Nobel Laureate Robert Merton on the Financial Markets

Well good morning everybody, and welcome to Dimensional's thoughts leadership series. I'm Mark Gochnour. I'm the head of global client services here at Dimensional, and I'll be joined by Gerard O'Reilly, our co-CEO and CIO. Now just to get it on your calendar, we are going to, want you to be aware of, we are sending out thought leadership series over the next two weeks. So next Tuesday, we'll have Ken French joining us. He's a professor at Dartmouth. And the following Tuesday, May fifth, we will have Eugene Fama, noble laureate and professor at University of Chicago. So we've got fantastic couple of weeks lined up here to talk about markets. We've got a fantastic session here today as well. We're very fortunate to have professor Bob Merton join us from MIT. And you look back over the last 50 years, Bob has made a tremendous amount of contributions to the field of finance, and I want to highlight one other thing here. Here at Dimensional, he does have the title of resident scientist. So there's a lot to get into here today. So Gerard, welcome. Thank you. Good to have you back in the studio here. You've gotten to know Professor Merton very well over the last several years. Tell us a little bit about that relationship, and then tie it into his role here at Dimensional, as our resident scientist. Absolutely, and I thought it would be useful to give some remarks about Professor Martin before we begin, and I've written down some for myself, because his list of accomplishments is incredibly long. We only have so much time here. We only have so much time. He's a distinguished Professor of Finance at MIT, and University Professor Emeritus at Harvard University. And you mentioned ground breaking research over 50 years in lifecycle and retirement finance, in optimal portfolio selection, capital asset pricing theory, financial engineering and innovation, and the pricing of options. Received many awards, and you alluded to the Nobel Prize in 1997, that he got for his work on options pricing. Formerly associated with Dimensional since 2003, and I've known Bob for quite a few years now. And one thing that I've always found is having conversations with Bob is that he thinks and approaches economic problems, or conversations about markets, in a way unlike anybody else that I've ever met. And every conversation I get some really unique insights, pearls of wisdom, that really, really shape how I think about a particular topic in economics or finance. Today, Bob is going to be talking about markets and so on, and sharing some of his insights about market prices and option markets. But with that, let's bring Bob up here on screen, and welcome Professor Merton. Very pleased to have you with us. Well it's good to be with you today, especially on Dimensional Week, to be with everyone. So I'm very privileged to be here today. Well very pleased to have you, Bob. And we thought that we might start off with a few questions on the background of how you began working with Dimensional. How did that whole situation unfold? Well as you alluded to in your remarks, I've been at this for over a half century, and I've known Myron Scholes and Eugene Fama, and Mac McQuown for just that long. I've always known at Dimensional from the time it was founded in 1981. I first met David Booth in 1997 at the Nobel Prize celebration in Stockholm, and then over the years, 2003, as you mentioned, I was invited to serve as an independent director on the fund boards, where I was until exactly 10 years ago really. I enjoyed that very much, and of course very familiar with Dimensional in that role. But then at that time, I was working on interesting things, things of interest in retirement finance, and related things, and Dimensional, and felt that was an interesting area. So we decided that I would come and work at Dimensional on those problems, and others, which I have been doing. And of course, I no longer could be independent, so I resigned from the board, and then took on this role as resident scientist for the last decade. Very good. How has it been working with the team here at Dimensional? Oh it's been great. Very quickly, others may not recognize this, but I've had the good fortune to be associated with a number of really first rate financial institutions, but none of them were founded on science, finance science. They use finance science, but they're not founded on it. And so someone who comes from that background always feels they're swimming a little bit upstream, because of the somewhat different cultural background. Dimensional was founded on science, and for nearly 40 years, it's been using science. Fantastic. So getting into today's team, Bob, let's talk a little bit about research that you've been working on recently, and things that are really grabbing your attention right now. Well, most recently times, there's two areas that I've been working. One was goals based investing, and just briefly, that's a way that you can improve performance to the goal, without having a higher sharp ratio, without having alphas, not that you give them up, but this is an increment to it. And it can be useful for both institutions and individual performance improvement. The other is the extraction of more and better information from market prices, to use in both implementing our plans of action, and improving on the formulation of them with better information, better data. So when you talk about market prices and markets in general, can you talk about some of the things that markets do, whether it's aggregating information, transactions, and how you think about market prices more broadly. Yes. The two core functions of markets is we know, one is transactions, executing our plans, and the other is the source of information, and indeed, that's the one I'll be focused on today, principally. And they are synergistic to one another in the following way, that the more we improve markets, more markets, more different kinds of securities we can use to better execute transactions, the increase in numbers of markets and types of securities also makes the information set richer. So as each of those has happened, which has been going on for half century now with innovation, they work together, but they do really distinct functions. Can you give us an example of the latter there, Bob, in terms of information and prices, and maybe pick some examples around insurance hedging, things of that nature? Sure, yes. Today, in these times, the focus on our minds most I think is on managing risk and uncertainty, and so I'm going to work on those areas, how can market prices improve in that. But then that will feed into how we can also do the other part we're interested in, enhance expected returns. So if you'd like to get started, if it's risk that we're worried about, let me go first to a slide to remind us what are the important ways of managing risk in general. Bringing that up now Bob. We'll bring up the slide on managing risk, so Bob can, there we go. There are essentially only three ways to manage risk. We all know diversification. We