Not Another Investment Podcast
Understand investing beyond the headlines with Edward Finley, sometime Professor of Finance at the University of Virginia and veteran Wall Street investor.
Not Another Investment Podcast
Demystifying Market Efficiency: A Journey Through Finance Theory (S1 E7)
Unlock the mysteries of the capital markets as we navigate the Efficient Markets Hypothesis with compelling insights from from the financial literature. This episode is your golden ticket to understanding why asset prices aren't just a roll of the dice. We're dissecting the bedrock of market behavior, from the long-term accuracy of pricing information to the inevitable influence of uncertainty on capital allocation. Join me, Edward Finley, as we scrutinize the traditional beliefs in market efficiency and invite skepticism into the conversation, revealing how the unpredictability of new information and the human element behind market movements challenge established financial doctrines.
Prepare to have your perspective on finance revolutionized as we contrast the trajectories of Amazon and IBM, demystifying the market's heartbeat through the lens of human behavior. The episode sheds light on how even the most rational investors fall prey to cognitive biases, weaving in personal anecdotes and empirical evidence that underscore the powerful sway these biases hold over financial decisions. Our engaging dialogue will be supported by academic studies that peel back the layers of overconfidence, conservatism, and representativeness, offering a fresh narrative understanding that goes beyond the numbers.
We wrap up with a potent discussion on the premiums and risks of diverse asset classes, diving into the behavioral finance theories that shape investor behavior in profound and often irrational ways. Unravel the enigma behind the higher historical premiums for small cap and value stocks, and explore the influence of loss aversion on equity risk premiums. As we pave the way for future conversations on the limitations of probability theory in market risk and return, you're left with profound insights into the complexity of financial markets that will keep you ahead of the curve.
Notes - https://1drv.ms/p/s!AqjfuX3WVgp8uGqU5kV1Yj937her?e=JXxNWQ
Thanks for listening! Please be sure to review the podcast or send your comments to me by email at info@not-another-investment-podcast.com. And tell your friends!
Hi, I'm Edward Finley, a Sum Time Professor at the University of Virginia and a Veteran Wall Street Investor, and you're listening to Not Another Investment Podcast. Here we explore topics and markets and investing that every educated person should understand to be a good citizen. Welcome to the podcast. I'm Edward Finley.
Speaker 1:This week we're going to spend a little bit of time talking about some basic concepts in finance theory. Why finance theory? Well, so far we've identified what the capital markets are, the role we think it plays, the roles of participants in capital markets. But one question we have steadfastly avoided so far is what do we think makes capital markets better at allocating capital to risky uses? And we're going to try to answer that question. To do it, we need to develop some understanding of finance theory and we also need to understand some of the limitations of the answers that finance theory gives us. So let's start with the most fundamental concept in finance theory to help us answer this question, and that is the efficient markets hypothesis. Like always, I'll try to define a term for us, we'll break it down, we'll test the limits of that definition and then we'll move on to the next concept. So let's define efficient markets hypothesis In a nutshell.
Speaker 1:We're going to call the efficient markets hypothesis tells us that accurate information over the long term is incorporated into price in a market system. Again, accurate information over the long term is incorporated into price in a market system. Think about it from a very basic perspective. If there were a way to predict perfectly what an asset's future price is, we wouldn't need a capital market. Traders would buy the asset predicted to increase in price and sell the asset predicted to fall in price, but holders of assets predicted to rise in price wouldn't sell for anything less than the predicted higher price and buyers of an asset predicted to fall in price wouldn't be willing to pay more than the predicted lower price. The result is that prices would immediately adjust without the need for trading and without the need for markets all incorporating that accurate information in price. But of course that's not the real world. In the real world, uncertainty plays a big role in how the efficient markets hypothesis works and why capital markets, we believe, do an excellent job of helping us predict that uncertainty.
