Transcription
I shamelessly got into a debate with a discretionary trader about this idea of alpha. And one of two things always happens when this concept comes up. Either the person you're talking to is going to cite alpha as some sort of return in excess of a benchmark like the S&P 500, or they're going to correctly cite it as some sort of performance relative to a selected pricing model that is unexplained. And of course, you're subject to your classical model and parameter risk in the second case. But unfortunately for my sanity, this one was the former, not the latter.
And he goes, "No [ __ ] Roman. You know, if I'm generating positive returns in a bull market, it's going to look like beta exposure. There's a correlation there." And I really, I really didn't even know where to start in trying to explain the concept. So I just followed it up with, "What happens when there is a sideways market or a bear market?" He goes, "Well, I don't do too hot relative to a bull market." And sometimes I just feel like I'm taking crazy pills, man.
So I hit up my boy and he's sitting on a VRP desk this summer, and I tell him, tell him the story, tell him the interaction. And he goes, "Well, Roman, you know, if you walk like really fast, it kind of looks like running." And >> [laughter] >> he got me pretty good with that one. And shout out to my boy, you're funny as [ __ ]. Good luck on that desk this summer.
The The point is, it is running, right? It is beta exposure. But this brings up a really necessary discussion, one that we really haven't focused on on this channel. And that's this idea of alpha, an elusive statistical concept when it comes to trading strategies, portfolio management, and the financial markets, right? Academia loves to propagate that alpha, right, is non-existent. We can't statistically prove that you have any capacity to outperform via stock selection, this, that, the other. And they love to articulate that probabilistically, right? There are just fund managers that are lucky via the effectively law of large numbers. So, what the [ __ ] is alpha? That's That's really That's really the question that that we're trying to answer within this video.
To understand alpha, first we really need to understand beta. And why we start with these pricing models. It's not to predict anything. Nobody knows what's going to happen next. In the context of these undiversifiable risks, these undiversifiable risks are producing return streams. So, you look at something like the broader market, right? SPY, VOO, the S&P 500 ETF proxies, right? Historically, it's done a pretty good job of generating returns, right? But there's no guarantee that's going to continue into the future. It's effectively a bet on the broader economy, right? The smartest people in the world doing battle to try to make progress, and of course, money for their companies. And we're all just getting a slice of it. But if there's to be a massive regime change in that particular set of risky assets, everybody is going to be subject to the same drawdown.
All right, is there a way to measure that? Well, as a matter of fact, there is. And that is with beta. We're not going to talk about other undiversifiable risk factors right now, extensions to something like a CAPM. We just want to start with understanding this idea of what it means to produce alpha relative to a pricing model. So, we take something like the S&P 500 returns, right? And what we want to do is we want to pair them up with the returns that we generate from a particular strategy. It doesn't matter how we actually get those returns for that particular strategy. It could be discretionary. I could be sitting there clicking a buy and sell button. I could have an algorithm trade it for me. I could be holding a particular set of assets over a day, week, month, year. It doesn't matter. You're going to have a return. And what you do is you pair up those returns with the broader market returns, and you create a scatter plot. And that scatter plot is either going to have a trend, up or down, or no trend at all, meaning it's completely uncorrelated, right?
Okay, well then where is alpha? That effectively is beta. The slope of that line, that regression line, is telling you your beta. How does that strategy tend to move with the broader market, which again, is just the direction the wind is blowing. Maybe it's blowing favorably, and you have a high beta exposure. That's going to be particularly good for your return profile. Maybe you have a very low beta, right? Then when the market rips, you don't generate as much necessarily. Here's where things get interesting. Where does that regression line sit on the Y axis? This is the notion of alpha. So, if we take a look here, we're going to see one regression equation which has a significant amount of beta exposure, but no alpha at all, which means the strategy is not producing any sort of idiosyncratic return, meaning it's really just relying on the market to generate returns. Okay. Now, we're going to take a look at a different strategy, and I want you to notice that you can see the regression equation is relying on the market to generate returns, but it's also shifted up. Which means relative to just the market driving returns, the strategy is also producing some unexplainable return relative to this particular pricing model. Maybe if we account for other factors, right? That alpha, that intercept goes away. But, for now, in the context of this pricing model, that alpha remains.
