In Episode 151 of the Art of Boring Podcast, Rob Campbell sat down with Portfolio Manager Manar Hassan-Agha to discuss the importance of position sizing in investing and how factors like behavioural biases, market structures, and optimal betting strategies under the Kelly Criterion can impact returns at varying position weights. The conversation explores limitations in precisely calculating probabilities and edges for stock investments and how frameworks and checklists can be used dynamically to thoughtfully consider odds, edges, and optimal sizing for investment decisions. This discussion highlights the various personas or strategies that investors can adopt in dealing with both the winners and losers in their portfolios. Many of the concepts discussed in this episode are the research and works of others. Manar talks through how we think about applying their lessons dynamically and from a first-principles basis to the day-to-day management of portfolios at Mawer. 

Disclaimer: This podcast is for informational purposes only. Information relating to investment approaches or individual investments should not be construed as advice or endorsement. Any views expressed in this podcast are based upon the information available at the time and are subject to change.

[00:01:12] Rob Campbell: Manar, it feels like not too long ago we sat down on a podcast. Welcome back!

[00:01:17] Manar Hassan-Agha: Yeah, I appreciate you having me here again, Rob. This is a topic I'm excited about.

[00:01:21] Rob Campbell: It certainly is a wonkier topic. Fair warning to our listeners. We are going to get into some detail on a topic that might seem pretty plain and vanilla on the surface: position sizing or how big of a weight to ascribe to a single position in a portfolio.

Manar, I first heard you talk about this specific deep dive at our Top Down Scenarios Forum that we do as a research team. Let’s start there. What is that? What does it involve? And why do we do it?

[00:01:50] Manar Hassan-Agha: We are a decentralized team at Mawer, and curiosity is an important pillar of our culture. When you think of the top-down scenario forum, this has really become an annual tradition at our fall retreat. Our research team members get together, and we explore economic scenarios, themes, trends, and related investment topics.

We share those ideas in a freeform presentation, either written or in PowerPoint –  I'm a big PowerPoint guy. We're bottom-up stock pickers, and we intend to stay that way because that's really where our competence and our value-add lies. But this is an opportunity for us to practice our top-down thinking, trying to harness the creativity of the team and get those neurons bouncing and colliding between the team.

I've had great conversations coming out of this presentation, and the ultimate goal is to try to improve the inputs into us making repeatable, good decisions. So that's where this top-down scenario form is involved.

[00:02:45] Rob Campbell: I've heard it described as just an opportunity to help us stretch, go outside our comfort zone a little bit and talk about something that we do a deep dive on. Yours is less of a macro topic, but it is a broader topic. I'm just wondering, how did you land on this idea of position sizing as something to spend a lot of time researching?

[00:03:02] Manar Hassan-Agha: This is fascinating. So Michael Mauboussin has this great paper in which he talks about the two ways of generating alpha. One is you have an edge, and two is through position sizing. I became really excited about this topic because I immediately thought, “Position sizing is this squishy thing that can generate money for our clients.”

But the topic seemed like an afterthought. There wasn't a lot of attention paid to it, especially relative to binary yes-or-no decisions like buy or sell. To me, that seemed like inefficiency or a missed opportunity. And then I also thought, “How do I reconcile all this with great investors like Warren Buffet, Ed Thorpe, or Charlie Munger at Druckenmiller?” They all emphasize, to some degree or another, that betting bigger when the odds are in their favor is something that they ascribed to.

I recently heard Druckenmiller talk about what he learned from George Soros, and that's that 70 to 80% of the equation is position sizing, and one of the key things was to size bigger. “Go for the jugular,” he said, "when you have the conviction.” So you have all these great investors telling you that position size matters, but there's little attention paid to it. The combination of these things peaked my curiosity.

So I set out to survey the landscape. What I'm trying to do is stand on the shoulders of giants and put together a mosaic of the different perspectives and then apply my own judgment and framing along the way to Mawer’s process. Think of this as an open-ended exploration of the topic and me just trying to grapple with it.

[00:04:35] Rob Campbell: You mentioned Ed Thorpe. I'm glad you said that you didn't get interest in this because you were a degenerate gambler, and I mean that half facetiously. There are concepts that we'll discuss here today that originated from gambling principles and were applied in that format in the first place. As time has gone by and investors like Warren Buffet have seen their utility, they've applied them to the stock market, as well.