Speaker 1:As new information is introduced, there's going to be uncertainty as to what part of that information is accurate and what part of that information is noise. Market participants are all engaged in trying to predict the accurate information and earn a profit on it before that information is fully incorporated into price. They make these decisions under uncertainty and, as we'll discuss later, when we say uncertainty we don't mean difficulty in forecasting from among known outcomes. That's sometimes thought of as statistical uncertainty. For example, if I have a coin and I can toss the coin, it will be heads or it will be tails. There is uncertainty whether the next coin toss will be heads or whether it will be tails.
Speaker 1:When I talk about uncertainty in markets and whether information is accurate, that's not the kind of uncertainty I'm talking about. I'm talking about the kind of uncertainty where the outcomes themselves are not knowable and therefore statistical methods are of limited utility. Here's how the famous economist John Keynes explained the kind of uncertainty that exists in financial markets. He said, quote the sense in which I use the term is that in which the prospect of a European war is uncertain, or the price of copper and the rate of interest 20 years hence, or the obsolescence of a new invention, or the position of private wealth owners in the social system of 1970, which for him, by the way, was the future. About these matters, there is no scientific basis on which to form any calculable probability whatever. We simply do not know. End quote. And so this type of uncertainty means that markets play a vital role in helping to determine the nature of information and what part of it is accurate, and to incorporate that into price in a way that does not admit ready statistical inferences. And so the result of a market that is operating under uncertainty is that, in the short run, prices will move a lot all the time as actors, as traders in financial markets engage in what is known as price discovery, meaning the action of trying to incorporate accurate information into price.
Speaker 1:Now it's rather important to point out that some of you will have taken perhaps an economics class and may even have a passing knowledge of efficient markets hypothesis, and you might very well have heard from one person or another that in fact, the efficient markets hypothesis is wrong. I'm going to suggest that that is really a misapprehension of the efficient markets hypothesis. I don't think that you can reasonably disagree with its statement, so long as you're making the statement correctly. I'll come back to that definition. Accurate information over the long term is incorporated into price. I think that's unassailable. The doubt that one has is whether, in the short term, that is true or not. And that's where those who disagree about efficient markets hypothesis don't disagree about in the long run. They disagree about the meaning of that thesis, that hypothesis in the short run.
Speaker 1:We can divide those people into two camps the strong camp and the weak camp. The strong camp says in the short run prices, asset prices in capital markets are random. New information it doesn't mean it's correct, it just means there's information. New information is not predictable. It's not predictable in how it's introduced. It might be by rumor, it might be by intuition or just pure speculation. New information is not predictable in its meaning. Interpretation is very prone to ambiguity. One person's interpretation of a new piece of information, even if it's accurate, is not going to be the same as someone else's interpretation and its consequences are uncertain. Information is often incomplete and people have to fill the gaps with instinct or intuition and that means that it might not be fully understood what the impact of the information is. So those in the strong camp say traders are busy trying to disambiguate the accurate part of new information from the noise and in doing so they're doing it under a situation of uncertainty and in that uncertainty the changes in price are really going to look random until all the information can be confirmed and its import can be determined. Now here, random with quotes around it doesn't mean traders are bidding random prices. It just means that lots of traders all simultaneously trying to forecast what each of them believes the accurate price is after information is incorporated, will, in the aggregate, look random, and in that case probability theory and statistics can help us in understanding price movements and analyzing the risk in markets in the short run. All right, that's the strong camp.
Speaker 1:What are the limitations of that view of the strong camp? Well, first, I would say the import of even confirmed information is far from easy to discern. Especially in the complex system like our world, it seems to me it's rather difficult to have a piece of information. Let's take a very current thing like the introduction of AI. It seems to me it's very difficult for anyone in a complex world like ours to be able to say with any kind of certainty what the consequence of AI will be on, say, the technology sector. We might have some ideas, but a world like ours is so complex, the connections so very and very varied, means that the import of that information is nearly impossible to discern. And so it seems to me that that's a very strict limitation to understanding markets as being quote efficient in accurately incorporating information into price. Second, since that kind of complexity makes it so difficult, when the information will be incorporated into price will take a very long time, and so we're saying short term. But we might really be thinking that it might take 10, 12 years to fully incorporate that accurate information into price and all of its import. But all the while prices are still moving as people are testing the strength of different views and new information is being introduced, and so it seems to me it's not terribly clear that the strong camp of the efficient markets hypothesis is really anything other than a completely theoretical way to see the world.