get the best diversified portfolio. Let us, that's a given. But if diversification doesn't give us enough risk mitigation, then what's the next step? Hedging. That's where we substitute the risk free asset for the risky, the best risky asset. So diversification alone would give you the 45 degree line for your portfolio, because you just told me the best risky portfolio. If you shift to hedging, and if you, then what you do is you move up the axis to absolutely sure thing, but you give up, upside, for the portfolio, because you don't have 100% there. There's a third way, which is insurance, and I'd ref that by saying we buy insurance that guarantees us nothing less than that same amount, but then we have a much bigger upside, because we pay for the insurance. Now the hedging and diversification come for free, but insurance is not free, we pay for it. So all I want to illustrate to you with this, just look at those three patterns. None of them dominates the other, and all three give different patterns. My final point to make on this is that most investing still today is focused on the first two, diversification and hedging. The application of insurance is a newer, less widely used, but I believe is going to be more and more important. And why only use two tools when you have three? Now in the insurance, when it comes to stock market, buying markets asset prices, it's done in option markets. Puts and calls are insurance. Puts guarantee you a minimum value for a period of time. So it's natural in this time to go and look at the price of insurance. If I look at the price of insurance, the price of puts and calls, they have different maturities, but the factor that has the biggest influence, everything else being the same, is how much uncertainty is on the thing that's being insured. How volatile are stocks? So as this chart shows you, that for say a 30 day insurance put, and it's on let's say the market, the price will depend on how volatile is perceived that the market will be during this period. If there's no volatility, then the insurance is worth nothing. But as you can see from this graph, the value of the policy, or the value of the put is higher and higher, the greater is the volatility, and therefore there's a mapping, in fact straight line mapping here, between the amount of uncertainly perceived for the underlying and the price. So if we can just map one to one, going backwards, start on the Y axis where the price is. If we take our, let's say, Black-Scholes model of pricing and set it equal to the current market price, whatever it is. Let's pick a price, let's say three. If you take three, if you go to three on the slide there, and then you move across and down, you'll see that it implies a certain volatility that's associating with that price. This is the fundamental mechanism for extracting from option prices, these assurance market prices, an implied value for the amount of uncertainty in the future. It is forward looking, not historical. And so this is how we fundamentally derive information about distributions, information about the future by looking at prices. And currently prices I trust the most, because they have all the information from the past, but they also have the most up to date information today, looking forward. And in these times particularly of crisis and uncertainty, the future may be a bit different than we've experienced in the past, and these prices are giving us information based on both the past, but also looking forward to the future. So then to recap Bob, for a moment, if you will. So we're thinking about these options prices, and options trade in the market place all day long, and we're going to see how much in a moment. And from those prices we can infer, here's the expected uncertain here, the amount of uncertainty that the market is anticipating. So we use a real market price to get a sense of the amount of uncertainty that the market is anticipating. That of course has become known as, one level of that, as the VIX. So people talk about those levels of the VIX in terms of level of uncertainty. Can you comment a little bit on the VIX? And what that means? And how people should think about it? Yes, well the VIX is a particular version of this implied volatility process, which basically looks at option prices traded on the S&P 500 on the CPOE. Through the process I just showed you, uses those prices to infer or imply a volatility, and that number is reported as the volatility index for the market. And as prices move, stock prices, everything moves, this implied volatility moves back and forth, and that's gives us information about the price of insurance. It's a standardized price. We can compare it across time and across the space. And of course one critical question is, when you're extracting information from market prices is you want to be sure both the market is working, that the market is deep enough, and used enough, and has enough volume, and so forth, so that the prices there are representative. Clearly when there's very little marketing, it won't be there. So we showed you on the slides just in front of you, the amount of volume that you see during this recent period on trading of options. By day, I call attention to the very long bar there. On the 28th of February, the volume of trading in these S&P option contracts, VIX contracts exceeded coverage of over a trillion dollars of exposure to S&P. That's just the trading volume covered in one day. So bottom line, these are very deep markets. The information, therefore, should be, gets extracted from it, we should have confidence and trust in. So couple of observations there, Bob. Thanks for that. That if we had a trillion dollars on that particular day, the total market cap of the stocks, and the S&P 500 is about 24 trillion, so it's about four percent of the value of the shares outstanding in the S&P, so a big number. You mentioned a couple of things though, Bob, that I think will be great to dig into a little deeper. So you said the volume is deep, and there's a lot of activity, and that leads to you having higher trust or confidence in the prices that you see from markets. Can you talk a little bit about what you mean by trust in market prices, and whether they're biased, unbiased, or what type of information, those types of things. Quickly, trust has two components, competency and trustworthiness. Competency means deepness of information and knowledge. The market contains enormous amounts of information, and aggregates it very quickly. Information that no one in the world has themselves. The market, very informed traders, central banks, sovereign wealth funds, they may have very good, special information about some part of what's going on, but they don't see the whole picture. The market, through its mechanism collection, aggregates all the information out there. So what you're doing in looking at market prices is you're getting information that no one themselves has, and therefore it's a primal source, which we all use, and so it's very rich and important. Second, it's certainly unconflicted. There's