Episode #502: Mikhail Samonov – Lessons from Two Centuries of Financial History
Guest: Mikhail Samonov is the CEO of Two Centuries Investments, which was established to create long term wealth for investors through its blend of innovative and behaviorally focused investment strategies.
Date Recorded: 9/20/2023 | Run-Time: 1:11:28 
Summary: In today’s episode, Mikhail walks through what led him to focus so much on ‘long history’ in his research. He shares lessons learned from studying two centuries of financial returns, including momentum and asset allocation. He also spends time explaining how he’s using AI to study intangible value and company cultures.
Sponsor: YCharts enables financial advisors to make smarter investment decisions and better communicate with clients. YCharts offers a suite of intuitive tools, including numerous visualizations, comprehensive security screeners, portfolio construction, communication outputs, and market monitoring. To start your free trial and be sure to mention “MEB ” for 20% off your subscription, click here. (New clients only).
Comments or suggestions? Interested in sponsoring an episode? Email us [email protected]
Links from the Episode:
- 1:29 – Welcome Mikhail to the show
- 2:30 – Why such a focus on long-history?
- 3:04 – Two Centuries of Price-Return Momentum; Two Centuries of Multi-Asset Momentum
- 12:41 – Investors often underestimate risks, ignore history
- 18:11 – Investment comfort zones vary by generation
- 25:15 – Innovation and creativity are crucial for quantitative investors
- 34:26 – Applying research on intangibles using NLP
- 41:00 – CultureLine uses AI to analyze workplace culture, aiding investors and enhancing ESG models
- 45:46 – Story about Steve Jobs adding “creative” to Apple’s 10K
- 49:46 – Deep dive into asset allocation strategies and their long-term resilience; A Century of Asset Allocation Crash Risk
- 59:42 – Why investors underestimate drawdowns
- 1:03:54 – What investment belief Mikhail holds that most of his professional peers do not
- 1:07:09 – Mikhail’s most memorable investmentLearn more about Mikhail: Two Centuries; CultureLine; LinkedIn; Twitter
Transcript:
Welcome Message:
Welcome to the Meb Faber Show, where the focus is on helping you grow and preserve your wealth. Join us as we discuss the craft of investing and uncover new and profitable ideas all to help you grow wealthier and wiser. Better investing starts here.
Disclaimer:
Meb Faber is the co-founder and chief investment officer at Cambria Investment Management. Due to industry regulations, he will not discuss any of Cambria’s funds on this podcast. All opinions expressed by podcast participants are solely their own opinions and do not reflect the opinion of Cambria Investment Management or its affiliates. For more information, visit cambriainvestments.com.
Meb:
Welcome, my friends. We have a special episode today. Our guest today is Mikhail Samonov, one of my favorite investment researchers. He’s also the CEO of Two Centuries Investments, which he established to create long-term wealth for investors through its blend of innovative and behaviorally focused investing strategies. Today’s episode, Mikhail walks through what led him to focus so much on long history in his research. He shares lessons learned from studying two centuries of financial returns. That’s a long back test, including value strategies, momentum, asset allocation. He also spends time explaining how he’s using AI to study intangible value in company cultures. There’s a lot packed in here. Make sure to visit the show notes at mebfavor.com. Please enjoy this episode with Mikhail Samonov.
Mikhail, welcome to the show.
Mikhail:
Thanks for having me, I’m excited to be here.
Meb:
Where do we find you today? This might be a first for the Med Favor Show. Where in the world are you?
Mikhail:
I’m in Avignon, France, which is a little town in the south of France with my family. Decided to spend an academic year here. My wife is French. We always dreamt of doing this and this is the year.
Meb:
Amazing. When I went there with my mom and now wife, I think I gained 15 pounds on the French wines and cheeses. How are you doing? What’s the vibe like in France right now? You picked up a few kilos while you’re there?
Mikhail:
Actually, part of my personal thesis was I was going to head the other way around, French life expectancy and all. I’ve been exercising, walking a lot, moving into the healthier direction. First couple of weeks, yes, we definitely enjoyed the croissants and all that stuff, but now it’s in the more farmer’s markets and a lot of walking, bicycling. Heading in a good direction.
