与 Methodical Investment 的 David Kaiser 谈规则化投资

We've got an interesting one for you today. I'm a little off because I went to the gym for lunch and the water line broke, so now I'm in my head. I'm just a nasty, nasty person because there was no shower, and I'm an unshowered host right now.

We've got a really interesting one for you today. It's David Kaiser from Methodical Investments. David runs a quantitative, rules-based firm, and I think that's going to come through in the conversation. It makes for a really interesting backdrop when you're saying, "Hey, you're coming up with all of these rules." A lot of the things I've talked about on the podcast are: How are these rules being impacted by AI? How do you think about doing something that's rules-based? If it's rules-based, can computers copy it? How do you think about evolving—or not evolving—with the times?

As I'll say on the podcast, Ben Graham, if you read The Intelligent Investor, is telling you to buy things for two-thirds of net current asset value. Well, guess what? You haven't bought anything but Chinese frauds in the past 40 years if that was the only thing you were buying. So, how do you think about maintaining a rules-based, value-based approach—almost a religion—but evolving with the times, or not evolving? I think it's a fun conversation. I try to start off with, if you are a fundamentally focused investor, what's one thing you can learn from rules-based investing? But I think it's going to really make you think about investing, sticking to principles, and everything.

This podcast is sponsored by AlphaSense. One of the hardest parts of investing is seeing what's shifting before everyone else. For decades, only the largest hedge funds could afford extensive channel research programs to spot inflection points before earnings and stay ahead of consensus. Meanwhile, smaller funds had been forced to cobble together ad hoc channel intelligence or rely on stale reports from sellside shops. But channel checks are no longer a luxury. They're becoming table stakes for the industry. The challenge has always been scale, speed, and consistency. That's where AlphaSense comes in. AlphaSense is redefining channel research. Instead of static point in-time research, AlphaSense Channel checks delivers a continuously refreshed view of demand, pricing, and competitive dynamics powered by interviews with real operators, suppliers, distributors, and channel partners across the value chain. Thousands of consistent channel conversations every month deliver comparable signals, helping investors spot inflection points weeks before they show up in earnings or consensus estimate. The best part is that these proprietary channel checks integrate directly into AlphaSense's research platform, which is trusted by 75% of the world's top hedge funds with access to over 500 million premium sources. From company filings and broker research to news, trade journals, and more than 240,000 expert call transcripts, that context turns raw signal into conviction. The first to see wins, the rest follow. Check it out for yourself at alphasense.comyavp. That's alpha-sense.comyavvp.

Quick reminder: nothing on this podcast is investing advice. You can see a full disclaimer at the end of the podcast.

With me today, I'm happy to have on, from Methodical Investments, David Kaiser. David, how's it going?

Good. Thank you for having me, Andrew.

Really excited to talk today. Before we get started, David, I'm super excited to have you on. Maybe this is a little bit of a different chat, but you run Methodical Investments. It's more of a quantitative, rules-driven firm, and I'm always fascinated by rules and quants in markets. Before I get there, why don't you give us a quick overview of Methodical, and then we can dive into the conversation?

We're fundamentally driven, data-focused, and rules-based value investors. I guess that's the best way to put it.

Cool. We'll dive into that in a second, but when I have most of the people on this podcast come and pitch single-stock, individual, qualitative-focused investments, I guess let's just start off with a headliner. If somebody's listening to this like, "Oh, we've got a quant guy on. I want individual stocks," what's just one thing that fundamental, qualitative investors—concentrated fundamental investors—could take away from rules-driven quantitative models that would improve their investing overall?

I'll start with my background: subjective, qualitative research—individual company research. That's where I come from.

Where were you before Methodical?

I was at Robotti & Company.

Cool.

Actually, I'm still at Robotti & Company.

I know. I know. Bob's a friend and a popular podcast guest. You've got to get the Robotti name when you can.

Yeah. I wouldn't be where I am now, even though this deviates somewhat, without Bob. I learned so much about fundamental, qualitative research: what drives company growth, what to look for, what causes stocks to go up, and what the important drivers are. I really got focused on—digressing a little—what those drivers are and how to exploit them in more of a rules-based arena. Do what you're comfortable with and stick with what you like. I like structure, organization, and process, and that's how I got here.

One of the things I think you can take away is that qualitative and subjective ways of investing absolutely have a place and can be excellent. Having rules and process can give you comfort, both in how it has performed over time and in an A-to-B framework. If X happens, you're doing Y. It's not about trying to figure out, with each individual company or each portfolio, however you look at it, what the next step is. I think that's a big difference with rules-based and quantitative investing: that level of comfort in knowing—and being able to communicate, if you have clients—what happens if.

I think this is going to be a recurring theme throughout the podcast. Let me riff off that for a second. You said "rules-based": if X happens, then Y happens. One of my worries as an investor—over the past 30 years, all investors have worried about when passive is going to replace it all and when computers are going to replace everything—is that the things computers and AI are best at are following rules. If X happens, do Y. It's replacing humans in all fields across the board, even before AI and software. If you have a rule that can be followed, eventually a computer will automate that.