no bias in the market to give us distorted numbers from this good information. It's not always right, but it's important to say it's unconflicted, so therefore it's trustworthy. So that's why I trust the markets. That's fantastic, and we share that sentiment here at Dimensional, as you well know Bob, and part of why we get along so well together, I'm sure. Now we've talked about the S&P 500 options, and the levels of the VIX. What have we seen recently from the VIX, Bob? And what can we infer from that? Okay, well first if we go, I wanted to just show you that this is a time series of the VIX, the price is insurance from the 1990s. If you look at that, you can see that it varies quite, and there's some big, extreme peaks. In 2008, of course, Lehman et cetera. And then most recently, in the piece of the slide to the right, we have a blow up of what's happened in the last month, and on March 16th, we had a record level of volatility, greater than 80. To put it in context for you, you don't have to do it, but back in the pricing graph I showed you, if the volatility were 10%, which it's been within the last year, that contract would sell for about one, one and a half percent of the value of the stock covered. At 80%, it's probably gonna sell at about nine dollars or nine percent of the value. So that's nine fold difference in price for the same insurance contract. So that gives you some idea of how volatility in the changing of the price of insurance occurs. So this is a rich set of information, and a detailed one, quite relevant to these times, as you can see. Hey Bob, I was just gonna make a question here for you too. You talk about the record high, can you just clarify, that's closing prices, right? That's not intraday high, where we've seen it in some cases get to high 80s. Absolutely. Those are closing prices, but as you allude to, that day, I believe that at one point it was trading at close to 90, maybe the very high 80s. So no, these are only closing prices. They, at times, have been higher. But the key message here is this is a market that conveys information and knowledge that's changing a lot, and quite important. So Bob, we're seeing here a, what people often refer to as 30 days, volatility over the next 30 days. Are there others types of options out there that might give you information about expected volatility over a longer horizon? Or said differently, the cost of insurance over longer horizons. Absolutely, there are many contracts traded on in the exchanges which are longer than 30 days. In fact, the CBOE on these contracts on the S&P, they go out, well they're every month, pretty much, for this year, and then the following year they go out contracts up to two years. So as a result, you not only have an implied volatility for the next 30 days now, but you now have from longer contracts, just like term structure of interest rates, you have information about what volatility or the price of insurance will be in the future, for 30 days. So I've drawn here for you, as you can see, a term structure which plots against exploration, how far the future of these contracts go, and the implied volatility. You see a pattern. This happens to be a downward sloping term structure, very much in the same lingo as the term structure we talk about of interest rates. So there's a lot more information, because it's telling you something about not only the current but the future. Now if you really wanna compare and see how to derive, if you like, ideas or inferences about the future, you really wanna be comparing apples and apples. So you like to look at prices for the same thing. So we use that term structure to derive forward prices for 30 days volatility. So we have the spot price today, which is insurance for the next 30 days. We have a forward price, and here's a graph of this that I plotted for you, for the forward prices for 10 days ago, in which we have the implied volatility for immediately, and then you have the 30 day implied forward price, or forward volatility a month from now, two months from now, all the way out there in this particular picture to the end of the year. Now what you can just see from looking at that, on today, the price of insurance for 30 days is 41. The implied price for the future, you could lock that in today even, is considerably less, about 31% less. So this is saying that the market seems to be suggesting to us, in fact offering us the opportunity to buy the same insurance contract for the same period, but buy it in the future, at a different time for a much lower price. So it's giving us that information. So if I might summarize a little bit Bob to be sure that I understand. Basically what we're seeing here is market prices, on what people can agree on today, to ensure a particular investment for 30 days right now, or 30 days a month from now, or 30 days two months from now, and so on and so forth. And what we're seeing is that people, our markets are demanding a smaller value, or smaller amount, to ensure for 30 days out in the future, than they're demanding today. Is that a fair summary? Yes. Does that change over time, Bob? Is it always downward sloping? Yes, sometimes it's upward sloping. And I put together in this next slide is a kind of dynamic of looking at this same forward curve, for periods surrounding our current situation. The top left one is from the end of the year, December, and you can see it's upward sloping. The minimum value's about 14, so you see that's quite a bit lower than 40 or 80 that we're seeing. And then it's somewhat upward sloping, and has a peak out about a year at 20, and that's showing us a certain pattern that will imply that the volatilities were expected to probably increase over this time. But then if you look at the end of January, that's the second one to the top right, you see that the pattern is, it's the back end, it stayed pretty much the same. But now what you're paying for 30 days as of January, the end of January, is back up very high, it's almost flat. So it's going from upward sloping to flat, and then we go to the end of February, and now we see two things. The price of insurance is basically doubled. The implied volatility is considerably higher, and this curve is sloping, and it's a considerable slope. The way you can see that is, as you see on February, you were paying 40 price, and the price you could lock in that same contract in six months from now is under 20, half the price. So you get a very different pattern with a very different inference. And finally I take the end of the March. And end of March, we see that both the price of insurance in general has risen, it's even more uncertain at the beginning of March looking forward. There are more uncertainty we resolve, but it has a similar pattern in the sense of downward sloping. And so we can use these to see the information is changing, and then we can use this information to help us infer what the market is saying about the future distribution, and the resolution of uncertainty. And I guess if I'm reading this right, I'll ask the two of you a question here, as I look at