So you talked a little bit about why position size matters, and to a lay person like myself, it seems make sense. If something's going to return a lot, apply a higher weight to it. I wonder if you can get just a little bit more granular with respect to some of the underlying reasons why position sizing might matter too.

[00:05:18] Manar Hassan-Agha: Sorry to disappoint you, Rob. With my wife, I always say, “Hey, you want to bet?” So I just won $20 yesterday, but I realize it's all in the same pool, so what's the point? But look, if we think about it from first principles, I think it's actually a very important question. One reason position size matters is behavioral. For example, your entry size often matters more than the entry price because your size is a key determinant of your ability to stick with a winning position.

Let me give you an exaggerated example. Imagine you had 100% weight in one stock, and it loses 50%. Well, you just lost 50% of your portfolio. Versus if you had a 1% weight and you lost 50%, you lost 0.5% of your portfolio. I suspect you'd feel a lot more sanguine psychologically if you only lost 0.5% of your portfolio, and you may be a lot more open to doubling down, evaluating the facts more rationally, and not passing up on potentially a great opportunity.

[00:06:19] Rob Campbell: So your point is that, behaviorally, in time period one where you get that draw down (of course, you'll never know this), even if the investment were to work out and then rebound quite strongly after that, you're saying that a larger position size increases the behavioral odds that I'll just give up and settle the position entirely as opposed to sticking with it.

[00:06:38] Manar Hassan-Agha: Yeah, you're spot on. There's a lot of emotional damage and it's much more difficult, from a resilience standpoint, to be able to take a look at that and not question your judgment.

There's an interesting concept that traders have that they call the uncle point. That's the point you throw in the towel because you just can't suffer the pain anymore from the drawdown. So getting that precision roughly right may let you not throw in that towel and squander a winning investment.

[00:07:03] Rob Campbell: Is there something about the structure of markets that leads to this idea that position sizing really matters?

[00:07:10] Manar Hassan-Agha: Yeah, I think so. Another reason that the position size matters is because we can amplify returns. Let me step back first about the structure. Stock markets are fat tailed, so they have this power law distribution similar to wealth. There was a good research paper on this; over 90 years of U.S. stock market data showed that only 4% of all listed U.S. stocks explained the net lifetime dollar wealth creation, and the
remaining 96% collectively just matched the returns of a one-month Treasury bill.

[00:07:39] Rob Campbell: That's pretty wild.

[00:07:40] Manar Hassan-Agha: That just astounded me. Here's another fun way of looking at it, courtesy of Nassim Taleb. He makes a technical observation that one year of data out of 56, years up until 2009, explain roughly 80% of the S&P 500’s kurtosis – a measure of the tailedness of the distribution. You had a disclaimer up front; we're going to get into this.

So if you took out that one observation, your whole distribution changes. I like to think of it like this. You walk into a bar and there's 21 people. You go around calculating the average wealth and you get $10 billion. You think, wow, this is a really wealthy bar.

[00:08:19] Rob Campbell: I was going to say – you hang out at different bars than I do.

[00:08:23] Manar Hassan-Agha: But 10 minutes later, one person walks out, and you go around recalculating the average wealth, and it's only $100k, which is the median U.S. wealth. And you're shocked until you realize the person who walked out was Bernard Arnault, the chairman and CEO of LVMH and the richest person in the world. So you’ve completely changed the properties of the distribution. It's the same thing if you took Warren Buffet out of Omaha; look at the distribution of wealth and how it changes.

[00:08:50] Rob Campbell: Got it. Let's go from kurtosis to something that is perhaps a little bit more relatable for listeners like me: baseball. There's a lesson from baseball here, too.

[00:08:59] Manar Hassan-Agha: Yes, absolutely. So our goal is to have bigger weight in the right tail events, the winners, and smaller weight in left tail events, the losers. It's about how much money we make when we're right versus how much we lose when we're wrong. But we often think about a lot in the probability or frequency space instead of an expected value with kind of payoffs or magnitude.

So let's think about baseball. In baseball, we have the batting average, and the batting average is how many times you actually hit the ball relative to how many times you went out at bat. Whereas there's another statistical slugging percentage and that tries to differentiate between the hit produced by taking kind of the total bases of a hitter and dividing that by the number of at bats. So for example, a double or a home run should have more weight. 100% stock return is more valuable than a stock with a 20% return, but the batting average won't take that differential into account.

[00:09:54] Rob Campbell: So you're saying that the slugging percentage is probably a better measure of the utility of the batter, and there is an aspect of sizing that's embedded in that slugging percentage, which is not in the batting average.