When you say having this rules-based approach—if X happens, then Y—the first thing I worry about, both from a quantitative and qualitative perspective, is: Why isn't that something that's already being done by AI, or that's going to be done by AI in 6 months or 2 years, whenever you want? Why isn't this something that turns us into dinosaurs eventually?

I don't have an answer as to how it won't turn us into dinosaurs. My understanding—and you certainly have a better understanding of AI than I do in this arena—is that it's learning and adapting.

One of the things with having rules and being consistent, and certainly the way I look at things, is that history repeats itself. It's not about adapting all the time. It's more about being patient and understanding that there are times when what you're doing may not create the effect you want, but that it will, and having that patience and fortitude—whatever you want to call it, chutzpah—to say, "All right, I'm not adapting in terms of trying to figure out what's working today. I'm working on what has worked over time." Look at value investing for 90 years and that sort of thing.

Obviously, there are different techniques and that sort of thing, but Methodical is looking to exploit those inefficiencies. Actually, I would argue that, to some extent, AI and its inputs probably feed into more opportunities and more inequity in stock pricing.

One thing you said there: You said "not adapting." I think you used that term 3 times. I've written and talked recently about adapting. When you say "not adapting," I could see 2 schools of thought. I think of the markets as a competitive game, and in every sport I've followed, the people who don't adapt—

In the 2010s, there were people who said, "Oh, the 3-point ball—if you base a team around the 3-point ball, you could never win the championship." Those people are out of the league now. The NBA is all about the 3-point ball. If you rewound it, if you went 50 years back in the NFL and said, "Hey, we're going to run a pass-based offense," they'd say, "Offense is run-based." Those people are out of the league.

So, on the one hand, I hear "don't adapt," and I say, "Hey, are you Phil Jackson? How are those 3-pointers holding up in the 2010s, when you're about to get out of the league?" On the other hand, there's the Lindy effect. There are these overarching principles, and all the time people say, "Hey, I can get away from these principles," and the cycle—the pendulum—always swings back to them. Buying things at a discount, that's kind of the Lindy thing.

So, how do I marry those 2 in my mind, where I hear, “Don’t adapt, don’t adapt, don’t adapt,” and say, “Markets are really damn competitive”?

One more idea I’ll throw out: if you go back and read The Intelligent Investor, Benjamin Graham’s whole thing was, “Buy stuff for two-thirds of networking capital.” That’s great, but those opportunities are generally gone, and the things that are still in the market are so stressed or so likely to be frauds. There’s no obvious way to do it.

What Warren Buffett did was adapt that approach, right? He said, “Hey, let’s make it intrinsic value, and it doesn’t have to be a hard asset.” So when I hear “not adapt,” how do I know it’s the Lindy effect versus the guy who refuses to shoot 3-pointers?

Yeah. So, okay, 2 things, in reverse order. It’s not just about discount, right? As Buffett changed from what Graham did, it’s about quality too, right? It’s that balance.

Graham was kind of a quant in my mind because he was very strict on criteria, right? His criteria were so strict in terms of what you would buy a company at that he would have no opportunities today, as you say. Right.

What I’m talking about when I say not adapting is maybe an approach and the rules—not that there’s no evolution in the portfolio. I don’t know if that makes sense. In other words, we don’t buy the same exposure, the same P/Es, or metrics like that every year. We’re using the same techniques and getting different results, not just in terms of performance but in terms of the complexion of the portfolio.

So I think you’re talking about 2 different things. One is adapting the rules over time, and the other is whether the portfolio evolves even with the rules. The rules of the NBA haven’t changed, right? To give you an example, the NBA rules haven’t changed; it’s the way teams have approached the game and offenses have evolved, right?

Is the game more physical or less physical? Are teams shooting more? There are all sorts of factors, right? But the structure—the court’s the same size, the rims are the same height, and those types of things—the rules haven’t changed in that regard. I think the 3-point line came in at 23 feet, and stuff like that.

You are correct. When you said the rules have changed, I was like, “Well, the 3-point line came in in what, 40 years, right?” It’s been consistent.

We have mentioned—not the NBA rules, but the rules that Methodical follows—for a long time, and in my mind, when you say it, I’ve got an idea of what they are, but I could be completely wrong. For my listeners, why don’t we talk about what some of the rules being followed here are?

Okay. So, rules in terms of the structure of portfolio construction, and also rules in terms of risk management and holding period, things like that, right?

What are some of our core tenets? We’re looking for a balance of quality and discount—probably not in that order: discount and quality. We consistently apply the same metrics, and the way we do that is we take a step back and look at the portfolio as a whole.

One of the rules is that pretty much everything is done at the portfolio level in terms of decision-making, so to speak. Two, we only hold profitable companies. I think that’s important because we rely on metrics, and we look at the complexion of the portfolio and how it compares to our benchmark, other benchmarks, the market, whatever you want to look at.