some of these different months. When you get into the end of March, so then that cost of insurance, looking out at least as of November, effectively had a 40% increase for the cost of insurance longer time, relative to even a month before, based on those November dates. So it's a significant increase in the cost of insurance, even longer a term. Is that a fair statement? Absolutely. By the way, that's the point. In times like these, where you have enormous amounts of information volume increase, much more information per unit time is going on right now, I needn't hardly tell you why. And secondly, the types of information, or the views of information that are in that bigger volume are much more dispersed. So you're going to get, as those information arrives, you're going to get lots of changes in prices, lots of changes in the insurance prices, and they may seem well those are crazy. But no, what they are telling us is that the information is very important. It's changing, and there's a lot of non-consistent, or dispersed information, and the market is doing its job to aggregate it, and then feed us that information. And if you're writing an insurance policy, it's telling you that you don't charge what you normally charge in a different period. So that's a really important observation, I think Bob, in terms of the comment that we've had a lot of market volatility. Prices have changed very rapidly in calendar time, short period of time, but that's exactly what we expect from markets and market prices in this type of market environment, given all the news that we're having about the coronavirus, and the price of oil, and the potential impact they may have on the real economy. So what we're seeing with respect to the VIX is what you would expect, conditional on these types of news events coming to pass. Would you comment on that statement? Yes, the whole point is that markets reflect the reality of the situation, and that's why it's more important to get this trusted source of information in crisis times. It allows us to have more information, aggregate it in an efficient way, to help us form strategies and responses. It gives us some degree of control. Obviously not control over the events that are happening, but control over our understanding of them, and our ability to make sensible, wise decisions in these very crisis laden times, with lots and lots of uncertainty. Very good. Can we compare this then to a previous time period, Bob, and try to get a sense of how it compared in 2008 time frame, these various curves? Yes, in fact, if we go to the next slide, I've put together, I put a parallel slide together to the one I showed you for right now, showing four periods during the last big crisis, 2008. August 29, just before Lehman, is the upper left, and you see a pattern very much like we have December. Upward sloping, numbers higher than our December, but very similar. The minimum was a little over, it was 20 in itself, and the highest was 23, so upward. Secondly though, you see in November, now we've had Lehman, AIG, et cetera, and we're in the middle of this crisis. We see that the price for insurance is quadrupled. It's gone from 20 to 80, and the curve, giving us inferences about information that uncertainty of resolution, gives us much lower prices as we go out. The deal being that during the middle of the crisis, lots of information, lots of uncertainty. Will it work? Will it not? Will we get a handle on it? We will know a lot more about it in two months, three months, or four months, and therefore the information, good or bad, will be resolved, and therefore uncertainty or risk will be reduced, as far as we can forecast, and that's how I would interpret one element of why you have a downward sloping curve. Another possibility, of course, which is open to testing, and we are, is that it also may be that the risk premiums are much higher in these prices, concentration periods, because of that. And so it could be that these are somewhat bias by rich premiums, but that's an empirical thing that we're studying. So can I pause you there for one second, Bob, before we get back to March and December. You say the risk premiums might be higher in these crisis situations. Saying that another way, is it expected returns might be higher in these crisis situations? Is that an inference to draw from that? Absolutely. It's just not reasonable for me to believe that the risk premium stays constant when the amount of risk you have to take to do it quadruples. I'm not saying it's going to move exactly linearly or anything, that's not the point. But we would expect that the markets here, by seeing how risk is changing, and seeing how it's profiled, we can get inferences about risk premium, and even maybe the time pattern of risk premium. And it may even give us insights as to when we'll get higher expected returns than in one configuration than in another, not because the market's foolish, but because the pattern of risk is changing in a way that it demands different risk premiums to clear markets at different points in time. Very good. Back to the slide, Bob, if you have more remarks to make there. Well I'd like to move on, but just let me point out two things. If you look at the slide that's on the end of March, and you compare it, of course it's nothing like November of 2008, and then by December of 2009. March, you may remember, was the bottom of the stock market. But by December, the curve had gotten back to kind of normal, upward slope is not too steep, et cetera. Now this is not to say that as the uncertainty goes down, that means things are good, it just means it's resolved. It happened in the case of 2008, thank goodness. It was resolved, and the market rebounded. But that's not what this is forecasting. It's forecasting the resolution of uncertainty, and the amount of uncertainty or risk that one has to take in forward months for the future. So it's the time pattern of the risk we're exposed to. We're becoming more detailed knowledge extraction from this, looking at the pieces. Very good. Do we have options in other markets, Bob? Yes. We've both done the VIX because everybody's so aware of that, but we have other equity markets. For example, the European equities. There's a European VIX. I use VIX in a generic sense, not as a commercial name. And as you can see here, the pattern is very similar, lots of volatility spikes, including spikes right on the same day that we had the spike here in US. Perhaps not surprising, but often we get surprised when something different happens in these patterns than we might have anticipated. So the fact that you say, well they're the same. They are different, and at times, during the European crisis, for example, you'll get a differentiated sense of the risk resolution and scale, than you would have, let's say, in the US market. But we have a variety, but we also have them in other asset classes. We have insurance or options on currencies. We have options on commodities, including oil, and the news yesterday on prices, again. And we have them on bonds. We have them on interest