Meb:
I love it. Listeners, Mikhail is one of my favorite researchers and writers. He’s written a number of papers over the years that really speak to me almost like a brother from another mother because he has an appreciation for history that’s somewhat unmatched in some of the depth and link. We’re going to talk about a lot of topics today. I figured we’d start with how do you even become such a historian? You can pick the paper. We can start with momentum side, the value side, either way.
Mikhail:
Price momentum gave me a big headache. That’s why it became my first paper, and the headache was that the fundamental folks didn’t like it because it was too technical. It didn’t fit some fundamental story within the firm’s philosophy, and of course the academics hated it, specifically Eugene Fama, who everybody looked up to and looks up to and it violates every efficient market hypothesis, all three levels. And also you got to keep in mind in 2004, my quant budget was very low. The data I had was only back to 1980, which is a Compustat industrial package for those of you quant geeks who still remember. They had a short history. You test price momentum in the eighties and the nineties and it looks amazing empirically, but since 2000 and the dotcom bubble and by 2004 it started to have this nose dive-ish drawdown compared to the last 20 years.
I’m sitting there and thinking, how much do I argue to add this factor to the model? And I come up with an idea from back from the seeds of college. I got to look at more data, more history. We buy the next historical database of Compustat tag. We get this 30 years of data back to 1950s and I’m eating up those decades of back test like it’s live, it’s happening to me. It’s not some irrelevant old data. This is like I’m testing all these factors I invented on new data. Price momentum survives and has a phenomenal run during those 30 years. I convinced the team, we added it with a small weight and it was good with a small weight, and everything was super, again, not univariate, it was dynamic, contextual. It was very diversified, but as an idea it made it into the model.
Then I make it to Wharton to do my executive MBA while working. It is a cool program Wharton runs parallel to the main MBA, but of course inside I always wished, I think I did a PhD, but I didn’t want to leave my job. The first place I go at Wharton is a library instead of the business school building and start digging up all the data they have access to and I’m just like, yes. First of all, I get crisp data back to 1925. I test price momentum right away and oops, there’s this massive drawdown in price momentum right after the Great Depression, 80% drawdown. And again, consuming history as if it’s live coming at me, I don’t just ignore it. Oh well, Great Depression, who cares. It’s not going to happen again. I’m like, this is a distribution of this guy, of the sucker.
It’s good to have a small weight, but then this whole notion around skepticism, does it exist? Is the data mining still in the back of my mind, the fundamental folks right, is Eugene Fama right? And then I see Professor Siegel’s famous stocks for the long run opening chart, the 200 years of stock market outperforming bonds and gold. I’m sure most listeners have seen that kind of breathtaking compounding chart. So simple and so powerful, and the question pops into my mind right away. I’m like, where’s all the underlying data, the stock level data that makes up the equity index before 1925? Merging this sucker together took a while because only we had common names as the only common identifier between the three data sets. We used some NLP back then, some early NLP to extract these names and correlations. Anyway, we merged it together, we run the price momentum and yeah, it’s basically what I said.
It was breathtaking because I could finally show around that this is not data mining for sure. I could say though, that momentum crushed a lot. You got to be really careful. You also start to see, and this was by the way, right before March 2009 when the market turned around from the 2008 recession and momentum had another one of its really ugly crashes, second worst since the Great Depression. But at that point I was kind of feeling good because I had a very small weight, I was ready for it. A lot of it was due to dynamic beta variation of the longshore portfolios, but that’s beyond the technical details that you can maybe protect against some of that crash, but maybe not. I got fascinated that history can teach you this. The third lesson that long-term history around factor investing unfortunately confirmed for me is that that last 20 year kink that you see, the inverted hockey stick, almost opposite from my brown graph, which was flat and then became upward sloping for 300 years of GDP per capita, factor premium, have this inverted hockey stick, which is when you combine value momentum, maybe some quality and look at it over 200 years. The last 20 years starts to jump out in this unpleasant way of just a flat mean beyond the drawdowns.