[00:10:06] Manar Hassan-Agha: Yes, I think they're both very useful because if you can improve your batting average, that's still good. Usually the odds are on your side as you think the probabilities are. So your hit rate is better. There are a lot of great investors where their hit rate is below 50%, but they still make a lot of money. It's really a result of how much money they make when they're right versus how much they lose when they're wrong.

In investing, we can really take this one step further because we can amplify our returns through position sizing. We can get a thousand runs where in baseball, it's capped at four. Jeff Bezos talks about this in business, as well. For Amazon, it's a thousand runs where it's capped at four in the slugging percentage in baseball. Of course, the challenge is ex ante. Beforehand, we don't know the significance of that skew or that result. And that's kind of revealed to us in the long term. That's the challenge for everybody. But this is just theory.

[00:11:03] Rob Campbell: We'll get into that a little bit later. Let's actually go into the theory though. This does emanate from the world of gambling. Let's talk about Kelly Criterion because I think this formed a big chunk of the work that you did.

[00:11:14] Manar Hassan-Agha: Let's talk a little bit about the development of it and then we'll go kind of more into the actual formula. Kelly takes information theory and extends it to gambling and, humbly, there's a lot of math that's above my pay grade in this. But the book Fortune's Formula does a great job explaining this.

What I find interesting is the premise that true information is unpredictable. Think of a series of random events like rolls of a dice. I thought the book example was pretty funny. Your spouse calls you, tells you to bring home shampoo, but instead you bring home Shamu, the killer whale. That's a pretty ridiculous misunderstanding because if your spouse wanted you to bring home a killer whale, I suspect it'd be highly likely that it would come with a lot of explanation as to what, why, how. And that requires more bandwidth. So that's kind partly the link to information theory. But how do you link money to information?

Imagine I sent you a zero or a one. That's a single bit. And I told you for sure that the Kansas City Chiefs are going to win the Super Bowl at even odds. So I told you one means they're going for sure win. I suspect you'd say that bit of information is worth a lot of money.

[00:12:24] Rob Campbell: So you're Joe Biden, basically.

[00:12:27] Manar Hassan-Agha: I got all the power. I'm little bit younger, I have to admit. Anyway, there are many iterations of the Kelly formula. But the most basic and intuitive one, at least to me, boils down to the optimal fraction of your bankroll you should bet. So the percent of your bankroll. So if I had a $100, what percent of that $100 should I bet? Well, it depends on your edge over your odds. An edge is how much I expect to win if I could make the same bet over and over again, and odds is measuring how much profits you would actually make if you win.

[00:13:00] Rob Campbell: Okay. So those are the two critical components, and I'm guessing that as your edge increases, the more you should bet, and as the ratio of what you win versus what you lose increases, you should be willing to bet more of your bankroll.

[00:13:13] Manar Hassan-Agha: You should bet more of your bankroll. I mean, it's trying to solve for the optimal. So it's actually saying: Hey, this is the optimal amount. We're working in the field of theoretics right now, and then we'll break it down into more practical terms later.

But yeah, you're spot on. It also says: Hey, if you edge is zero, you should bet nothing. So if you have no edge, bet nothing. But there are two central questions that that the criterion asks and one is, “what level of risk leads to the highest long run? And what is the chance of losing everything?”

So what it's trying to do is maximize the compounded growth rate of your money without taking any risk of ruin. I love that because philosophically, what it's emphasizing is survival. “Hey, maximize my return subject to surviving.” One of the reasons if you all held all constant is if you bet a fraction, less than a hundred percent of your bank roll. So you think of a percent of your bank roll, you should theoretically never go to zero because a percent of something should not go to zero, as long as it's not a 100%.

[00:14:14] Rob Campbell: Let's go back to one of your examples from earlier where you talked about if you had a 100% position in something and it went down by half, Kelly's saying never put 100% of your position in anything because there are probably no 100% for sure bets. But even if you had a 90% probability on something, I presume that if it got cut in half, that would still be uncomfortable. As you apply Kelly to what you do, which is finding companies and deciding what weight to ascribe to them, what are some of the pitfalls or things that either are maybe a little bit different in investing versus gambling?