rates. So I have a slide here, the 10 year, which is a pretty important thing for us here. We have the implied volatility history on the 10 year, on your left, and we have a picture of it on that same March day. Now as you can see, and that's why I put the yellow numbers in there, max and min, that although the slope looks very big, the absolute amount difference is rather small in basis point, as we would expect with treasuries. And what's this telling us? It's telling us something about the price risk we're taking in 10 year treasuries over the various periods. It was high at that time, but you could buy the insurance, price risk on the 10 year due to interest rate changes, much less expensively in the future, because the market's perceived that this is a time when there's much more information, and much more, therefore, risk impinging on the prices. So we cover many, many markets. Quick question for you Bob, make sure I'm following along. At that implied volatility of around eight, and based on some of the previous numbers that we had, does that mean it costs about 75 basis points to one percent of the value of your treasuries, to insure them against increases in interest rates, or falls in prices, if you're doing it based on today's price. Is that a way to think about that number? Yes, that's about right. It probably, at the vol of eight, it's probably about one percent or a little less. Okay, very good. I was gonna jump in here. A couple questions for you, I think one for each of you. Bob, I'm going to start with you. There's a question here, I wanna make sure we've got the right understanding of your comment. And it was back on the VIX, when we were looking at some of the future curves there, the term structure of the futures for the VIX. I think the comment was something in effect of we see less certainty, I'm sorry, less uncertainty, but that doesn't necessarily mean it's a good thing. So clarify what I was saying there, and I just wanna make sure that idea of doesn't mean it's good. Were you talking about doesn't necessarily mean market returns will be positive as uncertain declines. That's correct. It just says, the thing that's a big risk in our mind is there, the stuff that's going to have to be resolved, have big, wide, different impacts for us, and therefore they're quite important and uncertain. But when that's resolved, I mean by resolved, if you want to do it in terms of stock prices, the stock market is resolved, the information turns out to be better than we thought and prices go up, or worse than we thought, they go down. Whichever that outcome is, we've resolved it, and therefore looking forward from that time, that's not a source of risk. It's a source of pain or pleasure, but not risk, and that's what it's telling us. Okay, great, thanks for clarifying that one. Then Gerard, this question's for you, and we've had some questions come in just saying, what do we do with all this information here at Dimensional? How do you take some of this, from these different sources, and then apply them in the way we're managing our portfolios here? Yeah, I think that Bob alluded to some ongoing research that we may touch on at the end of the webinar. But there are some things that we do in the here and now that I think are very apropos, given Bob's comments about using information from market prices. And I thought I might share a few different examples to put it in the context of how we manage the money here at Dimensional. So one example that lots of folks are familiar with about what we do at Dimensional is use information in yield curves. And so a yield curve basically reflects the market price of a bond, or a set of bonds plotted out, and what we can infer from that is differences in expected returns across different duration bonds. We use that to increase the expected returns of portfolios. Bob talked about insurance, well what about insurance for corporate debt? They're general called credit default swaps. Well we can use that information to think about what's the implied credit rating, with respect to a bond, and we can use that as an input into our enhanced credit rating scoring system. Let's talk about stock prices and corporate events. If there's a merger and acquisition type of an activity, we can use the information in stock prices about the probability that that will actually happen. If the stock price is much below the offer price, it's maybe not going to happen. Much above, may be an indication that there's better offers out there in the market waiting to come. So there's a lot of different types of information that we can use there. Cross pollination of markets, I think, is another great one which is in the information in the securities lending market, and that's about stock prices. Let's imagine I have a stock A and a stock B, and somebody's willing to pay me a lot to borrow stock A, that has information about the price change of stock A relative to price B, or stock B, stock A is going to under perform stock B on expectation. One other example that I think is a pretty cool one is using equity prices in bond markets. If we look at two issuers, two firms, A and B, and they both have stock, and they both have bonds. If stock A goes down by a lot, it gives you information over the past five, 20 days. Information about what the price of bond A might be with respect to bond B, so information in that cross section. So I think there's a lot of ways that we use information in market prices, and it's a really exciting area of research that we might get into a little bit later on, about how do you keep on pushing that forward, and think about how do you use that information? What does it mean? And how can it be used, to Bob's point, to make better decisions? Because information right now is coming fast, it's coming often, and markets are one of your go to sources to understand how that information is being processed, and what it might mean for risk and expected return. Okay, well I like how you talk about this idea of research, and a lot of us will default to, well, maybe that's differences as we think about stock returns, or bond returns, but there's so much out there, research, in terms of better understanding prices from all sources of markets. So we'll come back to that at the end. The other thing I just want to allude to, there's been a lot of questions, not surprisingly, about oil, and how we might use information around that, say from yesterday's event. But I wanna keep going here. We'll come back to that, when we get into the Q&A session here. So let's go back to the slides that we're looking at there, and keep this idea of reacting to new information that comes out. Bob, I'll turn back over to you. Thank you. Well as you asked me about are there other markets, I pointed out that equity markets in different locations. We saw the fixed income currencies, including oil, commodities. But also I wanna go back to the stock market, and the CBOE, the VIX, and say there's even more information there than just the time pattern, because they have contracts with different exercise prices, but the