You start to see really extended flatness, which gets factor investors worried. Is it too much crowding? Is it arbitrage or should you hold on no matter what? That’s another question. But those are the three observations from long history that after we finished the US stock level data tests, the very natural next place was let’s look at global equity markets, fixed income currencies. All that data existed in global financial data. That was paper number two. The paper number three was the most painful one and the least appreciated one. I’m very proud of it, but I don’t think anybody cares, which is fine with me, but it was the commodities futures because the reason I’m proud of it because we actually ended up hand collecting all that data from taking pictures of this big book, CFTC’s annual Commodity Prices book and typing all this up and as you know, futures, they don’t just have the first month, they have the second month and the third month. And to do it well, you have to roll the future. You end up collecting a lot more data than you would if it was just a stock price.
Meb:
A significant undertaking. I almost picture you at bottom of some giant library flipping through these old books, handwriting down all these things and it’s sort of a quants dream scenario where you have a magical out of samples suddenly appear. Really at this point I feel like the only out of sample is moving forward for most of us, but in your case it was extending it backwards, which is pretty awesome. Let’s stick on those couple papers real quick and then we’ll kind of jump to the value in a minute. But the interesting thing about looking at the out of sample pre-history and what you, I think, call long history is you start to see some things that rhyme, some things that stand out, but you also mentioned some kind of practical implementation concepts, the crashes, the underperformance. Give us the cliff note of what you just described on the price and momentum side. Is it something that you think here in 2023 still viable the way most people think about the academics? Or is it something that you say, okay, it is, but it’s got these two, three caveats or asterisks by it? What did it really reveal to you as the main takeaways?
Mikhail:
Yeah. A couple takeaways. With price momentum specifically, there’s a huge market beta variation happening inside a long short momentum portfolio. As momentum portfolio long short is buying winners and selling losers, what just happened to the market over the last 11 months is going to show up in your portfolio. If market was up, then you’re going to have a positive beta, long short beta. If market was down, you’re going to have a negative long short beta. And so if the market is down, you have a negative long short beta and then the market turns around, usually volatility is high and that moment, so the turnaround is quick. Momentum has a big draw-down, so that’s if you’re going to apply momentum, watch out for the beta exposure of your portfolio because it’s going to sneak up on you and the variation is really large.
It can go from 0.3 to negative 0.3. Second is diversify, diversify, diversify through innovation, through tweaks. The degree innovation depends on your process. Maybe it’s tweaks, maybe it’s completely new factors, maybe it’s dynamic contextual, everything in between, because betting on one or two of these quantitative ratios is really dangerous. You got to blend it and innovate and mix it together, look for more or more data because if you can, sometimes it just doesn’t exist. But if you can, look for the longest data, because it will show you the distribution especially of the downside. And this was a lesson that, it was a big one. Are you ready for your factor or your asset allocation, as we might talk later, or your stock to drop… Or the stock market drops 90% in the Great Depression. The US equity is a favorite, best equity market out there. Risk management, which leads to the prior point about diversification. That’s why you diversify because these things crash a lot and then sometimes they have these dry spells that last decades. Again, what’s going to drive performance during those times?
Meb:
Every investor you talk to, A, doesn’t understand or appreciate what you just said. If they do, they still think it’s not going to happen to them, right? The amount of people say, no, I allocate to this strategy. I know that it can go years of underperforming. I know it can go down, in the case of equity 80%, but I don’t think it’s going to happen, right? I know it could happen, but it’s not going to happen to me. It’s not going to happen in my future. That’s something that’s happened in the past, but we had an old podcast alum, Wes Gray had an old article called Even God Would Get Fired as An Active Manager, meaning you had perfect foresight into the factors, but you still had these gut-wrenching drawdowns and periods of underperformance relative to something. You mentioned the futures one, which you love and most of the world wasn’t as interested in it. The commodities concept. We’ve had people on this podcast full spectrum, all right, on the biggest commodity bulls, you’ve got to have it in your portfolio, huge chunk. To other people that say, look, the returns are overstated, the implementation is problematic, the indexes are, blah, blah, blah, whatever. Where do you fall in that spectrum of how to think about commodities in a traditional portfolio?