Okay. So, how do we just get that alpha? That seems like a very like a very simple question to ask. How do I shift my regression equation up so that I have an intercept, right? Cuz if I can neutralize that beta exposure, effectively neutralizing that beta exposure means that I'm going to put another chart here. The regression equation is going to have roughly a zero slope, so the line is going to be just horizontal, and it's going to be placed above than the x-axis, and you're just going to be generating a return, which is the alpha, and that's completely separate from the market return. Then, it's saying, "Okay, it doesn't matter which way the wind blows, we're going to be generating that return because it has nothing to do with the broader market." Why is that return there? Right? Are you capitalizing on some sort of statistical inefficiency, so on and so forth? That is a quantitative research question. That's not a trivial question, and if anybody says that it is, please send them my way because I want to follow that strategy. All right, if it's so trivial and you know, easily explained as to why they have that that inefficiency that they're capturing in a particular strategy, I'm totally down to to run triple leverage on that strategy, but it's a quantitative research question why it exists and if it persists. >> [snorts] >> Okay?
So, that's this idea of alpha. How does it show up? And how do we find it? This is a really really interesting idea because academia loves to propagate that markets are efficient, right? We have the efficient market hypothesis and I'm not slamming the efficient market hypothesis. I think it's a very useful thought experiment, but it is nowhere near what we face in reality. Literally everything is mispriced. I'm not being dramatic when I say that. Every single option in equilibrium, hell, your [ __ ] car insurance, everything is statistically mispriced. In the context of a fair game, a zero-sum game, neither party should have an edge, which means statistically over time, if we keep engaging in that particular game, on average, we're both going to have net zero wealth. But, that's not the case in real life. Think about your car insurance, right? If insurance wasn't a profitable business, then nobody would sell it, but more importantly, I can 100% guarantee you that these contracts are mispriced because when you go out to buy your car insurance, do you get one quote for the same coverage? No, you get a bunch of quotes and they offer you different prices. Financial theory says, right? Law of one price. There should only be one price for a particular risky asset. But you just were quoted a whole bunch of different prices, right?
This is a part of a broader idea here. Because we're talking about this idea of alpha and capturing then this unexplainable return according to some sort of pricing model. Those mispricings create the inefficiencies that you can exploit which then produce that shift in the regression line on the Y axis to then capture what is recorded as alpha relative again to that particular pricing model. Where might this show up? I have a really great example in the context of everything I've been talking about recently with volatility drag, black swan events, and a hedge portfolio. So let's suppose that we are riding this market beta. So let's suppose that I have a trading strategy or I am invested in assets that produce a positive correlation with the market and I have a net zero alpha. All right, historically I have a net zero alpha. But I'm running a hedge sleeve. I'm running a hedge program on this portfolio which means effectively when there is another black swan event and the bottom blows out 20, 30 percent, what's going to happen to my portfolio? Well, all of the sudden, right, my returns are no longer going to be severely correlated with the broader market. Because I have that hedge in place, right, the efficacy of the hedge of course is going to determine whether or not you produce alpha, but because I have that hedge in place, I'm not losing money at the same rate I was gaining money relative to when the market was in a bull cycle. When the bottom blew out, my hedge kicked in and that's going to show up in the pricing model. Here's the the new regression equation as alpha, right? It's not timing anything. It is effectively saying, "Hey Roman, your positioning, right, was effective and it produced a return unexplainable by this pricing model." And the academics and purists will say, "Oh Roman, you know, that's a a statistical anomaly and this, that, whatever." And to that I say, think about something like baseball. When somebody steps up to the plate, that's the perfect example for every single professional risk allocator or personal risk allocator, right? You're stepping up to the plate and you have no idea what pitch is going to come and you're going to try to knock it out of the park, right? You have a batting average. What's expected of you? That's effectively something like your beta exposure. If you knock it out of the park, right, is that a statistical anomaly, right? Is that you performing at the upside of your variance window? It's positioning, right? You were in the position to swing at the pitch and it turns out when it was a fastball right down the middle and you crack [snorts] it and you knock it out of the park, right, that is you producing alpha effectively.
