TIP773: How Systems and Simple Math Shape Better Investing w/ Kyle Grieve

23:08Get started today at vanta.com slash tip. That's V-A-N-T-A dot com slash tip.

23:18Preston Pysh:All right, back to the show. So in the manufacturing example, let's say the CEO had 30 years of experience on the floor working without the use of automation. He's going to have all sorts of new problems to deal with. And since he wasn't too familiar with that technology before, he might not be the right person to lead the business as it scales up and adopt new technology. So scale is a very significant problem to think about. I really like growth companies. So one potential risk as a business grows is whether the business model can work at a bigger size. I remember doing a bull bear thesis with a member of the TIP mastermind community on an unnamed business.

23:53Preston Pysh:And one of the subjects brought up during that call was whether the business, which was a serial acquirer of niche industry businesses, would be able to continue using its acquisition criteria if the business scaled up in size. So, you know, when looking at it, Sure, they had success buying businesses for just a few million dollars, but could they have the same success buying a business for$10 million? It was a pretty challenging question to answer, and it requires some imagination and thought experiments to arrive at the best possible conclusion. Another issue that I've run into with scale is managers who don't adequately communicate the changes to their business as it scales.

24:25Preston Pysh:This could mean that management runs into unexpected problems, or it could be that they plan for these problems and don't feel a need to share them with the market until they make those actual changes. For instance, let's say a business is able to scale up production fast. Obviously, there's other problems that might arise, which I've already gone over. So they now, let's say, can double the product, but they just don't have anybody to actually sell that product. So as the business scales, investors might model for all new products to drop down to operating profits. But that's generally just not how it works.

24:56Preston Pysh:If you require new customers to fulfill your increased output, you're going to obviously need salespeople to go out there and sell the things. Another observation I've seen is that R &D spending often goes up as a business scales. So in a perfect world, the business could maintain R &D at an absolute level as its revenue scaled up. That's obviously the best case scenario in a scalable business. So you want to see how R &D spending has moved up as a company has scaled. You can do the same thing with things like sales, general and marketing. If a business has grown revenue by 20 % per annum for the last five years, what are R &D and SG &A as a percent of revenue?

25:32Preston Pysh:Is that number staying the same, growing, or shrinking? A shrinking number shows economies of scale. A growing number shows diseconomies of scale. So let's say it's 2019 and you're an employee of WeWork. One morning, you're at your desk early in the morning, and the excitement in the air is palpable. You see your CEO, Adam Newman, stride barefoot through headquarters, grinning from year to year. You're giddy with excitement as you've witnessed the company that you work for grow its top line from only 436 million to 1.8 billion over only two years. But how does a company that re-imagined the modern office end up with billions of dollars in losses from simply renting desks?

26:10Preston Pysh:Something just doesn't quite add up. As an employee at WeWork during this time, you not only observed epic revenue growth, but you also noted an epic growth in the headcount of people working alongside you. Scores of new hires seem to be drop-shipped in weekly. The business is growing so fast that paying for all this talent just isn't a problem at all, you tell yourself. And you're right to some degree. But then the bomb drops. In order to get up to$1.8 billion in revenue, you read an article on Apple News saying that WeWork had actually spent$1.9 billion to generate that new revenue. Now, as the market's love affair with WeWork comes to an end, so unfortunately, does your employment as the business's evaluation collapses to near zero.

26:49Preston Pysh:So what happened here? A business can obviously scale up poorly. And I would argue that paying$1.9 billion to grow revenue by$1.8 billion is not a good use of capital. And it all starts with incentives. If you thought EBITDA was a poor metric to achieve your KPI, then WeWork's abuse of KPIs was taken to, I think, a whole new level in terms of destroying shareholder value. So let's go over some of them. So they had this one KPI called workstation capacity. It was an estimate of the number of workstations available at open locations. Okay, that's not that bad. Then they had a membership count KPI based on demand and enterprise customers.

27:26Preston Pysh:Also, that's not that bad either. But now things started getting a little bit strange. So they had a run rate revenue. And basically, they would use gap numbers on just the last month. And then they would annualize that monthly number into the future, which obviously would create a very, very large number. and also assume that there was no fluctuations in the growth of the business, which obviously wasn't the case. Now, the real big one here is what they called community adjusted EBITDA. Community adjusted EBITDA, the gauge WeWork devised to measure net income before not only interest, taxes, depreciation, amortization, but also building and community level operating expenses.

28:07Preston Pysh:A category that includes rent and tenancy expenses, utilities, internet, the salaries of building staff and the cost of building amenities, which WeWork has described as our largest category of experiences. So basically, the business was incentivized to grow its top line by just any means necessary. So the next time you're looking at a scaling business, make sure that management is incentivized in some way to generate real profits for shareholders, not imaginary ones, such as this farcical community-adjusted EBITDA. So one way to avoid businesses with misaligned incentives is to create algorithms that filter these types of companies outside of your investable universe.