Mikhail:
Yeah. I guess I’ll give you two answers. One is the way I think about it today in my own portfolios is kind of simple as gold. And it’s a dynamic position, dynamic asset allocation where gold is part of the universe and sometimes the model buys it, sometimes it doesn’t and it’s a great inflation and disaster scenario. That’s my today’s actual implemented views. But academically speaking, or in other words, maybe if I was running a larger institutional book of money where I could do this sort of futures investing, what I basically confirmed, I can’t say discovered, but confirmed using long-term data is that, well I think it’s known that commodity spot prices are very different than commodity futures and you can’t invest in commodity spots most of the time, you end up investing in futures. When you study futures, returns, your strategies, you got to study it on futures, not on spot.
In fact, if you take price momentum and you test it on spot prices, you get an opposite result. It actually is consistently negative. And it was a surprising thing. I didn’t realize spots had that property. Other researchers have confirmed it, and spot prices go back centuries and centuries. You can get a 700-year back test of negative spot price returns if you did momentum, it’s weird, but when you shift to futures, then things become more normal. It’s a financial instrument. Futures markets for commodities were in… The first futures were in 1871. The contract was standardized, before that it was forwards. They weren’t structured, but in a structured contract, 1871 versus if you look at Bloomberg or data stream, you can’t get futures data for commodities before 1950s. I was looking at that going, ooh, yummy. I didn’t realize how painful it’s going to be to collect all that data, but it’s a good chunk of decades, and since you have to invest in futures, futures by nature roll all the time, so you have to be rebalancing once a month to get the next future, the next future.
They’re a perfect asset class for a factor exposure because you’re dynamically trading anyway. When you add momentum and value, kind of like the AQR thinking with values mean reversion in commodities and then you add the basis risk, which is a spread between is it backward dated or contango, the slope of the future’s curve. Those three factors together, if you combine them, even if you want to have a long only basket of commodities, you’re still rebalancing once a month, so your trading cost is the same. Now you’re just buying a subset that has those three factors over the century and a half. That was a great exposure to commodities. The premium is significantly higher than just kind of doing that for all the commodities together. I think it’s also getting popular and crowded, et cetera, over the last 20 years. Many banks had structured notes and structured indices based on these ideas, so there’s still some deterioration to that, recently to that premium.
I don’t even like calling it premium. I’ve spent so much time in academia, I call these things premium, but really one of my original ways to see it as I started on Wall Street was it’s all anomalies and it kind of became out of favor, that phrasing. And everything became a risk premium, but I think we can have a whole talk about that. It’s much safer as asset managers to call these things anomalies because then you set the right expectation, you watch the crowding, you keep innovating, you keep diversifying. Calling them premium is comfortable from getting the clients relaxed. Don’t worry, this thing will pay premium, but I don’t see it in the data. When you see this flattening out, confirms that things can get crowded, but from commodities, again, they obviously play a unique role, especially around inflation and supply chain problems, et cetera, but they got to be implemented, I think. You can’t do passive commodities in my view. You have to have some active approach there.
Meb:
One of the things you mentioned here that I think is a good lead in to this next topic but is threads of everything we talked about this far is what do we know? I mean if you look at, hey, we got this little 30 year period, and most investors around the world will base their investing style based on their very limited life history. I talk a lot on this podcast where my parents’ generation in the US, if you bought stocks and held them, you did amazing, right? But that’s largely because you invested the majority of your career in the eighties and nineties. There’s really been a fantastic period, whereas the generation that started investing maybe around 2000 got hit with two 50% bare markets in the US and then if you go ask somebody in Greece or Russia or China or on and on and on, they would have a very different takeaway and that’s just with equities.
I mean, goodness gracious, bonds and others have been super weird too, but looking at a data set and thinking, okay, here’s how much I can extrapolate from history. Also, I have to think about have the participants changed the markets? And what I’m leading into is for example, your value investing studies. You can talk a little bit about the value and then how you took it back because a couple of years ago, value was having one of the worst periods ever. ’99, pretty awful, but 2020, as bad or worse and trying to go through that and say, okay, well is this broken? Has it been commoditized? Talk to us a little bit about how to think about that because that’s, I think, something we all kind of struggle with.