Speaker 1:Take my point as follows If we have this idea that the market is going to incorporate accurate information into prices in the long run and in the short run that randomness of price movement can be interpreted statistically, it seems to me that you might look at that situation and say, okay, but before we ever know the effect of that information on price, there will be new intervening information which we will not have an understanding of, its import and what its effect will be on price, and so it seems to me hard to prove, very hard to prove that we think that it's right. Okay, that's the first big limitation. I think the second big limitation is that human behavior is usually not random, or doesn't even seem random. It's not usually subject to probabilistic rules like flipping a coin. That, by the way, is called a stationary process, something in which all the outcomes are known in advance and we can predict with some measure of reliability what the possible outcomes will be in the future. Instead, humans tend to rely on heuristics and psychological tendencies which have developed over the course of our evolution, when we make decisions under uncertainty. And doesn't that make sense, in a way that we are, of course, creatures of evolution? We have encountered uncertainty long before there were ever markets, and we've developed over millennia ways of solving problems, ways of understanding an uncertain world that do not necessarily rely on a notion of an efficient market that accurately incorporates all prices, all information, into prices, and so that would not look random and therefore statistical inference would be very difficult. ["statistical Inference"] All right.
Speaker 1:Well, for the reasons cited above, I think that the strong camp has some pretty serious shortcomings. So what's the weak camp? Well, the weak camp says in the short term, movements and prices are not entirely random, and they don't admit analysis by probabilistic means. Instead and here I'm gonna borrow a word from the title of a recent book by the former head of the Bank of England, mervyn King and his co-author instead, they reflect decisions made under radical uncertainty, which tend to rely on facts in context and the construction of a narrative to explain quote what's going on here. So Mervyn King and his co-author say that that's the much more typical way in which humans face off against uncertainty. They face off against uncertainty by making stories up, by creating narratives in their mind that help them understand what's going on here. And under those circumstances, we've developed a set of heuristics. Heuristic is a Greek word that just means rules of thumb, shortcuts. So we've evolved heuristics to help guide us in making decisions when we face off against radical uncertainty, and under the weak camp's view, since statistical probabilities aren't going to help us in the short run, as market participants are trying to figure out what the effect of any new information will be on price. Instead, what we can do is we can try to understand those predominant human behaviors and their effects on prices, and then predict market movements based on those effects.
Speaker 1:So if you're interested in having a look at the website, you'll see a couple of slides, but I'll describe them to you here in words and I think you can sort of picture it. So, on one side, I have a chart showing the price of Amazon stock, and it's the price every day for the month of October 2023. And it starts off a little below $1, and excuse me $138 a share, and then it climbs steadily over the course of October, getting to about $132 a share, and then it rapidly retreats to about $125 a share, where it stays for a week or so and declines further down to about $120 a share, and then very quickly rises back to $132 a share. And so when I used to teach this class to my students, I would ask them to interpret this chart of Amazon stock. What's going on here? Or do we simply have not? We don't have enough information to know what's going on.
Speaker 1:And, in general, what one might say when looking at this chart is that Amazon stock as investors we're trying to incorporate information into price sought to increase price steadily to about $132 a share. There seemed to then be some market resistance. Remember when we talked about how securities trade, bulls and bears are both competing in terms of their view about the direction of the price, and then the stock retreated over the coming weeks before it rapidly bounced back up to that $132. And so it's possible to describe a situation in which some market participants saw $132 as the correct price. Other market participants didn't, causing the price to fall, but ultimately that was the prevailing price. Sometimes people who do this sort of thing, called chartists, will say that the price had a ceiling, that the market was trying to break through that ceiling at $132 and it fell back from there.