[00:14:48] Manar Hassan-Agha: So the Kelly might tell you: Hey, bet a 100%. Plus it depends on if your edge is so perfect, etc. But you’re right. There are fragilities in optimized systems. If you look at just in time supply chains, and then the pandemic hits; what happened? Well that exposed the fragility in just in time, which is a highly efficient supply chain. And one of that fragility with Kelly is kind of volatility. That's not specific to the stock market. Now it's nice, theoretically, that you won't go broke. You're right; I'm not so sure you'll be happy losing 95% or 50% of your bankroll. And so the optimal allocation leads to a lot of volatility, and it's difficult to stomach those drawdowns.

[00:15:30] Rob Campbell: And probably just increases the chance of behavioral error associated with that, too.

[00:15:34] Manar Hassan-Agha: Exactly. There's a nice Nassim Taleb graph I've seen in his technical insert. He shows hedge funds’ Sharpe Ratio up until the financial crisis. So the Sharpe Ratio is your return to your risk, where risk is defined as volatility. So it's really volatility adjusted returns. A higher Sharpe Ratio is a hedge fund trying to optimize for what they perceive to be the return over the risk. And that failed to predict their subsequent losses in the financial crisis; let's say you had a high Sharpe/three Sharpe Ratio coming up to the financial crisis. In the financial crisis, you still lost a lot more money.

So in fact it was actually a weak predictor of failure. So some high Sharpes indicated they're more likely to go bust, so you had a high Sharpe Ratio coming into the financial crisis, and you were more likely to go bust. Even though if you think of the formula, it's like, wait, I thought you were optimized for return over risk. That kind of shows you the fragility. What you're doing is maximizing your returns and minimizing your measured risk, but you're maximizing hidden risk.

So the point is that optimized systems can be fragile. But let me apply it to stock investing and where there is fragility for stock investing. A glaring one is: How do we know our exact edge and odds?

[00:16:48] Rob Campbell: That was going to be my next question.

[00:16:50] Manar Hassan-Agha: So there's a fallacy of knowing the inputs. We don't have the certainty of a deck of 52 cards in a very controlled casino environment, for example. And so what are the probabilities of an asset price moving up or down? What's the odds? They're not given in advance to us. They're not static, if we can even calculate them at all. And then a second fragility is the law of large numbers. It’s kind of the engine behind Kelly. Our sample size is just insufficient. So when we invest in stock markets, we have significantly less sequential bets than the casino.

[00:17:22] Rob Campbell: You're not playing thousands of hands in a given day. You might be holding onto a stock for like eight years or so.

[00:17:29] Manar Hassan-Agha: Exactly. So the law of large numbers tells you: Let's say the more you spin the roulette wheel, the closer you should approximate the expected percent of times you see red. There's a fixed probability that you should see red, but you approximate better the larger your sample size. So you have a tendency to converge with a larger sample. And you're right; we just don't make that many sequential bets and that means there's way more randomness and noise introduced.

[00:17:53] Rob Campbell: Lots of reasons to not just blindly follow Kelly. I presume the reason that you look so much into it is because it's a useful framework, this idea of thinking about edge, thinking about odds. This maybe plays into the optimization piece, but to what extent do you go down the rabbit hole to try and identify these as precisely as you can versus using it as a framework for thinking?

[00:18:15] Manar Hassan-Agha: I think we can use Kelly by thinking through it dynamically, from a first principle perspective and less so from a mechanical perspective. Even Ed Thorpe, who famously used Kelly and is an excellent investor with a track record, he rarely had to precisely calculate it. So for each investment, we can think through what our perceived edge and what the odds are with a directional understanding that higher edge and more favorable odds should lead to a bigger position size and vice versa.

We introduce a red-line hedge against our own hubris with a 6% single stock limit. So it's really a finite Kelly bet. There's a restraint on it. You can apply conservatism to the investment fraction to the extent that you think the future probabilities are uncertain. If it's more predictable, then you might be able to be like: okay, well we can get closer to that six. If it's less predictable, then you might apply more conservatism. And really this weighting exercise is intractable. If we knew exactly everything that we own, 100% weight in one stock, we could talk about the satisficing process later, but that's how we think about it from a first principle perspective.

[00:19:25] Rob Campbell: In doing it and applying it to your day-to-day, you mentioned a couple things to go through. What does that checklist look like in terms of never getting the precise number in terms of your edge or your odds? What are some of those framing mechanisms that you use or that you go through when thinking about that?

[00:19:41] Manar Hassan-Agha: So we use a decision-making checklist, and it’s dynamic. It helps frame the odds for us. To be clear, this is kind of a supplement to our diligence and investment process. It's not like I do this checklist and that's the investment. So this is not a replacement, but it helps keep you in check.