same expiration date. And those different exercise prices mean these are insurance contracts that may have a deductible, if you like, you have on your car, where you take the first amount of loss, any limited loss, and then you're covered below it, and they sell at different prices. Now in the standard option pricing, simple option pricing models, if you were to look at the implied volatility for options on the same security, with the same expiration date, okay. So same underlying risk, same expiration date, you would expect that the price of insurance would be the same, the implied volatility would be the same, a flat line. But when you look at the actual data here, and this is displayed here, different strike prices. The ratio is the strike price to the current price gives you whether it's in or out of the money. You'll see that it's not flat, but actually, slope, actually sometimes it has a smile look, but this one looks more like a smirk. But the main point is that it's not a straight line. So what is the market telling us? It's telling us something, and the thing is that the Black-Scholes model that we were using there to get information, we can get even more information from this seemingly inconsistent, but really signal that the market's telling us something. If we introduce a somewhat more advanced model that includes both the kinds of volatility in the standard model, but also this from a possibility of event risk. That's a situation that instead of information coming continuously in small bits, very fast, very frequently, you get, at one point in time, a major announcement of information, and that causes the price to jump immediately. And that so called jump or gap risk, event risk, is very much, if you wanna think about it, is you're going along, the phone rings, and the phone ring is saying "There's gonna be an event right now. "We're going to tell you a lot", and then the market's told what it is, and the price instantly moves, it's a jump. That's a different kind of risk. When you put that into a more complete model, still incomplete, but a complete model, you're able to infer even more information from the market about what's going on with risk. You're not only getting the variation, volatility that we describe, but you're also able to extract estimates for the market of what is the likely frequency of arrival time of these big events. You've heard this happens every 10 years or every 100 years, that's telling you how frequently you think it's going to arrive. Well you sometimes find in these data that things are supposed to happen once every 10 years, when you can see the implied likelihood that it'll happen in particular periods, it can be, the expected time could be within the next two months, until the next arrival. And so it allows us to more refine even better our implication, our understanding of the distribution of future prices that we can extract from the market. So Bob, I have, to make sure that I'm following along, and then I have a question for you based on that. If we took this situation here, as we've seen in the graph, and put it down into a stock that's trading at a dollar today, and we want to buy insurance on that stock, one part of insurance will allow us to sell that stock for a dollar in a week from now. The other insurance will allow us to sell that stock at 90 cents a week from now. So we look at those two prices, and we say well, it's one stock, it's the same time period, so we think that the uncertainty that would be derived from those prices will be the same. But in a simple model, the uncertainty is different. When you then look at the market, you say it's not the market that's broken, it's my model that needs to be enhanced. If I use an enhanced model, I find out that I have a different distribution of returns. Things that are maybe less probable in normal times are more probable right now, so my enhanced model gives me information. So that's the way I think about what you've said. That leads me to a question, which I've heard you say many times before, which is the art of science, and the knowledge of what model to use and when to use it. Can you elaborate on that a little bit? And just describe how you think about that, and what you mean by that? Yes, we have fabulous models, improvement models over the last, mathematical models, data scraping, AI, the whole thing, but they're all models, and models are incomplete, and if they're incomplete, they leave things out of the model, and they put things in the model, and all models can fail because of that. The art of the science is, and this is true of any science, physics, life sciences as well as finance and economics, the art of the science is choosing your abstractions. When you take a model, as we're using here, either to interpret information, or to process information and make decisions, the model depends on what you assume. What do you take account of in the model? And what do you say "I don't need that "to lead me to a good decision "on this particular thing"? There's no such thing as an absolute best model. It's the best model for the application, and for the user. And so when you realize it that way, you sense that even in the hardest of sciences, and the ones with the most data, that always there's a form of judgment. As far as I know, we don't have any AI, any way to substitute for that. So experience, knowledge, and this art of choosing a good model for the situation, and frankly, that's the biggest difference between a model that works well in a situation, one that doesn't. So it's critical, but never forget, the mathematical models, or even the models in our heads which are abstractions, are precise in the model, but they're not precise in the reality that we are applying the model to understand, and never, never forget that. Always the art of the science is critical, and therefore judgment, and who's chosen the model, what data are used in the model, all of these things is necessary in order to have trust in what you're getting out of that model, is giving you the appropriate indication, albeit with risk or error of how they address the situation they're in. Fantastic, really appreciate that, Bob. So coming to the end of the time, I think you might have a couple of interesting examples of looking at market prices around some unique events that we've seen historically. Would you mind sharing some of those examples with us, Bob? I picked these out. These are both from 2008 during the crisis. The first one involves the interest rate market, the libor market, the financing market, and what we're doing here is there are things called caps and floors in that market, and there's so many of them in a sense, with all these different exercise prices. There are many, many, many of them, for many, many dates. And when you have that many securities, you could actually not only get an implied volatility, or an implied chance of a jump, you can get an entire implied distribution, and that's what you see on this screen. This is work of Doug Breeden and Bobby Litzenberger. But what they did is they looked at, during