All right. So, having that hedge in place, right, the proper positioning in the portfolio. This isn't about developing a trading strategy, right? This is about engineering a portfolio for success, which is exactly what I'm covering in my brand new four-week live class where I talk about running a personal hedge fund. So, if you have significant capital at stake, if you're triple long, which means you're long your job, you're long your house, you're long your portfolio, and you're looking to add a hedge sleeve to your individual portfolio because you're running it on your own without a wealth manager, then I highly recommend you check this course out. If nothing else, understand the risk that your portfolio is exposed to. There is a free quantitative research note on the course page, link in the description below, you can download it for free. At least give it a read to understand the risk that you're exposed to if you're uninterested in learning about how to produce a hedge sleeve for your particular portfolio and your portfolio goals.
Okay, so this is where alpha shows up. It shows up when positioning pays off, right, relative to a particular pricing model. That is effectively an inefficiency in pricing. I was able to fund portfolio insurance, and then what happened? The bottom blew out of the market. If the market knew that beforehand, that insurance would have been jacked way up in price. But because I had that exposure to it before the crisis, right, that payoff is going to show up as alpha in the particular strategy. It's a really important idea to understand, right? Because this is the same idea as an insurance company telling you what after you completely wreck your car? What? Oh, we're just going to give you the old rates. It was It was an accident, you know? You're not going to do it again. It's like, no, they're going to jack your [ __ ] rates up, right? They're going to jack your rates up. Once that insurance pays out, once the market absolutely tanks, insurance in equilibrium is repriced. Everything jacks way up. Implied volatility jacks way up. It's too late to buy insurance. It's too late to buy insurance. You already crashed your car. The market already crashed the car.
So, all of this goes into the notion of alpha. This idea of producing a return that is orthogonal to some sort of some sort of factor in a pricing model. Whether or not it produces a return in excess to a benchmark is subject for discussion. Sometimes it does, sometimes it doesn't, which is why it's such an infuriating discussion when, you know, finance majors or business majors, whatever the [ __ ] say that it's just return in excess to the S&P 500. It's like, "Oh, I produced all this alpha." It's like, "I went um I went double long um Taiwan Semiconductors and I produced all this alpha." It's like, "Nah, dude. You just bought a stock and that stock has an extremely high beta and maybe there's some idiosyncratic outperformance, but that isn't producing alpha. That's not engineering a portfolio for success. That's just buying a particular name, right? What does that look like over 5, 10, 15, 20 years, right? A portfolio with an effective hedge sleeve continuously monetizing those drawdowns, buying assets for 60 cents on the dollar when you have 85 cents on the dollar, is going to produce significantly higher compound annual growth relative to somebody who is just trying to pick stocks and achieve that excess return spread, which isn't even necessarily alpha, right? They're going to get slaughtered by volatility drag because of how high the beta is for those individual picks to produce that excess return in a bull market year.
All right? So, this was effectively just a discussion on what exactly alpha is. I hope it provided some insight into the risk that you're exposed to and what it really means to generate alpha for your portfolio. If you're interested in checking out that class, I'll leave a link in the description below. Again, if anything else, check out that quantitative research note to really understand the risk that you are exposed to in your portfolio. That's going to do it for the fireside chat today. I hope you enjoyed. I hope you learned something. If you like this video and you want to see more like it in the future, please like, comment, subscribe, share. It helps me out tremendously. It is always greatly appreciated. Other than that, I want to thank you so much for watching and I will see you in the next video.