28:44Preston Pysh:So what exactly is an algorithm? An algorithm simply turns inputs into outputs. They are the working parts of a feedback loop. And the best part of an algorithm is that if you put the right amounts of inputs in, it always spits out the same output. It's like baking a pie. If you have the ingredients, follow the directions exactly as given, cook it in the right temperature for the correct amount of time, then the pie will always be the exact same. But if you alter the ingredients, you know, maybe use a little less sugar or estimate the volumes and weights of ingredients or eyeball things, or maybe try to bake it a little bit faster by using a higher temperature, then your output is going to change.

29:19Preston Pysh:Sure, it'll probably still be a pie, but it's not going to taste the exact same as the original recipe. So I've already discussed kill criteria and the Kono uncertainty today, and they are all form of algorithms. I feed them information and they give me an output. Then it's up to me to decide what to do with the information that's given to me. As the saying goes, investing is part art, part science. The data that we're given is sometimes quantitative. We can look at a business's earnings release and make conclusions based on the numbers that we trust are real. That's kind of the science part of investing.

29:49But the art part of investing becomes more subjective.

29:53Preston Pysh:For instance, I can conclude that a business has a cone of uncertainty that is either widening or shrinking. What do I do in that case? Or what if I look at my kill criteria, the date expires, and the business feels like it might need maybe another quarter or two to avoid triggering my criteria? The point here is that while I love algorithms and they are helpful, they're only valid when the person using them takes action. If you have an algorithm telling you to take an action, but you're ignoring the action part, then the algorithm becomes a model with zero world applicability. So if you create algorithms for investing to improve your decision making, make sure that you're actually using them.

30:29Preston Pysh:Charlie Munger once said, the secret algorithm to life is doing more of what's working. So I wanted to take this chance to share with you a little bit on my life, which I think is working. And it's not a few things, but obviously they have large impacts. So the first thing is just spending time with my family and especially my son. And next is looking at investing, which is to invest in a mix of long term and short term opportunities. That's also working really well for me. Another thing that I've noticed has made a huge difference is practicing gratitude as often as possible. And another thing that I constantly am trying to work on is becoming a kinder and kinder person each and every day.

31:01Preston Pysh:And obviously, to do with kind of work-life balance, while I obviously have this job that is pretty demanding, I'm lucky enough that I can kind of optimize myself for happiness while also being good at my job. And the last point here is just self-care, you know, and that's physical, emotional, psychological, taking care of all three of those things. Because I feel like if you don't take care of them, it's really hard to thrive in basically any area of your life. So Warren Buffett, in his final letter to Berkshire Hathaway shareholders wrote,

31:48Preston Pysh:Now, I think Warren's internal scorecard and ability to live by the golden rule helped him find his own algorithm for a great life. And while a great life in Buffett's eyes might not be the exact same for me and probably the majority of other listeners here, that's not really what matters. What matters is going back to what Charlie stated and doing more of what works specifically for you. For investing, the long-term algorithm for success, I think, is to just focus on cash flow. If a business generates increasing cash flow in the future, chances are its share price is also going to perform incredibly well.

32:19Preston Pysh:This is where I put a lot of my focus. Note that I said cash flow and not free cash flow because I invest in many, many businesses that have a ton of reinvestment opportunities. The cash from investing part of the cash flow statement is often filled to the brim with new acquisitions. So this obviously serves to depress cash flow. So if you screened using just that number, many names in my portfolio might not look like they are generating too much cash. This is why I like using metrics such as Buffett's owner's earnings, which helps to show how much cash the actual business is producing before investing in growth.

32:51Preston Pysh:When I look at my portfolio that way, I can clearly see that my businesses generate a ton of cash. And I think they're likely to grow that amount of cash that they generate many years into the future. So if you examine any business whose share price is appreciated and look at its operating cash flow, I can guarantee you that it has probably increased substantially. So just look at the Magnificent Seven. The average cash from operation 10-year keggers on those seven great businesses is over 27%. At that compounding rate with just a fixed multiple, a company can expect its value to be an 11x over just 10 years.

33:21Preston Pysh:Now, I've said this many times on this podcast, but if you have a business that is capable of compounding its operating cash flow for many years into the future, you're making a very, very big error if you decide to sell it. And I'm going to continue saying it as I think every investor needs to understand it, including myself. But here's the thing about algorithms. Even in the cash flow algorithm that I just discussed, things happen in the meantime. Over 10 years, if a company compounds cash flow at 27%, it's going to be worth significantly more 10 years later. But in any given year, anything can happen.

33:53Preston Pysh:This is where critical mass comes into play as a mental model. So you see, if you were to own something like Tesla during the past 10 years to get the full benefits of its ability to compound cash, you would have had to hold shares through four separate 40 % or greater drawdowns. But for investors who did understand critical mass and could see that Tesla had drastically improved margins across the board and would continue to do so, Tesla became an absolute slam dunk investment for them. So what exactly is critical mass? It's a point in the system where enough participants or supporting resources converge, allowing the system to shift into becoming a self-sustaining or having radically increased momentum.

