on the practice of finance?

Wow. That's a great question. Very significant. One of the things of having been around, I've been involved in the industry for 50 years. I lived through it. I'm not making speeches about it. I'm just saying, it's been a remarkable trip. The real breakout in mathematical models being really used in the mainstream of finance practice, came starting in the 1970s. Really was in full bloom by the late 70s, into the 80s. I happened to have the great good fortune to be at the right place at the right time with both the work that we did academically and then implementing it. I’m just giving you that as background.

What drove that, by the way, was two things. The science, we happen to been lucky enough to develop models that were quite useful for investing. Circumstances, you talked about a real annuities, you're not going to get anything implemented, unless there's a need. The crisis of the 1970s, which you had double-digit inflation, you had all kinds of shocks to the system, creating enormous risks, created a response in the 1970s of huge innovation on risk. We created financial futures. We created the first option markets, which is financial insurance. I could go on.

The amount of innovation instruments and so forth that were created, driven by the enormous risks, remember, currencies were all fixed. Suddenly, all the currencies of the world were floating all over the place. No one had dealt with that in who knows how long. We had double-digit inflation, which no one had dealt with. Double-digit interest rates. You name it, the 1970s, the oil prices from OPEC. You wouldn't believe how many. But instead of people pulling back from innovation, it was an explosion. We created these markets. You cannot have done that stuff. You could not have created the options market. You could not have created all the portfolio things, international diversity, without the mathematic quantitative models and data that we had.

We had the tools, we had the stuff, we had the data. The two came together. That's when finance theory, or finance academics, finance science and finance practice became married. That marriage has continued to today. Having lived through this. How important are they? They're built into everything we do in finance. I mean, people tell you about FinTech and so forth. This has been going on for 40 years. You know what I mean? This is not new. In fact, mistakes are taken in a T. You're talking about betas. We don't have time now. I could tell you stories of what it was like to try to explain to fiduciary decision-making people who set the fiduciary rules, why holding a portfolio of index stocks where you don't even know what they do, was a prudent thing to do. Indexing.

I mean, this is a much longer story. I won't burn up any more of your time on that. The point being that mathematical models have been embraced, are embraced, and are embedded into everything we do. The options market wouldn't have been possible to operate without the mathematical models that were developed in portfolio theory. Then when you add in all the technology and FinTech today, and all that blockchain, all of this stuff, it's only going to grow.

That said, it's really important to understand that these are models. They're not reality. Models have error. Every model is incomplete. Therefore, every model has errors in it. To say, well, we have a big machine that gives you all these answers and churns out all the data is not good enough. You need to be able to trust what's in there. That is a major challenge. That has always been a major challenge. The fact that it's mathematical is very powerful because we can solve problems we couldn't. It also creates a discipline and a common language for measuring risks. You all are familiar with benchmarks and so forth. We've developed an enormous amount of technology for trying to monitor performance and test performance and so forth.

At the end of the day, they're models. They're only as good as the abstractions that are into the models. I'm going to get on my soapbox for just a second as a scientist. I don't care whether in economics, or physics, or life sciences, all right. Everything is a model. The most important thing you do as a scientist is to make the right abstractions. All models are abstractions from complex reality, right? If you pick the right abstractions, you create a great model.

It's all about picking the right abstractions. What do I leave in the model and what do I don't take account? Do I put in multi-period? Do I not? Do I do all these things? Those are the abstractions. It's the art of the science that's critical to good models. That's what made the difference. That would have been true of Watson and Crick versus Linus Pauling who were competing for DNA. It's always important to remember everything's a model. In the end, what makes a good model is abstractions, what you choose to accept or not. Then the rest churns out.

I noticed how this sound abstract you and I don't mean to play on words, that abstraction and abstract, but that's the reality. Trust in the models is critical. This stuff that says, you don't have to trust because I've got XYZ blockchain, or whatever. You have to trust the person who created the blockchain. Did they do it the right way? You can't get around that. That's why, ultimately, we do all the testing. At some point, you have to trust that they're doing it. That becomes a critical part of the whole process. End of standing on that soapbox. I think it's important to understand the fact that they’re mathematical allows to do amazing things. But the fact they're mathematical, doesn't mean that they're right, that they have the right assumptions in them.