Some questions we ask ourselves: is our edge behavioral, analytical, informational, or technical? So this is the BAIT framework. The most likely edge for us is behavioral or analytical. If you think of informational edges, a lot of its arbitrage doubt, there's proliferation of data disclosure regulations. Technical edges are unlikely for our style of
long-term investing.

Let me give you an example. We didn't know with any certainty at the time, but this is what we thought. Take Publicis, the number two ad agency in the world. We thought perhaps on the behavioral side, there was pessimism on the sustainability of organic growth rates, given some pre-pandemic track record of ad agencies growing. On the analytical side, we evaluated the base rates in this business, and then we dug into the mix of the assets, which suggested to us that, hey, maybe there's differences in the quality of the business relative to its peers and its own history including the cyclicality.

In this case, I know that I said technical is not very usual, but we don't know. It was getting a French company discount. It's a French headquartered company, and it was trading below U.S. peers. Interestingly, it has predominantly U.S. exposure. 70% of its profit is actually from North America, we think. So the exposure is more U.S.-based.

Other questions we ask: What are the base rates for this type of business? Have I considered the outside and the inside view? And what relative weights should I assign to each? For example, if it's a drug trial, you may lean into the base rates to estimate the probability that the drug goes beyond the clinical trials into commercialization because that's a little bit more luck versus skill-based. You might lean heavily into the base rates there.

We also ask: Where is the business fragile? We're trying to separate small negative risks from terminal ones. For example, we own KDDI, the number two communication company in Japan. They recently bought a stake in Lawson, which is the number three convenience store operator. So I love the egg sandwiches in Japan, but I'm not sure I'm going to get my data package with my egg sandwich. I don't know why a telco needs to own a convenience store. It's not very good capital allocation, poor strategic logic. But it's likely a small negative risk. It wasn't really a terminal one when I think of fragility; it was roughly 5% of the market cap. It wasn't going to plunder the balance sheet.

So something more terminal would be like what I call double trouble, which is financial and operational leverage. The double whammy we talked about on the last podcast. There's three more questions. Is there an asymmetry or convexity in this investment? Here you can do a reasonable upside versus downside handicapping. Can I handicap the downside relative to the potential upside? Maybe there's a mispriced option or significant margin of safety.

Take Novo Nordisk, the number one biopharma and diabetes care, for example. It had this obesity optionality that, ex ante, maybe you weren't paying much for.

[00:23:03] Rob Campbell: This is going three or four years before we all knew about this company?

[00:23:07] Manar Hassan-Agha: Exactly. Ozempic obviously. It doesn't mean you're going to get that home run from an optionality perspective. The option might just expire worthless but that gives you more asymmetry in the investment. Am I overconfident? We ask ourself that. Have I pursued disconfirming evidence? Charles Darwin reportedly would write down right away any disconfirming evidence he comes across because he knows the human mind, how it distorts information. It wants to hide away that disconfirming evidence because it's not going with my thesis or not going with what I thought.

The last one I'll say is what I call the Investment Turing Test, which is: Do I understand and can I articulate the short thesis? The more you can understand the nature of the bet you're taking, the exposures you have, that's useful to help frame the odds.

[00:23:51] Rob Campbell: You mentioned this is a supplement to the rest of our philosophy and process and our investment approach. I'm just wondering whether you can tie in this checklist, which I presume is designed to ensure that you are slowing down and thinking in Kelly-like terms. How does that compare with the matrix process that we have today, which, in some ways, is also designed to help with sizing?

[00:24:21] Manar Hassan-Agha: If I step back for a second, it's like, well, how do we know what we know? I think as investors, we live in this kind of gray zone between nihilism – everything is meaningless – and precision, which is I know everything. I think we need to combat that instinct to embrace the gray zone with precision.

So for example, we don't need to add more assumptions to our discounted cash flow to shore up our confidence and valuation. Because for one, we introduced the danger of fooling ourselves, increasing our conviction without increasing our accuracy. And for two, the unintended consequence of that is complexity.

So one way of simplifying and going back to basics for more is actually our matrix framing, which by the way, in itself is already an abstraction. So it's already pretty hard. It's hard enough. The required ingredients, regardless of what you use, is humility and self-awareness. So knowing what you don't know.