34:32Preston Pysh:Allowing the system to become self-sustaining or have radically increased momentum. If you've ever read Malcolm Gladwell's tipping point, you're on the right path. Now, how can we leverage critical mass when considering a business in many ways. The first way is to use it by focusing on threshold. If you're trying to make a change and need several people to support your journey, you don't necessarily need everyone to actually get you to where you need to go. You only need the right amount of people just to flip the system. So in Malcolm Gladwell's excellent follow-up to the tipping point, the revenge of the tipping point, he mentions this magic third.

35:06Preston Pysh:So if you can change a third of something, such as board representation, you can often reach the tipping point that you need. So as a board member, let's say a nine person board, if you want to make a change, you need to either sway two or more people onto your side who are already on the board or replace two of those members with new members who agree with you. Now, I know the likelihood of me being on a board is quite low. So instead of focusing so much on making changes to a business, I prefer finding businesses where the right people are already in place and they just need to continue to improve the business.

35:36Preston Pysh:So I like to look for leverage points. If you know a company is accumulating cash and has a large runway to deploy that cash at high rates of return, then I know the business could eventually reach a critical mass. For a serial acquirer, that might mean that they start with one person who just looks at M &A opportunities. But over time, they can create a system where the business gets broken down into small units, which can also find more and more M &A opportunities, which creates a very sustainable system, something like a Constellation Software or a Bergman and Bebbing. Another consideration is time.

36:06Preston Pysh:Critical mass is around the corner for one business while being just a pipe dream for another. It's essential to keep this in mind when considering your investments. Now, I like to look for companies that can not only grow their top line at a decent clip, but also improve margins. This will allow the business to grow cash even faster than the revenue due to operating leverage. But sometimes that operating leverage can take some time to actually show itself. And if I mistakenly believe that a company has operating leverage now, but in reality won't see it in meaningful amounts for the next few quarters or even years, that creates an environment where investors are very likely to be disappointed.

36:38Preston Pysh:Another interesting aspect of critical mass is that it can be reversed. So if an input required for critical mass is reduced enough, then it can exit the critical mass state. I often see this in businesses, especially in smaller companies. So a business might create a hit new product that its customers absolutely love. However, if the business does not have a competitive advantage, then competition can come out of the woodworks, replicate the product, sell it for the same or lower prices, and render the original product's advantage completely obsolete. So I once owned a cannabis company called Cannabis Capital.

37:11Preston Pysh:This business had these cigarette-shaped cannabis products that produced a lot of revenue and profits for the business for a very short time. And investors extrapolated that advantage into the future, myself included. But that advantage, unfortunately, was very short-lived. And while it helped propel the business upward while they were selling well, the business was unable to sustain its critical mass. and once the product was no longer selling, the business quickly lost all of its momentum. So where I have the most use for critical mass is probably in my investing in inflection point businesses.

37:39Preston Pysh:I could easily call these critical mass businesses. The reason I look for inflection points is that it's often a quantitative way to observe that a critical mass is actually happening inside of a company. For my inflection points, I'm looking for a business with two consecutive quarters of positive cash flow or profits. This means that the business is often at a point where operating leverage is coming into play. And many of its past expenses, which ended up being a drag on the business, are now helping that business grow its bottom line. As much of the revenue that's incrementally added over time is much further above fixed costs.

38:10Preston Pysh:However, as I mentioned earlier, falling out of critical mass is a real risk that I must continually monitor very closely. Business is complex. And being a profitable business is even harder. Only the very best companies managed by the best management teams will experience continued profitable success. So the approach must be managed for potential downside risk. And I do this by buying businesses that are generally pretty cheap when looked at on a forward basis. Unfortunately, there's always the risk that what happens in the next year is not aligned with my thesis, which is why these positions are generally smaller than my higher conviction bets on more established and deep remote businesses.

38:46Preston Pysh:So my overall goal in investing, whether that's from investing in more obvious, well-established businesses or less obvious inflection point businesses is just to find compounders. This is an excellent transition to the mathematics segments of this episode. Compounding is a powerful concept because if you get it and can set yourself up to take advantage of it, you're going to reap some mind-blowing rewards. But if you don't understand it, you can easily fall prey to compounding, but in the wrong direction. The most visible form of compounding is just probably compound interest. When you earn interest on money and reinvest it, you earn interest on an even higher amount.

39:19Preston Pysh:As this process continues over time, the potential for that pile to grow is simply mind-boggling. Where I think a lot of confusion comes from in the compounding process is from the less visible forms of compounding. Compounding happens all around us. But because the changes in the short term are kind of imperceptible, it can be quite challenging to understand the sources of compounding. Where I think many people misunderstanding compounding is actually in their credit cards. So credit card interest is a great example here. While I did say that compound interest is visible, I think most credit card owners are probably unaware of just how credit card interest actually works.

39:52Preston Pysh:And it's not their fault. I think credit card companies are intentionally vague about how it works. After all, if you pay off your credit card each month and avoid paying any interest, then the credit companies aren't making any money off of you. So let's say someone is currently carrying a credit card balance of$5 ,000 with an annual interest rate of 20%. A 20 % may sound like it's applied just once a year. But there's a pretty epic catch here, which most people don't know. And that's that credit card interest is actually compounded daily. So here's how it works. You divide the 20 % by 365 days, which gives us a daily rate of 0.055%, a number which seems, you know, completely harmless on the face of things.