We’re looking for a portfolio that looks better, is more discounted, and has better metrics, right? If we’re doing that, we need to look at it as a whole. By owning profitable companies only, we’re allowing for reliable information.

So, just as an example, if you look at the Russell 2100 value—or the Russell 3000, it doesn’t matter—and it says the P/E is 18, I’m just being arbitrary, the P/E is 18, but there’s a percentage of those companies that aren’t profitable. That doesn’t go into that, right? So it’s 18, but there are also unprofitable companies.

One of the rules, one of the tenets, is to own profitable companies, right, so that, A, they tend to outperform over time, and since we’re not picking individual companies in the same way as a qualitative, subjective, concentrated investor would, we’re wanting that data to be as reliable as possible. So, I think I’m getting a little off here.

No, no, that’s great. I actually had a question on data, but let me start at the smaller profit level, and then we can zoom out. You said we only buy profitable companies—

Correct.

What is a profitable company? Is it GAAP net-income profitable? Is it adding back one-time items? Is it EBITDA? How are you getting—

No, no, we use net income, and we exclude one-time items.

Okay. So let’s go to data, and then we can come back to that.

One of the things, when you say rules-based—and I understand rules-based versus quantitative are different based on what you’re saying, but they’re somewhat similar. With screeners, I always have this issue, and I think I emailed this to you. Listen, to my knowledge, Gotham ran The Little Book That Beats the Market, and they had, “Hey, you sort by your quality metric, which is your return on invested capital, and then your valuation is EV/EBIT, I think they used.”

You kind of blend the 2, and that’s how you get your blend of quality and value. Then they would have this huge asterisk that said, “Hey, we have to exclude biotech stocks because so many of the biotechs trade for way below cash. They’re too cheap; we have to exclude them.”

With yours, when you’ve got, “Hey, we look for profitable companies,” the first 2 things that jumped out to me are, A, the biotech-stock exception, right? And, B, a mining company, right? It feels like if you’re saying, “Hey, we need to buy profitable companies,” then you’re going to get a lot of mining companies when mining is really effing hot, right?

Gold is 5,000, and guess what? Gold miners are profitable and they’re thrown in there. History suggests the wrong time to buy gold miners is when gold is 5,000 and they’re trading for a 5-times price-to-earnings multiple, because it’s a super-cyclical high.

So I’d love to ask you—let’s start with the cyclical, and then we can go back to the biotech. How do you handle that cyclical process, right? It does seem like the portfolio could tilt really heavily into cyclicals at the top of the market. If you’re just like, “Hey, we only buy profitable companies cheaply,” that’s going to push a lot of cyclicals in there.

Yeah, it’s a good point. So, first of all, you talked about rules. We own quite a few companies, and we don’t have a concentrated portfolio in terms of individual equity names. So we will cluster in sectors, and we’ll invest more heavily in certain sectors than others, based on exactly what you’re talking about—the clustering, right?

But we’re not just looking at 2 metrics. We have a variety of metrics that go into the selection process, and it is really looking at a more holistic view, right? It’s not just focusing on A and B. It looks to create a portfolio in aggregate that has better valuation metrics and better quality metrics, like return on equity, right? We want that higher.

We want lower P/E, lower price-to-book, and lower EV/EBITDA, things like that, right? To do that, you’re not just taking 2 metrics, although I’d be lying if I said that Gotham wasn’t an impetus for my thinking about this, right? But I think it has to be simple but also more complex than just 2 metrics.

So do names get in that maybe shouldn’t, or they’re cyclical and that sort of thing? Can it happen? Absolutely. Do we mitigate it by spreading out the risk in terms of companies? Yes. And we rebalance. So by rebalancing consistently, if we’re wrong, we’re wrong for a relatively short period of time.

And again, there’s that balance of quality and discount. It’s not just about discount.

What sectors right now are—I'm sure over time, different sectors pop up, right? Anybody who’s running a value screen—what sectors are consistently popping up right now? What sectors are the rules and the factors leaning overweight right now? I think it’s really interesting to see what sectors are getting discounted.

Yeah. So right now we’re heavy in consumer discretionary, which is such a broad sector, right? There’s a lot of different things in there, but right now we’re long—we’re heavy consumer discretionary.

We have some energy, not as much as we’ve had in the past, but this year we’re pretty heavy in energy. Financials is up there. I think those are the biggest. I think industrials is up there too.

No, no, no, that’s fine. Let me ask you one question I like to ask every podcast guest, right? Again, normally it’s individual stocks, but the market is a competitive place. What are you seeing that the market’s missing?

Let me just ask you: the market, especially when you’re talking about applying a rules-based model—which, as you’re saying, is different from a quantitative model—is computer-based. You hire 1 computer programmer, and they can go run it. Renaissance runs with 30 people and can manage $40 billion, and they’ve got the best computers in the world, and they can generate 50% alpha forever. You’re talking rules, so it can handle a lot more money.