Speaker 1:Next to it, I have a chart for IBM, a very different kind of technology company from Amazon, but a technology company nevertheless. Same month, october 2023, daily prices and what you see there is IBM share starting at around $139 a share, increasing to about $142 a share and then precipitously dropping to about $137, staying at that level for about a week, dropping again to $135 before rapidly recovering back to that level at about $141 a share. Similar story. We can look at this chart and we can try to make some explanation, and it's very intuitive. If I asked any of you to do it, I bet none of you would say well, I don't know. I bet most of you would look at those charts and have some sense maybe not exactly what's going on, but some sense for what's going on. Well, to illustrate the point of those with the weak view is I show a third chart, and the third chart I just label unknown. In finance jargon that's abbreviated as UNC unknown. So I'm not telling you what this is, but I am showing you daily prices for October 2023, and it starts at around $100 a share, it rises in the first day or two to $130 a share, and then it steadily declines over the next week to $60 a share, where it stays pretty much flat for a good part of the month, to decline a little bit more to about $50 and ultimately back up to 60.
Speaker 1:And here my students would attempt to engage in the same sort of thinking, that is, as Mervyn King and his co-author said, to engage in some construction of a narrative. We're trying to understand with this uncertain information. What's going on here? Well, here's the big aha. The unknown is coin tosses. Simply put, every time the coin toss was heads, the price went up by 20%, and every time it was tails, it went down by 20%. And so the point that I like to make by showing these charts is that we're saying that in the short run prices are random.
Speaker 1:We look at charts like Amazon and IBM and that kind of convinces us of maybe the weaker camp is correct, maybe it's not random, because the charts don't look random. They don't look random at all, and so we construct narratives around them. But when I show the unknown chart which also doesn't look random, I think the big surprise is there is nothing more random than a coin toss, and so we have to be very, very careful when we say, oh, it is random, oh it's not random, et cetera. What if we were to do this Right? Let's turn our attention, then, to these heuristics and behaviors, because I think that, regardless of whether you're in the strong camp or in the weak camp, these are very important things that have some effect on the way markets behave and the way in which we think they do their job at allocating capital to risky uses. So, generally speaking and here I'm relying a lot on the work of Andrew Lowe in his book called the Adaptive Markets Hypothesis most of our heuristics and behaviors can be traced to something very well adapted to our evolution. The issue is that these otherwise rational behaviors in an evolutionary context are not particularly well adapted to other contexts, like financial markets.
Speaker 1:Let me give you an example of what I mean by that. So if you're able to do this, you should give it a try. If not, you could do it as like a thought experiment. But if you stand on a very shallow step not someplace you can get hurt. But if you stand on a very shallow step and you sort of lean over to look down and you feel yourself tipping forward, what's your instinctive response? If you feel yourself tipping forward, your instinctive response is to pull back. That's a very well adapted heuristic. We didn't have to think about it or reason through it, it just happened.
Speaker 1:Well, those of you that are familiar with flying airplanes might be familiar with the concept of a stall, and a stall is a situation in which the plane loses propulsion and very rapidly starts to descend. Well, when an airplane is in a stall I'm not a pilot, but I'm told by my pilot friends that when an airplane is in a stall and you are rapidly descending, sort of nose down, heading towards earth, your instinct is to what Pull back. The problem is that the laws of aerodynamics say that that's not what you should do. That, in fact, what you ought to do is go further into your descent, that is, increase the amount by which you're declining in speed in order to gain more airspeed and then slowly lift the plane back up. If you instead pull back, what you'll likely do is put the plane into a tailspin, and that's not recoverable.
Speaker 1:So there's an example to help you understand this idea that our heuristics and behaviors aren't themselves irrational, and I want to really counsel you against thinking about these things as being non-rational behavior or irrational behavior, because I think it misses the point. These behaviors are abundantly rational. They're just not rational, necessarily in a different context, like financial markets. These behaviors and heuristics I think can be grouped into two broad categories how we process information and how we assess risk and reward. And I want to go through these two broad categories because I think it gives us a really rich sense for a way to understand what's going on when market participants are trying to incorporate information in price and thereby allocate capital to risky uses.