40:27Preston Pysh:But each day the person owes, they're getting charged interest on a growing balance. Now let's imagine the person doesn't make a payment for a year. So their balance won't just be$6 ,000, which would be 20 % of 5 ,000 added to the$5 ,000. It would actually be$6 ,100 because of these daily compounding effects. You are paying an additional$100 solely from the compounding process. This is why when you go and pay your credit card bill, they have the minimum payment button, which is never the entire balance. The credit card companies know that if they get you to pay the minimum, your interest payments will accrue and they'll make more money off of you.

41:02Preston Pysh:Another very non-intuitive concept regarding compounding is how compounding is convex to the upside and concave to the downside, meaning it has positive asymmetry. So my co-host Clay Fink interviewed Gottem Bade on TIP 583, which I'll have linked in the show notes. And he just did a fantastic job of introducing this concept to me. Most investors are very familiar with the concept of compounding. Okay. If a business compounds at 26 % per year, it doubles in about three years. If it continues compounding this rate, you get a punter-bagger in about 20 years. That's great, but where the real power comes into play is in the differences between the upside and downside in a more realistic scenario, such as your portfolio.

41:42Preston Pysh:So let's say we have two stocks that we own for$100 each. One compounds at 26 % and the other one compounds at negative 26%. So the natural belief here is that they end up canceling each other out as long as they continue to compound at those rates. After one year, the winner is worth about$126 and the loser is down to$74. Add them up and we're still at$200 right where we started. But what happens 10 years down the road? Are we still at that same$200 number? The kegger is not zero, not even close. It's actually around 18%. Yes, one stock essentially is a zero, but the gains on the winners are so high that they easily make up for the loss, delivering strong overall returns despite that zero.

42:22Preston Pysh:Gottem's conclusion on the real power of compounding was that he could be wrong 50 % of the time and still make a great return. Let's take a quick break and hear from today's sponsors. Spending my days digging through the financials of the world's best businesses, and one thing becomes obvious fast. The companies that win are the ones that can actually see what's happening inside their own operation. And these days, they say that every day your business is late to AI, you fall two days behind. The competition is only moving faster, but fortunately, there's NetSuite Next. You probably already know NetSuite, the AI-powered business management suite that securely connects all your data.

42:58It's a unified suite that brings your financials, inventory, commerce, HR, and CRM into a single source of truth, trusted by over 43 ,000 customers. NetSuite Next is the next huge leap in how business gets done. Because now AI is built into everything you do. It's customized for a wide range of industries, so it supports the way your business actually works. Whether your company earns millions or even hundreds of millions, it's time for NetSuite Next, where your business meets AI. And look, if I were running a growing company and needed to get my financials and operations onto one platform, NetSuite is exactly what I'd use.

43:35For the first time ever, you can try NetSuite Next for free. If your revenues are at least in the seven figures, go to netsuite.ai.tip. Built for every industry, ready for every boardroom. That's netsuite.ai.tip.

44:16right from your phone. You can get started with just$100 and jump into the action. See something interesting? Once your account is open, you can trade it in just a couple of clicks. And if you're not quite ready yet, you can practice with a free demo account. No risk, no pressure. With 20 years of experience, Plus 500 makes trading more accessible than ever. Check it out at Plus500.com. Trading and futures involves risks of loss and is not suitable for everyone, not all applicants will qualify. Plus 500. It's trading with a plus. Before I joined the Investors Podcast, every what if you can imagine was running through my head.

44:51What if I'm not the right fit for this audience? What if I freeze up on camera? What if I can't keep up with the level of analysis people here expect? Well, betting on myself anyways turned out to be one of the best decisions I've ever made. And I'll tell you, those early days would have been a lot less stressful with a partner in my corner the way Shopify is, the businesses that work with them. Shopify is the e-commerce platform behind millions of businesses around the world, and 10 % of all e-commerce in the US, from household names like Allbirds and Gymshark to brands just getting off the ground.

45:25You get started with their design studio, hundreds of ready-to-use templates that help you build a beautiful online store that actually matches your brand with no design background required. It's also packed with AI tools that handle the work you didn't realize you'd have to do, writing product descriptions, headlines, even enhancing your product photography. And what if you get stuck? Well, Shopify is around 24-7 with award-winning customer support to walk you through it. It's time to turn those what-ifs into with Shopify today. Sign up for your$1 per month trial at shopify.com slash T-I-P. Go to shopify.com slash T-I-P.

46:03That's shopify.com slash T-I-P. All right, back to the show.

46:11Preston Pysh:So to show what this looks like in a more realistic portfolio with more than two positions, you can use something called contribution analysis. This shows which stocks make up the most of your returns. When Gottem did this during his interview with Clay, he noted that out of the 23 positions, about 80 % of his returns came from only four stocks. I have just checked the contribution analysis for my own portfolio year to date. So my top four positions year to date make up about 53 % of my gains. And two of those positions make up 42 % of my portfolio's returns. This takes into account the 19 positions that I've owned this year.