It's a lot more scalable. What are you seeing as your competitive edge that allows you to compete in the market against the market in general, but also all these quant models and quant-specific firms that are trying to compete here?

Yeah. I think, going back to what I talked about, which is my background, I didn't come at it from a scientific view. I came at it from a fundamental, qualitative view. That's my background. I think that gives me a little bit of a unique perspective in terms of competing.

I go back to—and I feel like I'm beating a dead horse—the consistency, not constantly or even frequently changing your approach, having that patience, and having the willingness to have the confidence that the market will give you an opportunity and you'll be able to profit from that opportunity.

Okay. So you've got a rules-based system, and it's heavily value-based, right? It leans toward value companies.

Very much so. Yes.

The past, let's call it 10, but I think it's been more like 15 years, growth has stomped value, right?

Absolutely.

And you'll hear a lot of people—and it's not just quants, right? You read the Einhorn letter, and he'll say, “Broken market, passive flows,” all this sort of stuff.

But I guess at what point—you said the confidence to stick to the rules—at what point do you look and say, “Hey, instead of—what's the Simpsons thing?” It's like, “Is it me?” And then he said, “No, it must be the kids, right? It can't be me.” At what point do you look in the mirror and say, “Oh, it's not the market. It's me?” How do you deviate? It's kind of, “Hey, I'm insane. The definition of insanity is doing the same thing over and over again: 15 years of growth beating value, versus, no, I'm sticking to the rules.”

So, it's a great question. I don't have a clear answer, right? I'm stubborn in terms of process. What is my breaking point, if that's what you're asking? I don't know. I'm certainly not there yet.

I think historical data—and talking about data—supports the unsustainability of what's happening now, that people are paying unreasonable prices for quality at this point. I have more confidence today than I probably did 3 years ago, just because of where things have progressed in terms of valuation, in terms of concentration, and in terms of how much people are betting on the future and not paying attention to what's going on today as much, in my opinion.

We are taping February 3, 2026. This is like a borderline Black Monday for payments and software. I'm curious: have payments and software started popping up in your models recently?

No, not really. We have very little exposure to IT and things like that. Also, we rebalance in January after tax-loss selling. There's an added bump in terms of cheap things getting cheaper—discounted companies getting more discounted. So, what's happened over the past month doesn't really affect so much what we're doing.

Let me ask about corporate governance. Corporate governance has been a big focus for me. A lot of the stocks I know that are the absolute cheapest are that way because of corporate governance, right? So, you've got a great business, great asset value, all this sort of stuff, and the CEO just seems determined to light the money on fire through dumb acquisitions or—

Pocket everything for himself.

And one place I could see a rules-based model really failing is accounting for—you know, it's very difficult to read a 10-K or 10-Q or proxy and say, “Oh, this CEO is going to take everything for himself.” But it's pretty easy if you're an investor and you read 2 conference calls to say, “Oh my God, I'm swimming with sharks here.”

How do you account for corporate governance when you're running this? And how do you not just end up in 15 different controlled companies that look very cheap, where the CEO is going to pay themselves $50 million per year for all time and shareholders will take what they can get?

Yeah. So, we don't have a specific fail-safe for corporate governance. You talked about really cheap, right? One of the things we do—and I guess we're going around a little bit with the rules—is we're not buying the cheapest, right?

When we look at the data, we're removing outliers. The things that are really cheap, in whatever metric you want to talk about—P/E, price-to-book, et cetera—they probably are for a reason. Also, that combination of metrics is important, right? It's not just that it's the absolute, or even one of the cheapest, P/E companies. That's not enough.

In terms of corporate governance, it's not something that we apply. I think, theoretically and in practice, we're applying kind of the law of large numbers. Even if you take the market and you were to only buy profitable companies, over a period of time you would outperform, right?

What you're talking about, I think, is more if you had a 15–20% position in a company. It'd be really important to know that. It's not that it's not important to me. It's that how do you consistently screen out companies that don't meet criteria that you're comfortable with? And also, how does that combat what's worked over time?

And that's the balance, right? You said, in the middle there, if you only bought companies that were profitable—

You would outperform the market over time.

And you mentioned you're referring to GAAP profitability. Yeah. Okay. So, what time period—what is the basis for saying that?

Okay. So, if you look at the S&P 600 over the past 30-some years, it's noticeably outperformed the Russell 2000. The main difference between the 2 benchmarks is profitability—this profitability requirement for the S&P 600.

There are other studies—I can't think of them off the top of my head right now—but generally speaking, if you're buying an index and you're buying profitable companies comparatively, you will outperform in the last 30 years that I've seen.

You mentioned, on the side of the discussion, rebalancing, and you mainly mentioned it, I think, as December and referring to a little bit of tax-loss harvesting. I'll just throw in: no one's a tax adviser here. Don't take tax advice, all that sort of stuff.

You mentioned rebalancing. How do you think about rebalancing when you're running this rules-based model?