For that, as far as I know, AI has not been developed anything close to be able to provide assessments of abstractions. It can do data mining, that's fine. We all know about that, the limits of that, and we should not forget it. This is not to disparage that approach, but it's to say, the idea that somehow that data will tell us the story is not what anyone I think seriously in AI, or anyone else thinks is a sufficient statement, off the soapbox.

What areas do you think finance practice is lagging behind the models?

What we're capable of doing? The reason I'm hesitating here is because we are making so much progress there. I think that the elements that in the industry itself, I think we should design our products. Now, this may not sound like this is answering your question, but I believe it is. We should design our products based on what people know, not expect people to become educated to understand our products.

For example, I'll give you a concrete example. Auto industry. If you got into a 1956 Ford today, you'll find a steering wheel, some kind of gearshift, accelerator, and so forth. If you push that accelerator with your foot, you know what I'm talking about, you know where it is, right? If you get into a 2022 Ford today, the accelerator will feel identical. That's not an accident. They don't put the accelerator on the ceiling, or on the door, or on the steering wheel. Because you know how to drive, and so they maintain that. They not just maintain location, they maintain feel.

I mean, I was a actual auto engineer at one part of my life, so I use automobiles as a mass consumed product. You go to rent a car, you've never driven that vehicle before. For the most part, you can do it. Why? It's not an accident. What it does is that design says, what do people know how to do, design a solution based on what they know, rather than design a solution and tell them they have to get educated. Example I gave this concrete before is if you show them income, they understand income. They do not understand wealth an indicator for what they can have in retirement, for example.

By the way, where else do you see this? In consumer electronics, right? We all have heard this story, if they have to look it up in the manual, you're going to have a failed product. They would take it out of the box and be able to use it, which means you have to design it, so what they know allows them to use the product.

Now, to come back to your question on practice. I believe, this is the most important – among the most important principles of design that we need to start to do. Instead of telling people, I got to get financial literacy, and they got to go – which is fine for some people, that's the wrong way. Look at what people know, and design on basically, what they do. Why I bring this up is because theory and science can help you to do that in design of your products. I think, that's really important. I'm optimistic, as much as I can be. Does that make sense to you, if you think about it, that shouldn't we make it easy for people to make decisions? You'd be amazed how people can make a decision if you frame it in the right way.

All of this is where I think most important in my mind is where practice needs to go to the communication, so that people can make intelligent decisions. Now, some decisions, they may have to make the ultimate one, but they can't. They can't do analysis, like you're going to your doctor or something. You're not going to do the testing and evaluation of whether type B, or type A, or something should be used. In terms of our obligation in practice, to me, it's to make the products and services and the decisions people make, important decisions, be as easy to understand and where they can do it.

I've designed, and I don't want to get off this, but I've designed with another person, we designed a bond, for example, and I'm not going to go into that, which was designed precisely to be able to be used anywhere in the world, with people with sixth-grade education, by designing it in a fashion that they relate to. I won't go any further. I'm saying, this can be done. Admittedly, it's a big thing with me. I just think it doesn't make sense to ask the masses of people to become financially literate, so they can understand finance speak, betas or whatever, or that kind of thing. I view that as our obligation of the industry. That's my big thing I wanted to say for about that today. I mean, I just think, by way, give this some thought. I've been trying to do that myself. I think it is doable. It's different from the way we – that I'm aware of that we produce things.

That's fascinating, because it is – There's people can get more financially literate, or products can be made such that financial literacy matters less. There are two solutions.

Fine. I don't have to understand how an internal combustion engine works to drive a car. I don't have to become – but somehow, we think that people have to understand interest rates, compound interest, blah, blah, blah, blah, blah, in order to do this, and that they should get educated that way. Wonderful, if you happen to have that knowledge, but not something – to me, it makes no sense. I think, it has implications for the way services are provided now and how they could be provided. This is my view. Very doable.