Mikhail:
Yeah. There’s a lot in there. And you mentioned this in your earlier question a little bit also about looking at shorter history and then you just mentioned generational biases. I think investments, the biggest thing that I’ve learned, and then I’ll get to value is basically there’s some stuff that’s popular and comfortable to do, and then some stuff is unpopular and uncomfortable. And my biggest check for that, it’s actually very easy to know. I tested it when I worked with the larger teams. I would walk into my office and then kind of lean over all the cubes and state something out loud. I think I actually, for example, last one I remember in 2016 I said, “Market is going to double.” I didn’t really believe that, but I said it with full confidence. It was very awkward to say, my gut churned on me. Everybody looked up thinking I was crazy, and I realized at that moment I, along with everybody else here, have this massive bear bias in me and it’s much more comfortable to say, I think stuff is going to crash and burn and the Fed is wrong and everybody’s wrong.
And so the premier exists on the other side because whatever’s comfortable is already priced in. Even if it’s correct fundamentally, there was a recession, but it’s already priced in so you don’t make any money where the money is made or market moves in an uncomfortable direction into this true surprise, not just an easy surprise, uncomfortable. Factors like value is a good one because by 2004, value was very comfortable because it just totally did great after the dotcom. Now if you think about why value works in the first place, we can debate again, I don’t think it’s a compensation for premier, it is a risk from point of view of co-variance and you can see the volatility is a factor like an industry is a factor if you’re building a risk factor model. But in terms of why does it earn a premium, again, I keep using the word premium, positive return.
I think the original papers and we all kind of as quants believed it is because value stocks are very uncomfortable to own. Traditionally in the eighties and the nineties, they’re the duds. There was a premium that, well excess return earned from that undervaluation. But then through all the data and computing powers and all the quant papers and quants themselves and then even smart beta took it a whole new level betting on duds through value investing became very comfortable and popular. Same with momentum. Momentum holding this winner stock, you feel it’s expensive, it’s ran its course, people sell too early, but then again through all the momentum literature out there, that’s now a comfortable idea to hold it. When things get comfortable, they get overpriced or at least fairly priced and the return goes away. Value then reached its heyday by 2008 and started its drawdown since 2008. We now, when I wrote that blog on value extension, I decided not to go for full academic. It takes nine years to write an academic paper, for me at least. I just decided to put it into a blog.
Meb:
That’s the beauty of the internet, man, is that you do an academic paper, you get two people that read it, you get comments back, it’s published in two years. You do a blog post and within an hour someone is telling you why you’re an idiot and you get feedback instantaneously.
Mikhail:
[Inaudible 00:23:59].
Meb:
You could do both, of course, but if you’re looking for feedback and input, the internet and social media will certainly whip it around pretty quick.
Mikhail:
Exactly. Sorry, I just wanted to give the message out. Didn’t want to wait any… Yeah, I think by 2020, value was very painful, huge drawdown that was building for a while and then nosedive on top of that decade. It was very uncomfortable to hold onto for those who were still holding onto. But those drawdowns have happened before. If you zoom out 200 years, you see them happening. But last time you saw that happening was in 1904. You had to go back that far to see the 59% long short drawdown on the metric I was using. And my recommendation at the end of the blog to the diehard value quants out there is hold on, don’t sell now, because there’s going to be some mean reversion based on just everything I know. There’s a different question of whether for the next 20 years after the drawdown is done, that extreme state is over.
I don’t know if it ever gets back to zero, but just at least the bounce back from the extreme lows happens. What is the average mean? The slope to the factor? That’s a harder one for me because there, I do think the markets evolve if things are very comfortable, like the value factor stays in a lot of models, that’s a comfortable factor to have. Unless that changes, I don’t see that mispricing returning to value and it’s much safer to continue to invent, I think, new types of value. Some people in your podcast were talking about, very exciting to listen to other like-minded folks, and continue improving it.