46:45Preston Pysh:So when I extended that timeframe from year to date to the inception of my portfolio, about 57 % of my returns have come from only five businesses. I also went ahead and looked at my bottom five positions. They contributed negative 17%. So here you can quite easily see the differences between convex and concave compounding. My top winners have easily surpassed my top losers, allowing me to continue to compound my money. So what this talk of convexity and concavity in compounding is really discussing something called power laws. So a power law describes a distribution in which a small number of outcomes accounts for a large proportion of the overall result.

47:21Preston Pysh:So a business itself can be governed by power laws. One product or one new service can easily catapult a business into the stratosphere. When you look at a business, the worst it can do is become completely worthless. But the best it can do is multiply its value by 10, 20, 100, or even a thousand times. So when you are viewing a business, imagination is very potent in seeing the upside optionality that a business has. And even though the likelihood of these initiatives might be very low, it can still be integral to imagine them in case you find a business that can truly scale quickly. If you don't have imagination to understand power laws and imagine a much brighter future, you're likely to dump a stock that could offer you life-changing wealth.

48:00Preston Pysh:Power laws also govern portfolio diversification. If an investor rebalances their portfolio regularly to cap the concentration of a position, this can be seen as a highly intelligent risk mitigation. But it can also be seen as poor capital allocation. In my opinion, businesses that follow power laws inside of my portfolio should be the biggest positions. And removing parts of them is a mistake because I'm really just eliminating the possibility of maximizing future returns. Is this greed? Maybe, but I believe it to be control greed, which I think is necessary to generate wealth. Power laws can explain why using averages from competitors can be misleading.

48:36Preston Pysh:When you're looking at the average business in an industry, there's no such thing as an industry that's full of competitors with asymmetric upside. So there's always going to be losers because most of the time, capitalism is really a zero-sum game. I've seen this with some businesses that are growing at very high rates and investors use growth rates in the future that are more in line with those of their competitors. So let's imagine now that it's 2015 and you're considering Shopify as a potential investment. You look at Shopify's revenue growth and you see that it's miles ahead of its competition.

49:05Preston Pysh:Business and its industries are growing at an average rate of 20 % per annum. So you assume that Shopify is just gonna decrease its revenue growth to that number along with the rest of the industry. You buy it and end up selling out in about two quarters because you believe Shopify will experience a massive reduction in revenue growth and you take a double-digit profit. Here's the problem with this way of thinking. Shopify just wasn't another competitor. It was an emerging power law winner. It was in the midst of improving network effects, building an ecosystem, capturing the top decile of merchants, and improving its partner flywheel.

49:39Preston Pysh:Competitors just did not experience these same advantages. The average competitor was dominated by losers with stagnant platforms, commoditized products, a declining customer base and eroding market shares. So instead of Shopify regressing to the mean, another concept that we'll cover shortly, compounded revenue at a 50 % kegger, it grew its gross merchandise volume by a kegger of 63 % and became a category defining platform with incredibly deep network effects. So nothing about this outcome can be inferred from examining its competitors' averages. Where this is most important is inside of your own portfolio.

50:15Preston Pysh:These outliers might already be there. So don't handicap yourself by selling your best positions and replacing with something that is merely average. But here's the thing about compounding. It sounds smooth on paper. It's this beautiful exponential curve that just quietly builds wealth in the background. But in the real world, compounding doesn't happen in this kind of clean and predictable environment. It occurs in a world that's just full of surprises, shocks, and uncertainty. And really the key point here is that compounding is a reward for surviving randomness. Those significant outliers that I mentioned earlier, you know, the few positions that drive most of the returns are great examples.

50:53Preston Pysh:No one can really predict ahead of time what will become the power law winner. That outcome only becomes apparent after randomness really kind of unfolds. So as we move forward, it's worth asking, what kind of world does compounding actually live in? It's not a linear world. It's a probabilistic one, governed by factors such as chance, variance, and unexpected outcomes. And this brings us to our next mental model, which is randomness. Unless you understand randomness that underlies compounding, you can't fully appreciate how compounding truly works. Now, randomness should not be confused with chaos.

51:28Preston Pysh:Instead, it's a lack of predictable patterns in individual events. You may be able to get a grasp on the distribution of outcomes, but you can never predict any single outcome with a high degree of certainty. We love to see patterns even when they don't exist, and randomness can often fool us into believing that certainty is available in a truly uncertain world. Now, it's essential to recognize that a well-designed process can still yield a poor outcome, and a poorly executed process can still produce a great outcome. This helps explain why short-term performance tells you almost nothing. When you think about it, daily stock prices, quarterly earnings, analyst expectations, and market sentiment are all just forms of noise.

52:06Preston Pysh:In the short term, sure, randomness completely drowns out fundamentals. But in the long term, fundamentals drown out randomness. So I just spent some time discussing power laws and randomness plays well with how power laws function. Since we can never be 100 % certain about which business that we own is going to have the best upside potential, identifying our big winners is kind of unknowable up front. But we do know the distribution of returns is knowable. I shared my own and got them Bade's numbers. And my guess is that other investors probably have pretty similar distributions. If we understand how randomness will work within our portfolio, we should be prepared for a variety of scenarios allowing us to take advantage of randomness.