Is there any popular financial advice that you find particularly misguided, or harmful?

Well, the reason I'm stopping on that, there's many kinds. I mentioned one of them, this idea that in the long run, it's not risky, is probably the classic one. I always point out with no disrespect, but since you asked me that question, look at Japan. Right now, it's the third largest GDP in the world. It was the second largest GDP in the world for the last – most of the last 30 odd years. It's a wealthy country. The Nikkei in 1990 was what? 39,000. Today, where is it, 33 years later? This is a country that has been politically stable, still wealthy, no major disasters. I mean, they've had their disasters. That's 33 years.

I mean, the Nikkei is not even where it was. To say, it will all work out in 33 years. This is not some small country. This is a big GDP, successful and still wealthy country. It’s not a country that had a revolution, or anything else like that. I only put that out as to say, “How can you go up and tell people that are in the long run of 30 years, or something for your retirement, you don't have to worry. It'll be safe.” When it is a shining, major counterexample to that sitting in the middle of the world? I'm sorry to get on that soapbox. But it just seems very strange to me that nobody seems to have noticed that. Either that or they say, “Oh, that's just special to Japan.” That’s the exception. I don't know.

How do you think your work on option pricing – we asked about the practice of finance, but taking an even bigger step back, how do you think your work on option pricing has impacted society?

Well, of course, I'm a bit biased, since I had to do it. I would say, certainly, within finance, it's had a really a very big effect. A lot of it is not even in the original use of options. The work that we share – I shared the Nobel Prize with Myron Scholes and whatever, with Fischer Black, for in-options. The actual statement of what they gave us the prize for was not for options. It was for a new methodology to price derivatives. The methodology, in particular, my contribution that I brought to the table was how to manufacture, or produce, or to understand the risks of any derivative product, not specifically an option. That ended up getting used.

Well, now, today, if you look at US alone, in last year, or the year before, they had 10 billion contracts that traded in options alone. In the world, there is between 600 and 700 trillion dollars notional amounts of derivatives there. That's world GDP is a fraction of that, okay? I mean, so these are enormous things that influence. They are incredibly powerful for transferring risk. Finally, I would say, if you think about decision-making, what is an option? An option gives you flexibility. It gives you choice after you've seen the outcome. You don't have to exercise it until you see what happened.

Think about how all of us make a decision. We make a decision. We know what we like to have, or we're offered something. Then we think about what could happen, the uncertainty of it. Then we say, “Oh, I'd like that.” Then you want to know the price. Option pricing gave you the price. It tells you what it cost to get the right to be able to make the decision afterwards. That's how we do decision-making. In that sense, options are fundamental to the process of decision-making.

The only thing I would add that's more relevant here, for the context of the show, is that if you look at the means of controlling risk, there's diversification, there's hedging, which is just going into the risk-free asset, substituting, and there's insurance. Options are insurance. They're one of the three fundamental ways of managing risk of all kinds.

They all are used a lot. My belief is that you'll see both in the institutional area, but also in retail, much more uses of the third way of managing risk. You see it, for example, in creating floor products in retirement, insurance companies writing, what are they doing? They're all about options. It's one of the fundamental ways of managing risk. It isn't just a product. Finally, the structure, if you look at corporate bonds, if you look at the capital structure of a firm, valuing the capital structure of the firm is a bunch of options. They’re derivatives. They’re derived from the value of the assets of the firm. The applications are very wide.

Now, I won't keep going on about it. It also changed, created what was really financial engineering. That would be a much longer discussion. I have some wonderful stories of big impacts on society, or not just in the United States, by using financial engineering. Now, it is some joke, but as a way to solve problems. That all got supercharged, when the option model was created. Then the option market was created. It was used there almost from day one, because it solved the problem. I could go on, but we created a whole industry of solving problems by creating solutions, using those same models.