Meb:
I think the challenge for a lot of people, and you alluded to this a little bit, which we can get into. I remember when I got started in the business and was looking at a lot of these multifactor models. There’s the very traditional sort of French Fama stuff, and you could build a very simple portfolio of multifactor names, but then you go type it into, at the time it was Yahoo Finance, and look at all the holders. And the holders would be, well now it’s excluding Vanguard, BlackRock, State Street because they’re just the massive indexes. But if you look at the concentrated holders, it would be LSV, D. E. Shaw, AQR. All the firms that had all the PhDs who had the same databases, you end up sort of with the same names. And to me, the question I was kind of thinking about is looking at the modern history last 20 plus years and thinking of something like the value crash is like, all right, has it all been commoditized at this point?
And do people need to start thinking in terms of, all right, we got to find factors that either are underappreciated people aren’t talking about, or is it a scenario where they work fine, you just have to put them in the context of history. Give us some insight on someone who’s gotten their hands dirty with the data. What’s your perspective and how should we be thinking about the commoditization of everyone having 100 PhDs on staff with the same data?
Mikhail:
Yeah. If we’re talking about active investing, like trying to beat S&P 500, that was bashed into my head as one of these impossible challenges right away in college, markets are efficient, that’s what we’re taught. And if you’re going to go out there and try to beat the market, good luck to you. And then what I ended up internalizing that as, since I ended up on that job by total chance, I was actually partially a filmmaker, very creative editing final cut videos together and then I was editing data together, building a beautiful model backed those charts. I love the beauty of it. I realized that for me, it was naturally a very creative space. A lot of innovation for me. Alpha, if you try and beat the index equals innovation, live or die. It’s like becoming a musician or a writer. Your next book, your next blog better be interesting, better be new, otherwise no one’s going to care and read it.
Building the first model in 2004, I look at univariate, universe wide, price momentum value. They’re awfully looking things to me because they have tons of calendar years where they don’t work, they have decades they don’t work, even on short history. Well after I had long history for short decades and so I start innovating and luckily that fundamental map that I got handed over had these groups of companies, I believe there were six groups of companies based on their growth rates, which were dynamically gradually evolving. And you analyze each group differently using different sets of factors. That gave me some room for this, later I found out this was called dynamic contextual modeling. Pan Agora labeled that term in a really interesting book they have. And so by the time, to your kind of point, we arrived in 2007, which was the first big wake up call for the quants.
Quants were around 10% of the market back then in terms of assets under management, according to my estimates, it was using data, 90% was fundamental. Those 10% quants, 75% of those assets were in the hands of three firms, the biggest three quant managers. And I’ve seen some of their presentations back then and I’m friends with many of those people and it was models based on six or seven factors, maybe sector neutral at best, but a lot of univariate. Meanwhile, at that point I had about 125 variations of all sorts of things I was just kind of thinking about, testing, if it was positive, I take it. I think one of the biggest misconceptions in quant investing is this absolute fear of type one errors accepting a false factor. I think it’s a type two error, which is inversely related to type one error.
The more you try to control the type one error, the higher your type two gets. Type two error means you’re rejecting something that’s actually true. It would’ve worked but too bad, you rejected it because you were too paranoid about accepting factor that was not real. If you think about it, a factor that’s not real means it’s random noise. You thought it was real, but you over data mined, well it didn’t work out. It’s randomness. The probability of it flipping from a positive T-stat to a negative T-stat over a long period is just really low. It could happen, but that’s not the likeliest scenario. Likeliest scenario they go random. They dilute some of your good ideas, randomness. Transaction costs are almost zero anyway. Okay, some would say transaction costs from randomness, but really it’s not a big deal. What you don’t know is which one of the factors is going to be the good one. And the more you innovate, the more chance you have to come up with some that over the next decade actually have a positive spread.