52:44Preston Pysh:This means that when we are actively investing, we need to expose ourselves to randomness while simultaneously protecting ourselves from its downside. For me, this comes down to conviction and position sizing. I've discussed certainty a lot today. And I think that if you are more certain on a specific position, the distribution of outcomes becomes narrower. All that bad stuff that maybe you envision could happen to derail the business becomes less and less likely to occur, leaving more upside and less downside. To me, taking money out of a position that you think is de-risked and allocating it back into a business that carries more risk is just nonsensical, but that's what many investors do anyway.

53:18Preston Pysh:So if you accept randomness into your portfolio construction, you need to ensure that luck has a place to show up. This means a few things, no capping winners, limiting downside through position sizing and using margins of safety when evaluating businesses, and then just ensuring that your winners can actually make a meaningful impact. That means if you're over-diversified, chances are that one winner isn't going to make much of a difference to your overall performance. Since humans love pattern recognition, it also leads us to believe that we can find patterns where none exist. Market timing is a great example.

53:50Preston Pysh:If you look at something like Twitter or financial news, there's never a shortage of people forecasting macro events. And they often appear to be knowledgeable because they understand patterns in the world that correlate to past events, and then they extrapolate those patterns to the present. However, randomness causes nearly all of these macroeconomic forecasters to be wrong with their forecasts or just to be lucky when they're actually right. When we examine randomness in the world, whether that's in the past, the present, or the future, we tend to only see survivors. We fail to see the paths that weren't taken or the ones that failed miserably.

54:22Preston Pysh:Randomness also means that rare events occur relatively frequently. And if we cannot predict rare events, the making forecast becomes completely pointless. When you go back and look at COVID-19, initially nobody really thought that COVID-19 was going to have too much of an impact on the markets. Then when we got the big giant shutdowns, everybody thought that the markets would absolutely crumble. Yet in 2022, the market was up over 18%. So I mentioned earlier about process and outcomes. And the challenging part about the process is that when we reflect on it and observe all the poor outcomes that we've had, we might actually reason that our process is flawed.

55:00Preston Pysh:But a good process will always make room for the possibility of being wrong. Let's say you make an investment that has a 90 % chance of doubling your money and a 10 % chance of losing half. You should make that bet every time. And that's a perfectly fine process. But due to randomness, there's always a chance that we lose half on the bet, even though it's not a very likely outcome. So in this scenario, you might have executed your process completely flawlessly. You had excellent decision making and sound reasoning. You had perfect evaluation logic and you controlled risk and you were in a great mental space when you made the bet, but you still lost.

55:33Preston Pysh:So the lesson in this case isn't necessarily that your process is broken. It's simply that randomness reared its head and made you make a bet that didn't turn out as well as you would have liked. You can say that randomness fosters humility. You must be willing to be wrong because you will be. It's part of the investing game. Buffett has made approximately 300 to 400 investing decisions in his lifetime, and he attributes the lion's share of Berkshire's success to about a dozen of those decisions. This doesn't necessarily mean he was completely wrong on all the others, but it does mean that he had a great process and stuck to it even during times that randomness of markets showed that he might have been wrong in the short term.

56:08Preston Pysh:So speaking of being wrong. Since we must accept that we're going to be wrong pretty often, we also need to prioritize survival. And this is an area of my own investing that I've begun placing more and more emphasis on. I've had my share of a few big losers, but some were the result of a bad process and others weren't. But either way, when I look at these losers and then observe the impact they had on my portfolio performance, I can see just how important it is to not lose money. The thing with investing is that if you compound at even modest single digit rates over multiple decades, you'll end up with a pretty nice pile of money.

56:39Preston Pysh:And instead of focusing purely on optimizing how large that pile of cash will be decades down the road, some investors might argue that you should probably optimize just to make sure you end up at the finish line. Too many investors disappear from the market because they blow up their portfolio and then they swear off investing forever, putting a prompt end to their chances of compounding. Now, the easiest way to avoid blowing up your portfolio is to use inversion and avoid common mistakes that can lead to portfolio destruction. The easiest thing to avoid is using margin. Borrowing money is an easy way to lose everything.

57:09Preston Pysh:If your bet goes against you and you have to pay back the lender, your portfolio can drop to zero. Now, another way to avoid destroying your portfolio is to just avoid shorting. So shorting does not offer the same asymmetry that going long offers. When you short, the most you make is a double and the most you lose is everything. When you go long, yes, you can lose everything, but your upside is limitless. Next is overconcentration. Look, I love being a concentrated investor, yet I've never put in more than 15 % of my capital by cost basis into one position. And as I gain more and more experience, it becomes increasingly unlikely that I will ever go much over 10%.

57:45Preston Pysh:If you had, let's say, 100 % of your capital in one position and it goes to zero, your compounding is over. But if you'd had only 10 % in that position and it still goes to zero, That sucks, yeah, but at least you have 90 % of your capital left that's either working for you now or ready to be deployed. And last year's market timing. So as I've already discussed, randomness renders market forecasting completely ineffective. Many investors attempt to participate in it, for instance, by selling when the market seems expensive, thereby missing out on many of the gains and ultimately buying back into the market at elevated prices.