Obviously, I may have some bias, but this idea that are we going to have derivatives? People say, “Should we have derivatives?” They are structured. Every central bank in the world uses them. They are there. They're done. They're as much a part of the financial structure as anything else. This is how the central banks do things. This is how banks – by the way, they're also an incredibly valuable source of information extraction. That's a whole industry in itself. The simplest one you know is something like the VIX, the implied volatility. Where you're looking at market price, important markets, being where there are serious people playing, and you're able to extract something from the price looking forward about the future.

These are not historical. You're running a regression on the past. This is the price today that someone not only who knows about the past. This is going back to efficient mind. They know everything else that they have around them. They know what's going on in Russia. They know what other things are going on that we don't know. They are paying these prices, which means these prices contain information looking forward, not historical. They contain. That's a very powerful tool. Central banks, particularly in the United States, I think, it's the Minneapolis Fed and the Bank of England, several of them have gone into extracting that information. You can get more than implied volatilities. You can get the whole distribution. This is a big use of information extraction.

I could go on for hours about this. That's a very important rule. Because remember, I said at the beginning of our discussion, that the market has information no one has. It's certainly not infallible, but particularly, at times of stress, I tend to look to the markets, because that's what real people are doing, who have things. I get more from that than looking at historical data and say, it shouldn't be this way, or something else because there is a forward-looking. I know I am going on here. This is to me, very exciting. I begin to be involved in it. This will allow us and does, I mean, allows you to forecast future volatility, much improved over just using historical data.

That means, among other things, that if you can better estimate risk and distributions that's improvement for everything. It also means that you can create the product. I'm sorry to run on, but I'm going to do it. You just let the tape run. It's your tape, but I'll send you a check if you need it for it. All right, the thing I wanted to point out to you, if you are able to forecast volatility, not to make money because you're forecasting for the market, so you can't make money on the VIX, because you're extracting it from things like the VIX, you understand. This is not a moneymaker itself. If you can do that, what can you do? It means that you can then stabilize the portfolio's volatility.

If I can forecast what future volatility is going to be, then I can adjust the portfolio for – in other words, if I can do a good job of forecasting realized volatility, what the market is up and down, just the volatility, then I can institute strategies to stabilize the volatility of the portfolio. You say, “Well, why is that important?” Well, that's a much longer discussion. Let me just point out the following thing. You've all heard of 60-40, 70-30. In your opinion, ask yourself, what does that mean to a typical client if the advisor says, 60-40, 70-30? We're going to do that one. It's an indicator somehow that you're stabilizing the risk, right? A lot of people lead, 60-40 is less risky than 70-30. We're stabilizing the risk by somehow keeping the same asset allocation mix we’re somehow stabilizing the risk.

Well, how else could they interpret it? They don't understand what an asset allocation is in the abstract. You're telling them, you're giving them the impression. I bet you, if you go into interviews and that's what they interpreted to me. Somehow you're stabilized at risk. Then the argument is, look, in general, unless we have a reason to change it, you got to want to take the same risk, until you change your mind. We'll keep you at 60-40, or 70-30, or 50-50, whatever. Then, okay. Now, if you accept what I just said, stabilizing the volatility of the portfolio is much better than stabilizing the asset allocation at 60-40, 70-30. Why? Because the one thing we know for sure is the stock market volatility changes all the time.

If I do 60-40, the risk of my portfolio changes all the time. It's not constant risk. I'm not trying to turn this into a class, but I'm trying to tell you some very fundamental out there to advisors. They think that by tying a fixed asset allocation, that somehow, they're creating a stabilization. When in fact, you can do much better than that because you can use these devices to stabilize the portfolio. Now, it's not perfect, but you can say to the client, “Why would you want to change your risk from one day to the next, or one week to the next, unless you had a concrete reason to?” The answer would be, probably not. You want to have same risk, unless things change.

Well, the way I can do that is by stabilizing the risk. You see that? That's better than say, I'm stabilizing the asset allocation because just look at it. We look at the VIX. The stock market that you're using in your asset allocation is volatility is changing all over the place. Empirically, you look at it. You’re very clear that the volatility is not constant. I'm trying to give you something very concrete and basic. That's an example of how we can use these tools to get much better forcasting. By the way, these implied numbers really do improve the forecast over historical numbers. It's not the historical numbers are irrelevant.