the crisis, they looked at what the market, these options were implying about the probability of the three month libor rate three years from now. So way in the future, and they looked at it in June, that's before the crash. That's the pink line, so you don't need to look numbers, just look at the colors. The pink ones, that's a histogram implied back in June of 2008. Then we did the same exercise in December of 2008, after we're full blown in the crisis, Bernanke and the Fed is taking a very aggressive driving markets. Now what do you see here? You see that pink is more centered and much higher than the blue, which makes sense, in the sense that you've really brought rates down, and the indication would be there. So this is an example of showing you how much the distribution, not just in fixed points, but in actual shape changes. But that's telling you something, you'd say well of course, that's what we expect. But they don't wanna show you the next slide. This is three months later. At that time, on March of 2009, the Fed announced, not only did they have rated very low, but by the way, they're going to keep them low for an extended period of time. They said "We're going to keep them low." Now let's look at how the market reacted in terms of the implied distribution. Pink here is now December, that's the one I showed you in December. The blue is for April, so after March 29th, and here's what you see. The pink is December, the blue is April, what can you see from the picture? The whole thing is shifted to the right. That means that the price is implied, not of interest rates, forward interest rates that people pay is going to be higher, not lower, even though the announcement would have suggested, as the previous one did, that rates are going to stay lower for longer. And you also see the shape of the distribution. You have a much more wider shape. You have a different, I'm not going to take you through the detail. Bottom line takeaway, sometimes the markets confirm, so these full distributions that we're able to drive, what common sense said you would be, although it quantifies it and that's useful, because it tells you more than direction. But sometimes, as in April of 2009, it shows you things that are very counter to what makes any sense, based on the announcement. I think that's really interesting information. That would be information to the Fed, in terms of understanding the response of the markets to its policy, and certainly interesting to investors and the libor market, or use the libor market for financing. So once again, when you have enough securities, you can get an enormous amount of information, in this case, the entire distribution. So that's fantastic, Bob, and from our perspective, our clients perspective, we always talk about yield curves and bond prices reflecting the expectations of investors, and that may contain information about what yields and spot prices might be in the future. But what's leftover with all that expectation built in is differences in the premiums that people demand to hold different duration bonds. So not only can this be useful to the Fed, this is useful to the market, and is reflected in the prices that people are willing to pay in the market, and the amount of expected return that they demand for holding different types of bonds at different types of time periods. Very useful information for both decision makers and people who are investing. All right, thanks for that, gentlemen. We've got a couple minute left here for some questions. I'm going to start with you, Bob. I've got a question for you, and there's been some themes to these questions that have come in here, when you talk about risk management. I've been asked a question, when you talk about risk, how should we be thinking about risk? What is risk? How is it tied to uncertainty? And then what are these tools we can use, whether it's diversification, hedging, insurance, depending on what risk is. Well, two parts to my answer. The first part is risk and uncertainty, this is a long discussion, but it's well known, the concept that risk is thought of as you know the rules of the game. You know the distribution, but you don't know what the outcomes are gonna be. So if you enter into that asset, there's risk in that you may get something different than expected. That's the usual measure of risk, and it's quantified, and you can talk about it that you know the distribution in some way, or believe you do, so that it's quantifiable and that's usually what we measure when we talk about uncertain risk and so forth. However, as we certainly are aware right now, there are a lot of things out there that we're very uncertain about, and we don't even have the model for them. We don't have the data. We don't even, we haven't accumulated the data, or we haven't acquired the model or the knowledges yet that allows us to perform that question of what distribution it has, and so forth. And that sort of thing is what we think of uncertainty. And in crisis times, it is typical that the amount of uncertainty goes up a great deal more, as does risk. And part of the risk of the insurance is surely going to be related to the uncertainties, which make it very, very hard to measure. Both are covered in these insurance contracts. They're giving you, from the market, are giving you assessments of the price for this, the amount of risk, if you think of higher price for insurance of the same contract, means there's more risk. So that's how I would look at those things. What I would then, on the second part of your question say, that this more detailed information that comes from more and more markets being created and are extracting them, allows us to both manage risk better, but also identify and higher expect of return potentials. Why? Among other things, as we all know, most models in finance are estimating expected returns, like the CAPM Asset Pricing Model, and so forth, use risk measures to determine the differences in risk premium. So whether it's beta or multiple betas, or Intertemporal Capital Asset Pricing Model, or whatever, the betas, they're risk numbers. They're variances and covariances. Those are easier to measure, that's one reason, but those are definitely risk. Most everyone I think would agree that if markets to clear, that if risk is higher, risk that you can't get rid of, then equilibrium payments of expected return should be higher. The trick is how can you estimate how they're changing? Very difficult. Are we gonna make progress on that? Absolutely. The times will come when we're gonna have much better models, models that will be able to give us a more dynamic picture of how risk premiums are changing, and therefore we have better understanding, better knowledge. Excellent, well you had several references there to expected returns. I'm gonna use that, Gerard, to bring back to the question we were talking about earlier, about how does Dimensional use some of this research out there? Because we talk so much about what we try to do in our strategies, just to maximize expected returns within designing the strategies. How