You mean in terms of what the reasoning is to rebalance when I—

More—not the reasoning. I think everybody can understand the reasoning, right? You buy a company that's trading at 5 times P/E. It's a windfall; it's trading at 25 times P/E. It's probably time to sell if you're running a rules-based model.

I more meant it in terms of the timing of when you rebalance. If you and I were running a quantitative book—3,000 stocks in the Russell 3000 or whatever—we're going to be long 1,500, short 1,500, net neutral. We're going to do it on quantitative value and momentum factors. That's going to be rebalanced basically not just every day, but every minute, right?

I'm guessing you're not rebalancing every minute. How frequently are you rebalancing, and what's the thought process behind that?

Yeah. So, we do a big rebalance once a year, and that is in January. You mentioned one reason why, which is the stock goes from 5 times P/E to 25 times P/E, and it's because it's increased in price, right? But what if it does that because it qualitatively falls apart? The example you gave earlier about earnings falling apart—that kind of thing—is another reason.

I think it's a balance between giving things time to be recognized and handing off a value stock to a growth stock, while also keeping things fresh or inexpensive, discounted enough that there's a margin of safety and some downside protection—not just upside. When we looked at this, we looked at more frequent rebalancing, and it doesn't give companies enough time to come to fruition.

Why is that? Because you're running a rules-focused quantitative model, right? Why should the answer not be once a month? If Company X reports a bad quarter and it's no longer profitable, or its stock trades up, why isn't it better to kick it out and adjust more frequently? You're running a rules-based model, and you said once a year. Why is the answer not once a decade, then? Why is it not once a day?

Right. So, first of all, we've looked at this over time. It's not an arbitrary number. Second, as I mentioned, we hold profitable companies, and we review that more frequently. That's a quarterly review. We make sure that the companies are profitable in the portfolio. So, if a company goes from profitable to not profitable, it's no longer in the portfolio.

Another risk tenet is that if the portfolio gravitates to being more expensive than the benchmark for whatever reason, then we're going to be in significantly higher cash.

Why would that happen?

That would happen because of either the portfolio being up noticeably, deterioration in fundamentals, or some combination.

Okay?

Or the portfolio holds up and the market falls apart. There are multiple scenarios where that would happen.

They’re very unlikely, right? Especially since, in addition to the discount and the quality metrics we look for, we look for substantial differentiation from our benchmark where we fish. The time period, like I said, is about not holding the leash too tight but also being true to value.

You mentioned being value versus something else in quant, like momentum and that sort of thing. We want to consistently be a value book. So, yes, we rebalance, but we rebalance frequently enough that the portfolio isn’t running away.

What I mean by that is, if we held for 3 years, there’s a high probability that things have changed enough in the book that our valuation is no longer advantageous. We don’t have the leverage that we do by giving it, let’s say, a year.

Let me go back to data. I think I asked this, but I just want to make sure I’m clear on it. If I do a Yahoo Finance screen, one of the tough things I find is that I’m going to sort and say, “Hey, show me the cheapest companies on a price-to-earnings basis,” and the first 30 it shows me are unusable.

The first 10 had a one-time gain. You can edit that out. The next 10 are obvious frauds—Chinese reverse-merger frauds or something. Then the next 10 are a data error on Yahoo’s price.

Maybe the company did a reverse split and the stock hasn’t adjusted for that yet. Or, I guess it’s more likely they did a split, and it’s showing, “Hey, this company earned $100 per share,” when actually it should be $10. The stock did a split, but it shows $100.

How do you guys—doing rules-based investing across basically all the larger U.S. companies—handle data integrity?

For data integrity, we rely on Capital IQ, after testing the validity and reliability of its data over time. It’s interesting that you mentioned Yahoo, because Capital IQ feeds into Yahoo, I believe.

One of the things is to have checks and balances in the way we look at the data. We don’t just take it verbatim. You mentioned mergers and acquisitions, right? That’s something we look at, and if a company is involved in an announced transaction, it’s not something we’ll buy.

There are things like that that we look at. Again, can there be an error in the data? Yes, of course. That’s another reason why we spread out the portfolio, and if a bad apple gets in because of bad data—certainly if it’s related to profitability—it’s not going to be held long, and it should not have a significant impact on the portfolio.

You think about how this has been cleared up a little bit with bringing operating leases onto the balance sheet, which I think was about 8 years ago. One of the issues I remember Gotham used to run into was that retailers looked unbelievable because they used operating leases, and it was all off-balance-sheet.

There are other companies where they sometimes do it. Look at Facebook right now. I have a post I’ll have to do at some point, but Facebook is working with these complex JV structures that are honestly reminiscent of Enron, right? They’re working with these complex JV structures to keep their data centers—these huge data center buildouts—off their balance sheet.

I’m not accusing them of fraud. I’m just saying that literally is what Enron did as well, right? They wanted to look asset-light. You see Intel right now, Verizon, and T-Mobile doing these complicated JVs to keep these fiber buildouts off their books.