But let me give you some examples. You're absolutely right; we have the matrix. There are some things that the matrix can’t account for and doesn't mean we should change the matrix. It just means we're supplementing it with additional considerations. Let me give you a simple example. Illiquidity is not accounted for in the matrix. We own a tail position in Bravida, a dominant Nordic servicer and installer of HVAC, electrical, plumbing systems, et cetera, and it plots really well on the matrix from a quality and valuation perspective. It has one of the largest gaps between the theoretical weight that the matrix would suggest to us as a guideline weight versus our actual weight. But that's a technical constraint. We can't get enough position into it. There are also times where risk management dominates. And the matrix won't appropriately reflect that judgment because it applies a linear weight to risk as one of the three categories of quality assessment.

[00:26:11] Rob Campbell: So this is where the fragility in your checklist becomes that much more important.

[00:26:15] Manar Hassan-Agha: Absolutely. So let me give an example. We own XP, the number one independent Brazilian broker dealer. It plots really well on the matrix and the suggested matrix weight, let's say, is higher than the actual weight. But there are some compounding risks of – well there's a capital market cycle; it's in Brazil, so emerging market risks; and there's a black box with financials. So we likely tap out at a certain risk budget so to speak, even if the matrix says go to 6% because that's the quality times of business. And that's because the risk may be understated in that particular point. We don't want to be optimized in that to be compliant to our matrix.

[00:26:59] Rob Campbell: I imagine this applies to anybody allocating capital, whether it's a client with manager selection and just dealing with new information. One of the interesting things that we've seen in studies is just that over time – and this is not specific to us; this is just institutional investors at large – there's a tendency for their buy discipline to be better than their sell discipline. Is there anything that Kelly offers in explaining that?

[00:27:25] Manar Hassan-Agha: I came across a study which effectively said that the institutional buyer tends to make great buying decisions but not very good selling decisions. They attribute it to what they call selling fast and buying sow. And I thought that was very interesting because what it's saying is, ‘Hey, maybe we're not allocating as much cognitive resources to our selling decision as we do to our buying decision.” What there's a tendency for institutional investors to do is, for example, look at the tails of their portfolio. So at the end of the portfolio and say, “ah, well this one has been underperforming, and use that as a selling decision.

[00:28:04] Rob Campbell: Not necessarily the sizing tails. You're talking, “Hey, I've got this great new buy idea I need to make room for it in my portfolio. Let me look at the top third of the portfolio. It's been up more than 20% over the last 12 months and choose from among there as opposed to the better method, which would be to look at the whole portfolio and their expected returns and select from there.” Am I thinking about that right?

[00:28:25] Manar Hassan-Agha: I think you're spot on. The better quitting decision, so to speak – and Annie Duke talks a little bit about that – is what's the expected value? So look through your entire portfolio and look at what's the lowest expected value for this for this opportunity here? We do this process. If you think about it, our matrix process is quarterly, so it’s dynamic, and we look through all our holdings as we rank them through on the portfolio. But still, we've implemented an additional circuit breaker. And that forced us to consider all of our holdings for sale, not just the ones in the tail.

And it's a checklist question. We love checklists. You could do this in different ways, but we strip out the tails of the portfolio. So these are the ones that buy stock return, the ones that contributed the most as to the top contributors and the bottom contributors because those are already salient in our mind. We know, roughly speaking, those we kind of outlined the stocks, what I call in the mediocre middle. And put them at the forefront of our brain. Have you considered these for sale when you're thinking of selling stuff?

[00:29:32] Rob Campbell: It's amazing how much of this comes back to behavioral stuff; it's pretty incredible. Let's shift to another situation that hopefully you find yourself dealing more with the former than the latter. What happens when you've got real winners in the portfolio over a long period of time, they're up significantly versus
losers in the portfolio those that have already drawn down quite a bit? What are some insights from your deep dive into how to deal with those types of situations?

[00:30:00] Manar Hassan-Agha: That's a good problem to have, Rob. One of the books called Art of Execution talks about one of the make money strategies for having a winner. And he talked about behavioral. What's the typical urge for if you have something that goes up? You want to sell.

[00:30:18] Rob Campbell: Unless you didn't own it, in which case you probably want to buy it out of FOMO (fear of missing out).

[00:30:23] Manar Hassan-Agha: Or you resent it. There are different reactions to it depending on your temperament. So let's say you own it. Scratching the itch, so to speak, he calls it being a connoisseur. So if you think of connoisseur, they have all these fine wine aged bottles in their cellar. But just don't drink it all. Drink one bottle every now and then. So effectively what it's saying is do a little haircut, trim it, scratching part of the itch. So having a connoisseur approach to that is an interesting way to approach some of your winners. The difficulty is you don't know if it keeps winning. That's one. There is some prudence to kind of trimming and doing some haircut to that.