Speaker 1:So let's start with information processing. In talking about information processing, I'm going to really think through primarily three different heuristics or behaviors. There are many, many, many. The research on this is replete. I think any discussion of this is misguided if it doesn't make some reference to Kahneman and Tversky's great work on the subject. And, by the way, when I mention books or I mention articles in these podcasts, you'll find the link to the text, or at least you'll find a citation to the text on the podcast's website. So I'm going to take just a couple. I'm just going to take three of these behaviors to help illustrate how market participants acting in the short run to understand information and incorporate it into price probably can not be understood to be acting in a way that is random in the statistical sense of random, but maybe are acting in more predictable ways.
Speaker 1:First, overconfidence People tend to overestimate the precision of their forecasts and overestimate their abilities. My favorite illustration of this is if we were all sitting together in a room and I presume that the people who listen to this podcast are roughly drawn at random from the population as a whole I would say to you raise your hand if you're a better than average driver. I've done this many, many times for years in teaching this class and it always gets a good laugh. But far, far more than half the people in the room raise their hands. But statistically that's impossible. More than half the people cannot be better than average, because that's the definition of average. This is a very well-adapted trait for human evolution. Without overconfidence, we would not take the kinds of risks that have allowed our species to thrive, primarily by being curious.
Speaker 1:When you take that behavior out of its context and you put it into a different sort of context, you get some perverse results, not irrational, but just strange results. So, for example, in a very well-documented paper by Barbara and Odeon, they find that women trade less actively than men, especially single men, and that the more trading activity there is is directly linked to poor investment performance. Women trade less actively than men and the higher the trading activity, the poorer the investment performance. That suggests that there's some behavioral thing at work here that might be perfectly well-adapted to our evolution but in financial markets may cause distortions in price depending on who the traders are and when they're behaving. A second paper by Mahlmendir Antaid talks about the tendency of CEOs to overpay for target firms. This has been repeated in study after study, but CEOs tend to overpay for target firms. Well, what's going on there? It would seem that CEOs are highly overconfident about their ability to incorporate that new business into their business by the cost savings they think it will engender by the new sales they think that they will be able to develop. All of which are examples, I think, of overconfidence at work and very relevant in terms of thinking about when you see new information, how do we incorporate that into price.
Speaker 1:Second is conservatism People tend to be slow in updating their beliefs based on new information. They initially underreact to the news and then gradually incorporate that information more thoroughly. What this tells us is if that's a Again, one can imagine this being a very, very important trait in human evolution. If you discover that there is this new delectable-looking thing that we now know to call an apple, but the first human encounters it. It's new. You don't know what it is. It could be poison, it could be good, someone might eat one and they don't die. But the way humans are wired is we don't then all immediately rush to eat this new fruit. Some people might eat it, some people might just eat a little of it, and we kind of wait and see and we let time develop in front of us before we take the decision this is okay to eat.
Speaker 1:Well, this, I think, is probably something that's very real and happens in markets all the time. This could potentially explain something that economists have observed in equity prices, which is called momentum. Momentum just means that when the price of a stock in the last 12 months has been falling, the stock price tends to keep falling, and when the stock price in the last 12 months has been rising, it tends to keep rising. It's very hard to explain why that would exist in equity prices if the strong camp were right, if the strong camp were right, and all that's happening is people are trying to incorporate information into price and it's more or less random. There should be no momentum in price, but there is. We see it, we observe it and one explanation might be the conservatism.
Speaker 1:Third, representativeness. Here the idea is that people tend not to account for the sample size of their data when making decisions. They just assume that their sample is representative. That's why it's called the representative tendency. Their sample is representative of the population as a whole. I'm sure that any one of you listening to this podcast can imagine a situation in which you have access to certain information that you've observed and you draw conclusions from it because you believe it's accurately reflective of the greater whole, the greater, whether it's the greater society, the greater market or what have you.