Those markets are giving us important information that we can use to improve that. That's where I see a whole suite of products coming. Now, I've burned up more tape than I can afford to pay for. I'll turn it back to you. I can't resist that because it's so exciting that you can do that. It's being done, by the way. This isn't hypothetical. It's just at its beginnings.

You mentioned financial advisors. What role do you see for financial advisors?

Well, obviously, it depends on what. They can have multiple roles. First of all, that financial products are complicated and they're not transparent to the holders. Like a doctor, they can give them advise on those products and explain it to them and tell them that things that are there. I mean, that's pretty basic. They can put together plans. They can talk to people, and we can get much better at it. Some are doing that by looking at the lifecycle of the person and extracting from that, what they need to do, rather than just mechanically saying, 60-40, 70-30, whatever, all right?

The advisor themselves can do a huge job. The reason is that you're not going to hire an advisor if you don't trust them, I don't think. If you trust them, they're in a position to, therefore, help you get to a good solution, you can't get to yourself, all right? Trust is an absolute essential. That's what I see the advisor providing. As I already said, mathematical models by themselves are not trusted. Despite what they say, technology by itself is not trusted. I don't think you would take your cellphone and put it in, ask, “What should I do about my sore leg?” If it came back and told you, cut it off, that you would do it, all right? You don't trust your technology by itself.

Technology combined with things, yes. That's the role of the advisor, I say, that however they do it is providing that trust. The same trust that you need for your medical. At some point, they're going to explain things to you, at the end of the day, you're going to have to trust the person that’s in the operating room, or you're going to have to trust the person that's prescribing, because you have no other choice. I'm sorry to run on about this. I view that the issue of trust is absolutely essential. You cannot verify most performance. We know that. If you could, there will only be one provider. You only have two choices, three choices. Verification, then you don't need trust because you could just do it. Transparency, which means you really understand what you're saying. That's limited. If you don't have transparency and you don't have verification, then you have to have trust.

I think the role of the advisor is trust. If advisor is your advisor, they have that asset. Maybe they shouldn't, but they do. They have your trust. Therefore, that's the important role I see for the advisor in the mechanisms. I don't see it being replaced directly by someone who just claims that they have a mathematical model that does it, unless the provider’s a very trusted entity and probably, that doesn't work.

You became affiliated with Dimensional after your Nobel Prize. Some of the other people who have Nobel Prizes, they were affiliated beforehand. Presumably after a Nobel Prize, you could have worked with any asset manager. What attracted you to and continues to attract you to Dimensional?

Well, many things. The first is that I was a part of developing the science of finance way back and have continued, and then bringing it into practice. I mean, I was, as I mentioned earlier, that's a great kick because you're doing something. You're signing. You have a sense that you're performing something meanwhile. Dimensional from the beginning, was a science-based firm. David Booth who started it with Rex Sinquefield, it was all based on the finance science. All the stuff they've done is that. The way they run the firm, the checks and balances and the way the arguments, discussions go on, what to do are all done along those lines.

It has to be, what's the series? What's the principles? Why do you think it would work? Then put it to the test of the data and recognize this. That's a very comfortable place to be, if this is something you're doing. From that point of view, it was very comfortable to be there. I was a director and independent director, meaning I had no affiliation with – other affiliation with Dimensional Funds for a number of years. I also was on the other side of the table representing people, and I saw what that board did, not counting myself.

I will say this, and this is an opinion. I have never seen mutual fund board members of the kind that they had, and still have. If you look at the list of who they are, if you watch what went on, I don't want to turn this into an advertisement. This is just a fact. They know what's going on. They are not people who don't know about medical devices around the board, and just have to listen and hope that what they're told is true. They really know that stuff. They're not shy about bringing it up.