[00:31:01] Rob Campbell: Provided your fundamental analysis suggests that the business is intact and fundamentals are doing well, the connoisseur approach allows you to take some risk off without missing out on what might come next.

[00:31:13] Manar Hassan-Agha: Absolutely. Part of it's like, yeah, is it a crazy stretch valuation? Or are the fundamentals are eroding? And it's still winning, then that might be a puzzle you want to solve. I think another mental model in this connoisseur approach putting it into perspective of the fundamentals is, am I conflating the cyclical with the secular?

Is this a cyclical thing where it's outperforming today because there is a temporary demand supply imbalance, or is this secular? Like it's taking market share, it has a long runway, it's cookie cutter runway because it's doing the same thing over and over again within its competence. So that's another mental model to apply that – to think about is this cyclical or secular?

[00:31:53] Rob Campbell: What about on the other side? I imagine these are the ones that actually cause the greatest anguish, which is, “oh gosh, I own this thing, and it's down 30 or 40% since I first bought it.” What are strategies, informed by Kelly or otherwise, that have proven to work in dealing with those situations?

[00:32:11] Manar Hassan-Agha: That's the worst side to have, Rob. So two ways that the book kind of outlines of making money in terms of approach. This is kind of a money management approach. One is being an assassin and two is being a hunter. An assassin cuts their losses quickly. You're on the losing approach, and there was a tendency that either a third loss of the stock, they just cut their losses. They're like “it’s not going my way.”

Now, of course, the risk is you don't give it enough time to play out, and we're long-term investors, so always keep that in mind. This is not a wholesale implementation of this; this is just, “hey, what's the perspectives out there?” Another reason an assassin might cut their losses quickly is what Peter Lynch calls the pig in a trough approach. The bigger pig kicks out the smaller pig in the trough. This is where a new idea is better, or perceived to be better, and kicks out an existing one. So that's another way of considering it.

Now the other make money approach in the author's perspective is a hunter. A hunter is more contrarian. They double down or they average into the position. And that requires more patience and discipline. So let's say you have a stock; it's coming down so you buy one third of your desired position, and it keeps going down. You buy another third. So you have three bites at the apple, so to speak, and you don't put a full stake in immediately. That requires discipline and  patience. You have the risk that doubling down eventually leads to ruin. You double down enough times and all of a sudden, your bank roll's gone.

[00:33:44] Rob Campbell: So far, we've been thinking about these situations in the context of a broadly diversified portfolio, but looking at individual investments, I've been thinking of them so far as over time. So it's been five or six years, and this has just been the trend. But some of these movements happen quickly too. They could happen in a day. Maybe there's a great earnings announcement and the stock just shoots up. Are the approaches different for those intraday movements versus these long-term trends?

[00:34:10] Manar Hassan-Agha: Michael Mauboussin has a great concept. So there's two times where you find yourself on the extreme ends of the emotional pendulum. One is the person is overboard. So you think you're overboard in the water. You're drowning. That's when the stock kind of drops in one day, significantly against some type of benchmark. So a one day drop of more than 10% against the benchmark, let's say. The other side is ecstatic and euphoric because your stock went up 10-15%, more than 10%, he calls that celebrating the summit. So you're on the top of the mountain Mount Everest. And, of course, most people die on the way down, not on the way up.

So you're on Mount Everest and you're on that extreme. What's useful for these type of scenarios is what is called the “redo checklist” from Checklist Manifesto. A redo checklist is where you read the checklist and you do it as you're reading it. Whereas the other one is “do confirm” and do confirm means you're doing the stuff and then occasionally check the checklist to make sure you're confirming like, “Hey, yeah, I'm on track.”

So a redo checklist gives you, in a stressful moment, an ability to take away that emotion and think about it rationally. And so again, this is a checklist that focuses on base rates, and the idea is, “what's the base rate of this stock that's down on one day?” Let's say it's an earnings or non-earnings event. And then you go through a series of things in terms of momentum, quality of the business, and the valuation of the business. And, of course, as you go through that, you're sacrificing the sample size for a better fitting case. As you go through that, it kind of says, “Hey, 30, 60, and 90 days after that one-day event happened, here's the base rate of what the stock did with this sample size.” So let's say there's 50 stocks that dropped 10% or more in one day against the benchmark with these characteristics. And the base rate is it underperforms in 30, 60, 90 days subsequent to that event. And that's interesting because this is just a supplement to fundamental analysis.