Speaker 1:I used to give this illustration to my students and it used to make the point very clearly. I apologize in advance if this doesn't make sense to listeners, but I taught at the University of Virginia, and the University of Virginia is a big, heterogeneous student population but where there are definite stereotypes of different students in different divisions at the university. And so here's the illustration I gave, just to sort of see if we can't make the notion of representativeness a little bit more real. The situation that I gave students is imagine that they meet another student at a party. This other student is intelligent, highly creative, very technical, has a lot of attention to detail, their style is very fashion forward and they're very social. Now, for those of you that are listening, who may be UVA grads I hope you're sort of chuckling because I bet you're immediately picturing who that student is and for those of you that aren't UVA students, you're probably basing your understanding of this person from your college, imagining that, well, uva can't be terribly different.
Speaker 1:And so I would then ask the students based on this description very intelligent, highly creative, technical, attention to detail, stylish, very social is this student in the College of Arts and Sciences, in the engineering school, in the architecture school or other. In regular year after year after year, students predictably select architecture school as the likely place where this person lives, and you can see the logic to it somehow. But what I tell them is that if you had no idea where this person studied, all you had was this information, if you didn't suffer from the representativeness behavior, what you would say is well, what's the? What's the distribution of students among those different divisions? And I would tell you well, the College of Arts and Sciences is 72% of the students, the engineering school is 17%, architecture is 3% and other schools account for 9%. And so someone not suffering from representative behavior would say well, without knowing anything more, the most likely answer is the College of Arts and Sciences 72% of the people are there, thank you. But again and again, year after year, the answer I got was the architecture school, which only accounted for 3% of students. We tend not to account for the sample size of our data when making decisions and we assume that it's representative of the whole.
Speaker 1:Here again, another academic paper that shows how representativeness can find its way in financial markets. A paper by Chopra and his co-authors called Measuring Abnormal Performance found that stocks with the best quarterly performance suffer reversals in only a few days as market participants correct for their original beliefs as being too extreme. Stocks with the best quarterly performance suffer reversals in only a few days as market participants correct for their initial beliefs as being too extreme. Let's then turn our attention to the second category of behaviors and heuristics, and that's about how we assess risk and reward, and here again I'm going to talk just about three. There are many, many, many, but these are three that I think really help paint an accurate picture of market participants in the short run doing something other than just cold, calculatedly incorporating information into price. The first is framing. People will make different decisions about the same risk depending on how the choices are framed. So here's an illustration Imagine that I offer to play a game with you and I offer to play a game with you based, or I offer to play a game with any one of you based on who will pay me the most to play this game and what's the game.
Speaker 1:Well, I would say the game is a coin toss with $50 for tails, and in my class I would say how much would you be willing to pay to play this game with me? A coin toss for $50 for tails, and we would get a set of answers. And then I would say, okay, now, new experiment, new game, same thing. I'm going to play a game with any one of you who pays the highest to play this game. But this is going to be a different game. Here I'm going to give you $50 and that gift of $50 is conditioned to the game. Gift of $50 is conditioned on tossing the coin and it coming up tails. Now, in both games you end up with $50 if it's tails and zero if it's heads. So when you're deciding how much to pay for that, you would think well, I have a 50% chance of getting $50 and a 50% chance of getting zero. So I would never be willing to pay more than $25 to play this game repeatedly.
Speaker 1:But what's interesting is that the first game is framed as a risky gain and the second game is framed as a risky loss. And when you frame something as a risky gain, people are willing to take more of that risk, which is to say, they're willing to pay more for the right to have that chance of a gain. But when you frame it as a risky loss, people generally are willing to pay less for that game. Now, of course, they're exactly the same game, exactly the same risks. It's just how we frame it, and it suggests that how one frames a set of risks will have a bearing on how that risk is incorporated into price.