By the way, to Dimensional’s credit, that's what they wanted from their board. They want to get the feedback because they believe – and I won't go into that. Let's just say that it was a very comfortable place. I have a great deal of respect for how they did it. Then, when I developed a bunch of solutions and so forth, I gave up my independence and became affiliated with them. I'm not an employee or anything. I'm a consultant to them. I became affiliated with them, because why not? I don't want to turn this into an advertisement for Dimensional. Let’s just say that it was a good place to go.

I've been some very good places, by the way. I'm not trying to say that it's the only one. This was built on science from the beginning, so it's a very comfortable place. It's a discipline that you don't see very often, where it's all driven by confronting it the way you would in testing hypotheses and so forth. That's how the whole thing is run. When they develop new products, that's how they do it. It's a very comfortable place for me, and one I believe in. I guess, that's enough.

Great answer. I've just got one follow-up from earlier, before we before we let you go. Should be quick, I hope. We were talking about value and growth. You talked about how a young person should be exposed to growth. That's old thinking and product design has brought us past that. What does product design mean?

Well, what I mean by that is categories – look, when we're just having a conversation, I tell you about growth, or something, that's fine. The way I look at it is we're solving a portfolio problem. We're using models to do that and they help us to quantify it and to monitor it because it's very important that you also use these models, not just to create the positions, but to monitor how the positions are and see if they're behaving as they should. It's an ongoing process. That discipline, you can't do without it. That discipline is really important. Well, I don't know. Is that what you were asking?

Yeah. I don't know. John Cochran had a paper fairly recently about basically, ICAPM and how we should maybe be applying it. One of the points in the paper is that maybe young people should have more exposure to grow stocks. That's why I asked the question.

Oh, if you can identify a theory that says growth stocks happen to fit, I had no problem with that at all. I'm saying that the approach I have is, I look at the data we have. I look at the CV and say, “Here's what I think should happen.” I have a whole talk I give on sources of alpha, for example, and so forth, as a way to understand what's going on out there. I won't bore you with that. To come back to it, if that can be shown to be as part of a portfolio decision, the general portfolio decision, but I want to see it anchored to the portfolio. Go back to the portfolio model and say, “Why are we holding this?” The ICAPM is the equilibrium version of the portfolio theory that takes account of that because it's dynamic. It takes account of the changing interest rates and the changing all these other elements that are important. That's generic in that sense.

How we apply it is as we do with the CAPM, we can use the model to try to identify dimensions of price risk, or dimensions of return. Then use those in the portfolio model to figure out how to mix everything. That's what I mean. Of course, when you say in terms of value, growth, size, and all those things, those are just elements that go into the model. I have no problem with that. It's just that I don't think young people should be associated with growth stocks, for example, off hand. I don't see it. If John – I hadn't seen this particular one, has a theory of which growth stocks somehow have that factor. That's why growth and value have the elements there, that's fine.

We could go off into another question is what's the underlying explanation for value and so forth? That'll be for another day. The point is, that, of course, is just to understand. Going back to the fundamentals of portfolio selection, I push back on trying to oversimplify it when you don't need to. You can simplify, but simplify along the lines of saying, “We're trying to do portfolio theory, or we're trying to choose portfolio analysis to get people to a good place.” We need to know estimates of the means variances, covariances, etc., to do that. We have data and we have other things to do that, and then we need to do some optimization. Then we have to develop the products to execute that. All of that is part of the analysis.

I don't see some shorthand that says, young people should hold growth stocks, or young people shouldn't hold growth stocks. If someone provides that analysis because I already have mentioned, human capital is an important element. To the extent, if you're a stockbroker and you make your money from that, probably, that should influence how much stock you put in your financial account. Contrary to what some people would think. Intuitively, the person on the street would say, “Well, they're stockbrokers, so they really understand stocks. They should put their money in stocks.” Not understanding that they have a huge amount of their wealth in stocks, in their human capital, and that it doesn't – it may be wise to diversify that with their financial capital. Do you see what I mean it?

I don't know the context completely of what you're asking in this question about growth. I don't think that those are the monikers that can be used, should be used. I think they should be used in terms of looking at the portfolio model and relating to that.