58:17Preston Pysh:So when you sell a position at a loss, then repurchase it for more, that's just a recipe for failure. The interesting thing about randomness is that it fools us in the short term into making subpar decisions based on a multitude of noise that surrounds us. Most businesses that outperform do so for a pretty short time. Then once their competitors learn more about what they're doing, they attempt to replicate it. And if they do this successfully, a short-term outperformer will just simply fall from grace. This is an excellent example of regression to the mean. Here's what Parrish writes in Volume 3 of The Great Menomodels.

58:47Preston Pysh:Luck is random, so outlier results with luck components are probably followed by more moderate ones. This is regression to the mean. So the concept originated with Francis Galton in the late 19th century. Galton had been researching the heights of parents and their offspring, and he found that unusually tall or short parents tended to have children of a more average height. A set of tall parents would have kids that tended to be shorter than them, and a set of short parents would have children that tended to be taller than them. Galton argued that if regression to the mean concept were absent, humans and other organisms would be predominantly comprised of giants and dwarves, which obviously the data does not support.

59:23Preston Pysh:Now, another way of looking at this is through a sports analogy. I'm a big time NBA and basketball fan. And in 1985, a fascinating study was conducted on the NBA by three researchers, including the renowned Amos Tversky. In basketball, players are often thought to get something called the hot hand, where it doesn't seem like they can miss a shot. Is this true or is it simply a product of luck? The research concluded that it was actually the latter. A good player will eventually have multiple shots drop consecutively. And since NBA players are among the most incredible humans on earth that shooting a basketball, chances are very high that they're going to experience a hot streak at some point in their career or even in a given year.

59:59Preston Pysh:So someone exceptional shooting such as Steph Curry is probably going to have many hot streaks. But in the end, their shooting is just going to regress to the mean. For someone like Steph Curry, who has shot a career average of 42 % from the three point line, it means that on average, he's going to hit about four out of every 10 of his three pointers. But due to a mixture of skill and luck, he may hit four straight shots during a game, then miss the following six. While hitting those four shots, it gives the mirage that he's on a hot streak. But over the long term, he's just going to regress to the mean of his own career average.

1:00:27Preston Pysh:One of my favorite stories of regression to the mean is from one of my guests, Scott Barbie, who I interviewed on TIP 651, which I'll link to in the show notes. Scott went through some very challenging times during the great financial crisis. His AGEs fund had a 72 % drawdown between 2007 and 2009. If Scott had succumbed to this extreme event, he might have just called it quits and returned his investors' capital or maybe just drastically changed his strategy. After all, he was feeling the heat of Wall Street and from his investors, who probably weren't very happy with that big of a drawdown. But Barbie had a system that he knew worked.

1:01:00Preston Pysh:And even though the market didn't like his decisions at the time, he simply felt that his portfolio was now de-risked and the upside was even greater than it had ever been before. So instead of giving up or rapidly changing his strategy, he just stuck to his guns. He never used these words, but when I interviewed him, I think he was essentially taking advantage of positive regression. So far, when discussing regression to the mean, I've talked about how positive events are mostly a product of luck, and once that luck runs out, the results will generally move towards becoming more average. Well, for an investor like Scott, who has consistently beaten the market, he thought that an adverse event would eventually pass.

1:01:33Preston Pysh:And once it did, he was likely to be highly rewarded. And that's precisely what happened. The fund bottomed in 2009, then took part in just a face ripping rally on its value place. Aegis ended 2009 up 91%, getting Scott onto the front page of the Wall Street Journal, all because Scott was just patient enough to take advantage of regression to the mean. So what are the key lessons that we can learn from regression to the mean to help us become better investors today? There's four that I'd like to discuss in some detail here. The first is that extreme outperformers aren't repeatable. I review several fund letters pretty regularly.

1:02:05Preston Pysh:And when you go through the annual results of these outperforming funds, you're going to notice something. The performance is pretty volatile. They may have strung together one or two good years, crushing the index, but then they'll have a year or two where maybe they fail to beat the index. However, if they have a good process, then the results will tend to regress to the mean. For an exceptional investor, it means it will be higher than the average investor, but the point here still stands. If you looked at a fund manager's performance in a given year and expect them to duplicate those returns regularly, you're going to be sorely disappointed.

1:02:36Preston Pysh:Now, the second point here is that if you misunderstand regression, you'll misdiagnose skill and luck. In Scott Barbee's story during 2007, 2008, and 2009, observers of his fund probably thought that Scott had lost his edge and his skills as a stock picker were starting to fade. This was because these observers were underestimating the effects of bad luck that had plagued Aegis Fund during those years. For investors who had faith in Scott's skills, they were very well rewarded in 2009 and onwards by simply doing nothing and allowing the forces of regression to boost the fund's performance. The third here is just to never forget base rates.