How do you look at companies that are maybe structuring things to stay off-balance-sheet to make their sales look asset-light, or where accounting rules simply make them look more asset-light? How do you think about those types of issues?

That’s a great question. The way we combat that is a combination of not relying on any one metric, like price-to-book, and spreading things out. There are going to be times when a company gets in the portfolio either because there’s a profitability error—not necessarily an error in the data, but a one-time item that wasn’t scrubbed out—or because of something in price-to-book, an offshoot, something like that.

We’re not immune, but it’s not a frequent phenomenon either. It’s a great question, and one of the things I’m thinking about as you’re asking a lot of these questions is that there are a lot of things to account for, but it’s very difficult to create a perfect system.

One of the ways I think about this, when I talk about the holistic approach and looking at value from different angles to create a portfolio with a certain complexion, is that it’s important to recognize that we’re human and fallible. I’m Jewish, and we talk about a Hebrew word I’m not remembering at the moment that means “missing the mark.” It means you’re human, you’re fallible, you’re going to make mistakes, and you want to do the best you can to come as close to hitting that mark every year and being the best person you can be.

I try to apply that thinking to how I look at the portfolio. I want a good portfolio. If I shoot for the bullseye and try to create a perfect portfolio, I’m going to miss things too. Even qualitatively, some of the names that have worked out in the past, if I looked at them, I would have said, “I don’t know,” and I would have had other thoughts.

By sticking to what works and putting my—pardon the expression—faith in the data over my own expertise, we’re able to do this. You’re asking great questions, and a lot of those things, if I were doing qualitative research, are all things you would check off. They’re all the boxes.

But when you’re buying, let’s say, 50 to 80 companies in a portfolio, it’s not that those things aren’t important. How do you account for all those things and then not exclude things that you would want in the portfolio? I think that’s where the complexity—the problem—occurs with coming up with rules.

That’s what we’re talking about, right? How do you consistently find companies that have the potential to pop and grow and make money for your portfolio? How do you do that consistently? And I’m sorry.

No, no, that’s great. A lot of it comes back to the rules, right? So how do you come up with the rules?

The basis for all of this is things that investors have looked at over time. The rules are mostly based on testing to see what works, but a lot of it has to do with—I don’t want to say common sense, but things that value investors would think about.

If my portfolio is upside down, in the sense that it’s not discounted and doesn’t create the discount that I want, that’s not an exposure I want. Things like that. They’re not out of left field, I guess, is the best way to put it.

You talked about some of the data and some of the ways you look at the data, and whether I have unique data and that sort of thing. I don’t. It’s how I look at it that differentiates me. It’s not that the data is unique or that I have some crazy rule, like, “If Company A does X, Y, and Z, and it’s the third Sunday of the month and falls on whatever,” I’m not doing that.

A lot of people are familiar with the Buffett indicator, right? Buffett used to say, “Hey, when the stock market’s value trades in excess of U.S. GDP, it’s overvalued.” You would hear this time and time again from investors.

If you followed that rule, the only time in the past 15 years you would have been able to buy stocks was at the absolute depths of the global financial crisis. I could make 2 arguments. One is, “Hey, we need to stick to this rule, stay in cash, and cash is king. One day we will get a shot. One day there will be another correction.”

But I think another way of thinking about it would be, “Hey, if you’ve got a buy signal that says one time in the past 30 years it was good to buy, the buy signal is outdated now. It’s not a good buy signal.”

Why is the Buffett rule outdated? In part, the U.S. used to be a place where the public companies were GM and Ford, and they were selling all their cars in the U.S., so GDP was a good tracker. Now it’s Apple and Facebook. It’s global.

How do you think about the rules? I’m guessing some of the rules are, as you said, profitable, trading for a low price-to-earnings ratio, and trading for a good ROE. When you say those rules and that you backtest them, was the backtest 100 years?

How do we know the next 10 years aren’t different from the last 100 years? The rules from the last 100 years could become the Buffett indicators, right? I could imagine a world where, when I say low price-to-earnings and good ROE, that’s probably going to push you a little bit heavier into banks a lot of the time, which tend to trade for low price-to-book and good ROEs.

That’s historically been a great exposure. But banks have all this fintech risk, right? How do I think about the rules evolving and sticking to them? I think I’ve thrown a lot out there. I don’t know how to quite bring it together, but I’m sure people can understand where I’m going based on the Buffett point.

First of all, we are looking for opportunity in every market. We’re looking for relative opportunity. We don’t have—and I should be clear about this when we talk about rules and wanting discounted P/E and that sort of thing—a cutoff, like 5 times earnings. I’m just being arbitrary, right? We’re taking what the market gives us at any given point in time. When we rebalance in January, it’s, “What are the opportunities today?” So that’s number 1.