Okay. That makes a lot of sense. I could ask you more questions, but we've taken up a ton of your time. You've been extremely generous. We appreciate it. Professor Merton, thanks. Thanks so much for joining us on the podcast.

Okay. Well, it was a pleasure. I hope you got something that you could use.

We sure did. Professor, thank you so much. It was an incredible conversation, and we wish you all the best for 2023.

Okay. Well, good luck and I’ll look forward to hearing about your future, what you're up to and doing.

John Cochrane was actually referencing a John Campbell paper, but I think it was that growth stocks are very sensitive to changes in the market discount rate. Therefore, they hedge against future investment opportunities, which makes them more suitable for investors with long time horizons. That was my understanding.

Well, if it's that context, which I had been aware, I mean, I know what John Campbell's done and just wasn't aware of that was the context. That's fine. That's simply not – I say simply. That's saying, I want to be able to look at how the opportunity side is changing, capitalization rates are changing. I found a vehicle. That's the one that is my point. My point is if on the model, it tells you that's what you want to have. They're saying, how you could do that maybe with growth stocks. I'm saying, fine. That may be. But actually, growth stock is a little more complicated because growth stocks have lots of options in them.

Growth stocks have huge amounts of intangible options, technology potential, etc., etc. Well, my view, and just off the record – not off the record, but just my view is that when you're trying to value that thing, you should value them as options, not by a discount rate. The same reason you don't use a discount rate to value options. Doesn't work very well because it's highly non-linear. There isn't a constant discount rate because the discount rate changes with risk. The right tool for the job, I could – as a financial engineer, I tell my students and for myself, I want to have the biggest toolkit I can when I go out on the job. I want to have everything. It's very important to choose the right tool. Discounted cash flows, or cost of capital is very useful for some things and a disaster for others.

You have things like options that are highly non-linear, that's why it didn't get priced that way. That wasn't the way it was priced. It is not priced that way. No one prices options that way. Precisely, because there's no constant discount rate across the capital, on the actual – and you don't have the expected cash flows.

Finally, the good thing about option pricing is you don't have to know the expected return. You don't even have to know the expected cash flows to get the right. That's what, by the way, I'm running on, that's what drove the industry in 1970s because everybody understood that it was putting your finger to the wind to say what the expected return was. You do your best, but you know. Here, you get a valuation theory that doesn't depend on that. That created a whole industry. Because you can measure volatility, whether it's the standard deviation, or beta, much more accurately, that’s what you can means. That's what drove the whole financial engineering.

When I look at growth stocks, Stu Meyers, my old colleague from MIT, these are growth opportunities. Those are options. If you're going to value those, don't put in a discount rate. That's a very crude way. Now, you see what happen when you do it? Had you brought up that. I get your point. You're correct. I thought I said that these are the risks to look at and then put it into the model and you will get it right. I don't know what else to say. I mean, I’ll just say, I use the best tool I could find for the job I'm trying to do. I don't try to put someone who needs a size seven shoe into a size four, and make it work, when I don't have to. Maybe better, no shoe at all.

Is there an error in the transcript? Let us know! Email us at info@rationalreminder.ca.

Be sure to add the episode number for reference


Participate in our Community Discussion about this Episode:

https://community.rationalreminder.ca/t/episode-234-prof-robert-c-merton-icapm-retirement-and-models-in-finance-discussion-thread/20748

Links From Today’s Episode:

Rational Reminder on iTunes — https://itunes.apple.com/ca/podcast/the-rational-reminder-podcast/id1426530582.
Rational Reminder Website — https://rationalreminder.ca/ 

Shop Merch — https://shop.rationalreminder.ca/

Join the Community — https://community.rationalreminder.ca/

Follow us on Twitter — https://twitter.com/RationalRemind

Follow us on Instagram — @rationalreminder

Benjamin on Twitter — https://twitter.com/benjaminwfelix

Cameron on Twitter — https://twitter.com/CameronPassmore

Prof. Robert C. Merton — https://robertcmerton.com/