What's interesting about it is, I've done this before, and it immediately makes you just think rationally, and it removes any psychological damage or anything like that. You could do the fundamental analysis even better and put it into some perspective or context. Now, of course, our horizon is longer than 30 or 60 or 90 days. Our turnover is 10 or 11% most recently for a Global Equity strategy. We're long-term investors, but it's still useful from a timing perspective, maybe don't add immediately. Be patient. You might have had the inclination, like you're so excited, you're celebrating the summit, and you're like, “yeah, keep going.” You become a momentum junkie or something to that effect. So it just grounds you.

[00:37:03] Rob Campbell: I imagine diversification's important there, too because base rates require a larger sample set, as you said.

[00:37:10] Manar Hassan-Agha: Yeah, absolutely. At some point, the more you go down the road of while these characteristics fit in the case, the sample size shrinks and sometimes to an extent where it was 50 while and for how long in that particular scenario. So the more you step back, the less relevant it is to your case. There's randomness and noise in this, so keep it always in context. But I find it useful as I've gone through it, grounding psychologically and behaviorally. There's some intuitive stuff that comes out of that checklist, which we can touch on later.

[00:37:42] Rob Campbell: I do want to go back to something that was on my mind from what you mentioned earlier, this idea of okay with losers: there's a hunting strategy; there's an assassin strategy. Can you clarify for me, because both seem to be diametrically opposed to one another yet both have proven to be successful, is your point that these are two approaches that successful investors have employed situationally? Or that successful investors have tended to exhibit more of one versus the other?

[00:38:14] Manar Hassan-Agha: The underlying point is that you have to fit your style, your own emotion, your own behavioral biases. Some make money managers will tend toward having a more of a hunter style. They're more contrarian, they're patient, they're disciplined, they like to average down. That approach works for them. Some others have a tendency that they can move off their positions quickly and change their mind very quickly to some degree. So both can make money in that scenario. The study showed that the one thing that doesn't make money is being a rabbit; you're just kind of scurrying into your hole, and you effectively hope that things will turn around.

[00:38:56] Rob Campbell: The do nothing approach, statistically in that analysis, was worse than either the hunter or the assassin approach.

[00:39:03] Manar Hassan-Agha: In that analysis. Now, there are times, by the way, that do nothing is a very good approach. It is a very intelligent activity because it's really a risk management. You have a risk budget. It's something to keep in mind. And on the winning side, the losing approach in that study was being a raider, they call it. If you think about, it's like the moment you got a little bit of money, so it's winning, you quickly raid that position and close it down, locking it in.

Usually, raiders were rabbits. That's probably a pretty toxic combo. You quickly take away all your winnings, and when you're losing, you do nothing. So if we go back to the top of our conversation, while stock markets are fat tailed, they have a power law distribution. So if 4% of all listed U.S. equities is contributing all the net wealth creation, you don't want to be a raider when you're also a rabbit.

[00:39:57] Rob Campbell: Can we wrap up, Manar? We've covered a lot. A lot of theoretical. We've gotten tangible in terms of specific situations, examples. As you look back on this deep dive that you did, what are the real key messages that you take out of it?

[00:40:12] Manar Hassan-Agha: So the key messages for me really are protect the downside. So think about that paranoid conservatism of the philosophy. Survival is critical. Bet bigger when odds are favorable, more favorable, so directionally and not mechanistically. So think of it from a first principle perspective with a finite red line hedge against hubris. This is requires humility and awareness. Diversify, which is another way of taming Kelly. As you diversify, you want to tame some of that volatility. And hey, we just don't know, we could be wrong. So there is a diversification element.

Be disciplined in your money management. Have strategies before ex ante. When do you double down when you shouldn't double down? When do you cut your losses? Are you a hunter? So be disciplined in that kind of money management. And ultimately, there's more than one way to skin a cat. I've heard Howard Mark say that there's more than one way to skin a cat. It depends on your style and your philosophy. You have to fit your own style and psychology.

[00:41:19] Rob Campbell: Great. A self-awareness associated with that as well. Well, Manar, thanks for taking us on that tour. Thanks for sharing all the research that you did, and I hope to see you back on the podcast again soon.

[00:41:28] Manar Hassan-Agha: My pleasure, Rob. Thank you for your time.