You mentioned financials, and earlier you mentioned biotech. Biotech is not something we invest in. There’s too much variability in earnings, right? You can have a drug that hits, and it’s going to go away in 2 months, that type of thing. It’s going to go generic or whatever it is. So that’s 1. Financials are another area where we limit exposure, and we do that because exactly what you said: when you’re looking for low P/E or low price-to-book, high return on equity, and things like that, you can get a lot of those, and those are not necessarily the companies that are going to drive performance over time, right? So we limit exposure. We don’t eliminate exposure to financials, but we do limit it for exactly that reason.

That’s 2 areas of the market, right? So I guess you just said, “Hey, we don’t do biotech.” And I’m with you, right? Biotech is really fucking hard, and you’ve got drug cliffs, up-and-down coin flips, and safety approvals. Safety approvals are the one where it’s like, “Hey, you can have everything right and then”—sorry, not safety approvals, safety issues. You’ve got everything right, you’ve done all the analysis, and out of nowhere it could be, “Hey, this drug caused liver failure.” How are you going to catch that—as an individual investor, let alone as a big investor? You just don’t know.

I understand that’s a risk, but those are truly out of left field. You do that, and the drug goes to zero anyway. But you said, “Hey, we don’t do biotech, and we systematically limit our exposure to financials.” I’m sure part of that is, hey, these things aren’t going to drive the beta up. Financials are one of those funny industries where Lehman Brothers looked really cheap on Friday in September 2008, and then on Monday it was zero. So you’ve got those risks, too.

That’s 2 pretty big things where you’re stepping in and imposing limits and systematically overriding the rules. How do you think about that versus, hey, maybe if the system says we should have 15 banks and 24 biotechs right now, maybe we should have less than that?

Right. Let me go back 1 second. We don’t have a hard limit on our exposure to financials. When we’re creating the portfolio, we look at it with financials, and then we do a second run actually eliminating financials.

Okay.

Then there’s a lot of redundancy. I’m not going to get into the whole process of putting the portfolio together, but we’re open to financials on the first run. On the second run, we’re limiting them. We tend to have, exactly like we just talked about, high exposure to financials—more so than would help us in terms of performance over time—and that’s the answer.

What I’m doing, I don’t think, is the most complex thing in the world. It’s kind of how I look at it and the consistency with which I look at it. Again, we’re using the same metrics everyone else is, but it’s—sorry, I lost your train of thought again.

That’s great. One more question, and this has been at the heart of it. We’ve touched on some things, but it’s just one that keeps popping up: melting ice cubes. These are investors’ least favorite things. The one that I think of right now is Disney, until 2016, when it comes out and says, “Hey, we’re losing ESPN subscribers,” if I remember correctly.

Linear cable channels are, as many people said, probably the best business in the world. I kind of disagree with them in some ways, but ESPN especially: you’ve got scale because of that; you can afford to pay for the NFL—no one else can; you get the ads; if a cable channel tries to kick you, all their subscribers are going to churn, or they’re going to be so angry; you’ve got great pricing power; you’ve got this huge network effect, huge scale benefits, all this sort of stuff. Until 2016, it’s the best business ever, with very low capex. After 2016, it’s death.

Go pull up the chart, and forget the tiny media companies like AMC Networks—you can look at them. All the regional sports networks, in 2016, were great; by 2020, they were all going bankrupt left and right. Look at the chart of Disney. Media companies are the shining example of melting ice cubes. They’re near and dear to my heart, and they’re also one that value-based, rules-based models tended to love, right? Again, capital-light. After 2016, a lot of them started trading really damn cheap, and they’re just cheap—down, down, down.

We can probably think of other examples of melting ice cubes, but how do you avoid getting a portfolio that, because you’re using trailing numbers, looks great on those numbers and is just buying left and right—melting ice cube, melting ice cube, melting ice cube? I understand some of that is, hey, melting ice cubes tend to be overly discounted, but I would just point to the media example and say, hey, if you did these over the past 5 years, you’re no longer in business because it blew up the price of AMC Networks.

We tend to have fairly frequent turnover. I think—we’re not even going to compare it—but we generally have fairly substantial turnover when we do a rebalance. So the idea of a company that is not executing the way it should in terms of quality metrics, as opposed to just discounted metrics, staying in the portfolio over a long period of time is unlikely, right?

Also, it’s the combination of metrics we look at. There can absolutely be melting ice cubes. It can happen, but we also talked about the yearly rebalancing, right? The focus isn’t a year; that’s just kind of the grander scheme of how we rebalance. The idea that we try to create a perfect mechanism for 1 year is not how we’re looking at it.

We’re looking at consistency. Maybe to use a sports analogy—baseball—we’re consistently looking for fastballs, right? If we get a curve, we’re going to hold off, right? The long-term play is to create alpha over a period of time. Again, because of the way we spread out exposure in terms of companies and because of the metrics we use, is it something that frequently happens, that we get a plethora of melting ice cubes? No.

Last question. You know, a rules-based model, a more quantitative model—it seems like there’s lots of opportunity for AI in the research process, in the fundamentals. How are you thinking about and using AI, not in terms of competition or as a risk to the underlying companies, but just in terms of assisting or helping with portfolio construction, with the rules, with the backtesting, whatever it is?