1:03:09Preston Pysh:So you can think of base rates as kind of like a global average. For instance, most people think they are good drivers. In reality, approximately 50 % of people should be above average drivers. But research done by Ola Svensson concluded that about 80 % of drivers believe themselves to be above average. So when you consider our skill levels in nearly anything that we do, we must assume that we're somewhere around average. And if you're a stock picker, you are implicitly admitting that you're probably better than average. And the only way to find out if that's true or not is to track the results that you've had over a long period of time and observe how well or poorly you do compared to a benchmark.

1:03:45Preston Pysh:In my case, I've settled on using the S &P 500 as my benchmark. And fourth, systems with high variability will show the strongest regression. Just like Scott Barbie's Aegis fund experienced those massive drawdowns, he also experienced a pretty wild ride up. If by some miracle during that same time period, his fund had only declined by let's say 5%, it's highly unlikely that he would have experienced a strong regression back to the upside in 2009. Maybe it would have been 10 % or 20%. This is an excellent piece of advice, I think, for investors that are experiencing drawdowns. If you're confident in the long-term fundamentals of your businesses and your portfolio is down substantially, then there's an excellent chance that you're in store for some positive regression in the future, as long as you can hold your businesses through the volatility that the market offers.

1:04:27Preston Pysh:So as we wrap up today's episode, I want to bring everything we've discussed together into one place. So each of these mental models on its own is powerful, but as Charlie Munger has taught us, the real strength of multidisciplinary thinking lies in layering the mental models on top of each other. When you can closely examine just how these models are interrelated to one another and work together is where you're going to reap the most significant benefits. That's where you will build the foundation of long-term thinking, intelligent investing, and a calmer, clearer way of understanding the world and markets around you.

1:04:56Preston Pysh:So we started today by discussing feedback loops. These are the invisible machines that power every system we use, whether consciously or unconsciously. Reinforcing loops create exponential growth, and balancing loops create stability. As investors, you must understand which loop you're participating in and which ones you're trying best to avoid. From there, we broke down one of my favorite mental models, kill criteria. These are the pre-commitment devices used to force your hand into intelligent action and overcome things such as noise, emotions, and narratives that can easily hijack our judgment.

1:05:27Preston Pysh:Then we discuss the cone of uncertainty, the invaluable idea from Sleep and Zakaria that reminds us that the future actually has a shape. And while we can't predict the future, we can observe whether that future is becoming less or more certain over time. Then we explore scale. Scale, as you learned, can make a business much stronger or more fragile. And these two opposites are what determine whether a company becomes a compounding machine, or if it just collapses under its own weight. Then we moved on to algorithms, those beautiful recipes that can turn a messy hodgepodge of ingredients into a predictable and consistent output.

1:05:58Preston Pysh:Algorithms interact very well with feedback loops, because if you put the right inputs into a system, you'll get exponential growth or stability that you're looking for. Then we really zoomed out. We discussed some key mental models from mathematics. We discussed the invisible forces of compounding, which can quietly build unbelievable outcomes over long periods of time, provided that we don't interrupt it. Within compounding, we discuss convexity, where a few big winners can easily compensate for the mistakes that we will inevitably make. Then we discuss power loss, which indicate that the majority of our results will stem from a very small number of key decisions.

1:06:31Preston Pysh:After that, we discuss randomness and how uncertainty is just a given in life. Since randomness is a feature of life, we must accept the role that luck plays both in our successes and in our failures. And we must understand that the key to succeeding is to simply finish the race and not attempt to take dangerous shortcuts that could easily prevent us from ever getting to the finish line. Lastly, we concluded with regression to the mean, a reminder that extremes, whether they be positive or negative, don't last forever. Life will go on as usual, and base rates act as a gravitational pull that reels in those outlier events.

1:07:03Preston Pysh:If you have the patience and risk strategy to survive extremes, chances are you're going to do very, very well. However, beneath all of these different mental models from systems and mathematics, I think lies kind of a unifying theme. And if you can use them regularly in your thinking process, you offer yourself a way not only to survive in a noisy world, but to thrive despite it. If you can build systems that help reduce the adverse effects of emotion, if you can recognize the feedback loops you're operating in, if you can embrace randomness rather being dominated by it and can commit to long-term compounding, then you put yourself into rare company, the kind of company that succeeds over a long period of time.

1:07:37Preston Pysh:Just realize that long-term success comes with its share of roadblocks. As we saw from randomness and regression to the mean, reality is just littered with extreme events that often hurt us in the short term, but act as nothing more than a tiny blip in the long term. If you set yourself up to withstand these blips and can think a little further into the future, you'll be able to avoid some of the flawed decision-making that causes most investors to get results that are far below average. Making this episode really reinforce the concept that you don't need to see the future to succeed. Your focus should be on avoiding the mistakes that will stop you from having one.

1:08:08Preston Pysh:And while you avoid failure, expose yourself to the possibility of positive surprises that may come from the most unlikely places. Thanks for spending time with me today. If even one of these models helps you make a clear decision, avoid a costly error, or see your portfolio through a smarter lens, then today was well worth it. And if you'd like to continue the conversation, please follow me on Twitter at Irrational MRKTS or connect with me on LinkedIn. Just search for Kyle Grief. I'm always open to feedback, so please feel free to share how I can make this podcast even better for you. Thanks for listening and see you next time.

1:09:07Thank you.