Speaker 1:Second, assessment of risk and reward behavior is something called regret avoidance. So regret avoidance describes how people tend to blame themselves more for a bad result when the choice they made was unconventional. And given that they blame themselves more for a bad result when the choice was unconventional, means that people tend to avoid unconventional choices, and if they are going to make an unconventional choice, they must be compensated more for it. So let's do a little thought experiment around that as well. Let's say you own a blue chip stock and that blue chip stock has gone down 15%. What you'll find is that people will be unhappy that the blue chip stock is down 15%, but they're more likely to attribute that to either bad luck, short-term volatility or quote the market end quote. But if I tell you you own a small cap stock or a startup stock or a new IPO stock and it's down 15%, people will be equally unhappy that it's down 15%, but they will tend to attribute their unhappiness to the to some way in which this might have been a bad decision, and it will tend to force action, and so this creates a situation where investors that own less conventional risks might require a premium over and above the premium to take the actual risk, merely because of this notion of regret avoidance.
Speaker 1:Later in the podcasts, when we start studying asset classes, we'll see that small cap stocks and value stocks historically exhibit higher premiums than stocks on the whole. But one could argue and those are explained typically explained by different sets of risks that small cap stocks and value stocks have a different set of risks. But one could argue that the premium for small cap and value stocks are not for the risk of size or price relative to book value, but instead reflect a premium for regret avoidance. Small stocks and value stocks are not conventional choices and in order to take that risk one needs to be paid something more than the otherwise underlying risk. In a very important paper by DeBont and Thaler called Further Evidence on Investor Overreaction, they found that investors typically focus on gains and losses of an individual stock rather than their portfolio as a whole. They become more risk averse concerning stocks with recent poor performance and then they discount the cash flows of that company at a higher rate, making the price go down even further, and there's no rational way to explain that behavior, according to DeBont and Thaler, other than to imagine that it's some version of regret avoidance.
Speaker 1:The last of the risk reward assessment behaviors I'd like to describe to you is called prospect theory, sometimes also known as loss aversion, and I should point out here this is a good moment to point out I bet that those of you listening will note that these behaviors overlap a lot, and so therefore it's very difficult to ascribe the behavior to just simply one or another. It may be multiple, acting in simultaneity Prospect theory or loss aversion. People derive utility not from their total wealth but from changes in wealth from the current level. That's not novel, that's pretty classic economic theory, and therefore increases in wealth from current levels have diminished utility also classic theory, but where the prospect theorists or the loss aversion theorists go further is, they say but decreases in wealth from current levels that are proportional to the gain that I just described are more painful than those proportional gains. That is, we derive less satisfaction from gains than the amount of pain we derive from losses, and therefore loss aversion tends to lead us to taking higher risk in the face of losses.
Speaker 1:Now, conventionally speaking, if you're suffering from losses, one could rationally describe how the thing to do is to reduce risk. But what we observe in behaviors again and again is that traders people tend to take more risk in the face of losses. We don't know why exactly, but we know that it might have something to do with this notion of prospect theory or loss aversion. It could explain the high historical risk premium for equities. When we get to equities as an asset class, we'll discover that equities earn a much higher premium over the risk free rate T bills, for example than any economic theory would predict. But loss aversion could help us understand why. If we imagine that the pain we suffer from potential losses is so much greater than the benefit from potential gains, then to invest in equities, which is a highly uncertain asset class, we would demand more compensation to take that risk. Not because there is more risk, but because our experience of losses is more painful. There was a study done of traders in government bond futures that found that traders exhibited higher risk appetite in afternoon sessions after they had a morning of losses than they did in afternoon sessions following a morning of gains. So this stuff is real.
Speaker 1:Last time, when we come back, I'm going to talk a little bit about probability theory. Even if markets aren't entirely random in the short run and I think what we've discussed today suggests to me they are not even if they're not entirely random in the short run, and even if the uncertainty is not statistical uncertainty like a coin toss, but instead is radical uncertainty like Mervin King and his co-author describe, nevertheless, I think we can still use probability theory as a tool, albeit imperfectly, to understand risk and return in markets and to help us express the strength of our conviction about a particular risk or about a particular return for that risk. And so that'll be next time we'll talk about probability theory, just to introduce some basic definitions that we'll find ourselves using again and again when we turn our attention soon to asset classes, as always. Thanks for listening.
.jpeg)