At this point, we’re not really using AI. I think AI is a great tool, as you said, subjectively. If you’re going through 10 years of 10-Ks and trying to find a trend, or things that are talked about and written about consistently, things like that, I think it’s really important.

Is it something that we’ve thought about? Yes, and it’s something that we could potentially implement. But again, it goes back to the fact that we believe in the opportunities that we’re finding and will continue to find. So we’re not looking to evolve. I think that’s really where AI helps, right? AI helps you evolve a process.

If everyone else is using AI and they evolve, I have the risk of being a dinosaur, but I also have the risk of really being differentiated and sticking with something that will continue to work.

No, AI is something I’ve thought a lot about. I did a post on Market Your Podcast, and AI is funny because when I would say AI was a risk, a lot of people would email me and say, “Hey, when you talk about it as a risk to investors, you forget that investors can use it.” Buffett, of course—somebody sent it to me—had a great quote for this. It’s the standing-on-your-tiptoes-at-the-parade idea: if you do it, everyone else does it, so it’s a counter.

I do think what you said is interesting. Someone else sent me this: a lot of times, what happens is that when these games get so optimized, right? The best example—it’s a difficult one—is daily fantasy sports, where everyone started using optimizers. They would optimize your lineup if you were in a competitive contest.

When everyone used them, the edge actually went to people who didn’t use the optimizers. They could build good portfolios—in this case, good teams—but they weren’t optimized. If everyone buys Albert Pujols and Chase Utley, and every team has them, well, if you don’t have one of them, you actually have a huge edge and huge variance.

It’s just interesting, because I guess where I’m coming with this is, hey, if everyone else is using AI, they’re running into the tiptoes problem, and maybe there is alpha on the edges of an old systematic process. I’m not 100% sure, but that’s one of the things I’m not 100% sure about.

Yeah. I do think the opportunity exists, as you said, on the fringes now, right? If everyone is using AI—excuse me, adapting, learning, trying to keep up with the Joneses—and even, like you talked about, 15 years in the history of the stock market is not a long period, right? It’s a substantial, noticeable period, but it’s not huge, right? These things do tend to be cyclical in terms of what drives market performance.

And right now, safety isn’t where it’s at. It’s FOMO—fear of missing out. And, in my opinion, that’s what’s pushing the market.

The idea that there will be a return to caring about where you are today and what your safety level is, relative to what you can potentially make—I think the idea that that will not happen is foreign to me. Can I tell you when? No. But I banked on it, right?

I think that’s the differentiator and the fringe you talk about. Yeah, I guess I’m on the fringe now, right? For me, it’s value, being consistent in what I do, and what’s worked over time. Does it work in the future? I don’t know. But I believe it does, and I’m betting on it.

Last question. I think a lot of this comes up as: Hey, I believe in value. I almost have a religious-type belief in value, in these metrics, but we mentioned backtesting yourself a few times. I’m curious, when you think about backtesting—and this was sparked by your saying that 15 years is not a long time in the stock market, which I agree with—how long do you think about backtesting an idea as a concept when you’re looking at it?

Ideally, I would say several market cycles—periods where, when I say cycles, I don’t just mean ups and downs in the market, but different techniques and approaches, growth, value, whatever you want to say, have excelled or worked over time. I think you’d have to look at it this way: If you said the past 15 years were more growth-oriented, you’d have to go back a lot further to times when value was more in favor, right?

I guess the reason I ask is, again, I think about this in evolution. Let’s just hypothetically say a lot of the value outperformance comes from 1980 to 2000, and a lot of the value outperformance comes from 1940 to 1960. Right? So I just use 20-year cycles.

Well, 1940 to 1960, I’d tell you, get out of here. Buffett comes along in the end, but you’re buying completely different things. The market is much less efficient. There are lots of pink sheets. It’s just crazy out there, right?

So if you’re doing a backtest and saying, “Hey, I’m backtesting to 1940, 1950,” I say, “I don’t think that’s relevant.” If you’re doing the backtest in the ’80s and 2000s, now we’re talking about a more modern market, but computers still aren’t around there. There’s still a lot. I just wonder: When you’re thinking about backtesting, forgetting the market cycles, how are you thinking about market evolution?

I’m trying to remember the date. It was in the ’90s when everyone had to report digitally, right? What was that?

I think that’s the SEC.

Right. So I would think that would be a fairly good period to be looking at, because you have enough information and the data would be whole, so to speak.

Cool. Okay, this has been fun. David, where can people find you if they kind of want to learn a little bit more? Methodicalinvestments.com and, uh, davidthodicalinvestments.com is my email and feel free to reach out anytime.

Perfect. David Kaiser, Methodical Investments, this has been great. Thanks so much for coming on.

Thanks, Andrew. I really appreciate it.

A quick disclaimer, nothing on this podcast should be considered investment advice. Guests or the hosts may have positions in any of the stocks mentioned during this podcast. Please do your own work and consult a financial adviser. Thanks.