are we using some of the research, that's been highlighted today, on a go forward basis for that? So I'll make references to some research that we've done, and then what the research, fruit it will bear going forward, time will tell, as that research uncovers new gems in different marketplaces. But examples, I gave some examples of things like using information from the securities lending market to infer something about stock prices, or from stock prices to infer something about bond prices. Another example that's very relevant for the right here and now is the commodity futures market, and of course commodity futures, in agreement to buy or sell a commodity at some point in the future, some future date. We saw unprecedented moves in the WTI, West Texas Intermediate oil market yesterday where those near term futures, so if you were agreeing to deliver oil in May, then you had to pay to deliver that oil, they went negative. So that's an unprecedented move. I'm sure the reasons of that will be discussed and debated over time, and people will grow to understand it better. The research that we've done on that particular topic suggested if you look at the futures prices. So just like Bob showed different horizons for the VIX, different expiration parts for options, you have different expirations for commodity futures as well, different points in the future. And if you look at those curves, you get to infer something about differences in expected returns across different commodity prices. What those curves were telling us in the past few weeks is that the longer term oil future commodity contracts had higher expected returns in the shorter term. And so for strategies that use that information, most of their allocation to oil futures would have been around November to January 21, so November 2020 to January 21 type expiration dates, rather than the May expiration dates. So this is an example of research in the here and now. There's a lot going on between Saudi Arabia, between President Trump's remarks, between Russia. That's being incorporated into the oil futures prices, and then managers who use that information can position the portfolio to improve expected returns. When Bob talks about that research that we think about pushing things forward, and Bob alluded to this very well. As new markets come into being, new securities come about, and those markets increase in depth and breadth, that's a really fruitful area of research where you could say, what information can I glean from these new market prices? How to manage risk, or how to improve expected returns in the portfolio. Bob has given some great examples on how to manage risk, and hopefully I've illustrated some examples on how to improve expected returns on portfolios. All right, great. Gentlemen, we've got about a minute left here. If you don't mind, I'm going to hijack that minute, because I have trivia for the two of you. Okay Bob, you ready for this? I believe, was it your masters you got at Caltech? Yes. Okay. Gerard, you got your PhD at Caltech. That's right. So are you ready for some Caltech trivia? Sure. Okay. Okay. What was the original name of Caltech? Any idea? I'm gonna guess something to do with Throop, but I don't know. Maybe you know, Gerard. Dang! I think Bob's right. Throop University. I thought I might get you on that one. Throop University, founded by Amos Gager Throop in 1891. So very well done on that one. And this is for the audience, I think, I'll position it to them. Did Einstein ever teach at Caltech? How is the audience going to respond to you, Mark? They're thinking. Oh okay. You know the answer, I'm sure. Bob, did you know he taught there? I can guess, but I really don't know. He was a visiting professor for three winters, 31, 32, and 33. And then the last question I have, and Bob, I'm going to go for you on this one, since you got the first one right. What is something Caltech is most known for outside of anything academic, Nobel, science related? I'll give you a clue, in the sporting world. I'm sorry the internet glitched me, I couldn't hear you. Could you repeat it quickly? What is Caltech most known for in the world of sports? I would say clever ways of putting motorcycles and other things in peoples rooms. Just remarkable cleverness in doing what seems to be impossible to do. Impossible to do, kind of ties to this. Gerard, do you know? I think they have the longest losing streak of any team anywhere in basketball, potentially. It's the longest conference losing streak. Conference losing streak. 510 losses, 26 years, and they finally fixed that in 2011. What's the Caltech motto? Motto? For you, yeah. The Caltech motto is something, I've heard you say once about, the truth will set you free. You got it. Is that it? That's it. That was pressure. Okay. All right, well listen, thank you very much for joining us here gentlemen. Thank you Bob. We very much appreciate your time, and I think Gerard, you opened it up with "You always learn something", kind of an ah-ha takeaway talking to Bob. I think there's a lot of those in this one here today. I learned about smiles and-- Smirks. Smirks as well. So Bob, really appreciate your time. Thank you so much. Okay, could I just say one quick thing? Absolutely. Yes, I hope out of this is a takeaway that the markets are wise, and we need to find a way to extract them, and extract them, and use the extraction wisely. And these are very tough, scary times for all of us. We don't need to be reminded of that. I hope, however, from this flow that we have a sense that in at least our areas, very important areas of finance, we're making progress. We have been, and we will continue, and we're going to get better and better information. More markets are going to allow us to do better and better jobs of managing and controlling peoples risks, and I think the most important one is a sense of trust. Trust, not because there's somebody who knows the future, that doesn't exist, but rather trust that those who are involved in the markets, it's all worth creating information, are competent and trustworthy, and that there's some remarkable things we're doing. And so to have the confidence that the people who are involved, give you a sense of control in an uncontrolled situation, that which you know. Well thanks Bob, and I love the energy you're just bringing every day, the innovation on a go forward basis as well. So we look forward to all those future innovations and contributions as well. And Gerard, thank you for your time. Thank you. And for that today. Thanks to Bob, very much appreciated. Absolutely. And thanks to all of you for joining us here this morning. We appreciate your time. There was some questions here if we will make this available from a recording, and we will. There's a lot to digest there, and something to listen through for a few more times there just to get the full depth of it. So thanks everybody here. Be on your calender, looking at the next two weeks with Professor Ken French, and Professor Eugene Fama on the next two Tuesdays. And everybody have a fantastic rest of the week.