The random ones will bring it down, but you’re still beating the index. The sum of zero plus positive is still positive and 98% of active managers don’t outperform. Even if you get 10 basis points, you’re already better than 98%. This brings me to all the innovation that was not done enough, to my opinion, even with existing. The way I was doing it, there was dynamic contextual models, but let’s say I like also this concept of forward-looking innovation, not reactive to what just crashed, but proactively looking at your models and thinking where do I go all in for the next six to 12 months? I remember looking at earnings quality, I had version one of it built by 2005 or six. It was flattening out. A lot of papers came out by Sloan and others and I had very basic versions of earnings quality. And then I decided that, let’s double down on this one.
I had my reasons. Hired an intern, really talented guy and we just went and ripped through every earnings quality paper out there and played with balance sheet cashflow, all the versions of accruals and profitability margins. Built together this super-duper cluster from all these little ratios, dynamic contextually applied, and it ended up being one of the best performing overall mega cluster for the next 10 years. And then once that research would be done and production wise, I would kind of refresh. It’s a very creative process. I take a subway New York back home and sometimes an idea pops in my head. I take it back to Wall Street and stay up in the office till 3:00 in the morning. It’s like a painting and then you go out and have a drink and sleep in and excuse yourself the next day. You kind of go a little mad building these things because like art, it becomes really personal and creative process.
My favorite thing with students, I teach a bunch and asking this question before you read anything out there, which is a good idea to read, but first take out a clean sheet of paper, a pen and start writing down what’s in there, what kind of questions come out, what kind of ideas come out because like a different musician, everybody’s got a different style. And there’s many ways to invest. Well, some are shorter terms, some are longer term, some are more fundamental, some use AI. A lot of the best ones actually data mine like crazy to go back to the type one to type two error like Renaissance Technologies, they can’t explain most of their stuff, I think. Very unpopular to do that, but it works for them. Of course there’s bad ways to data mine. I’m not saying if you just completely data mine, you’ll get a flat outer sample for sure if you overdo it.
It’s more like idea mining. You got to keep generating ideas, test them, and then don’t over torture yourself about being theoretically perfect around this idea because guess what? You never know until history will tell you. And a lot of academics, they sound very smart and math heavy, but at the end of the day, even value and size are the two most validated premier out there and they haven’t done much. Have the creative process, figure out what your personal style is. Hopefully it aligns with the firm, which they give you enough room to be you in the creative sense, and then you have a lot of freedom. Either you’re tweaking existing value like value, you can just live in value. If you say you love value, there’s so much you can do just within value, right? You can try to be like Warren Buffet kind of value. I love what you do with buybacks. Total yield. That’s definitely a huge difference between dividend yield and shared buybacks. That’s cool value together, keeps you up with the growth kind of side of value or whatever way you see it, but that’s already innovation, that’s big and then you just keep going and going and you end up with intangibles, which I ended up there. We can maybe talk about it, but…
Meb:
Yeah, let’s hear it. That’s a perfect lead in. We’ve done a few podcasts on kind of intangibles and it’s fascinating to me because it’s not something I spend that much time thinking about until people really started to bring it to my attention. Give us an aru. I know you also have a startup, Mikhail’s Two Centuries, which is such a great name for an investment company. I think if you named it two decades, people would be like, oh my god, that’s too long. Two Centuries, that’s such a great one. But also you got a new startup if we can talk about it, but tell us a little bit about intangibles. What led you to it and how do you think about them? What do they mean? All the good stuff.
Mikhail:
Yeah. This was also another very personal, kind of continuing on the story I was just sharing around… I graduated Wharton and then 2008 happened right in the middle of my MBA. I was working for AIG investments, AIG falls apart. My performance for the quant fund is great. We’re beating the benchmark, but fundraising is totally dried up. Quants are outer favors because quants just blew up in ’07 and again in ’09. I graduated in ’09. By 2010, I’m telling everyone let’s keep innovating, keep innovating, but myself, I’m kind of running dry a little bit. I take a sabbatical, I go to France and commit to coming up with a great new factor and I spend three months, I come up with something, I bring it back, I plug into the overall model and it barely moves the needle. Nothing improves in the overall model, even though the factor backed us was great. And I was like, oh my god, I can’t innovate anymore. Having an existential crisis. What am I going to do? And I realized kind of thinking about it is that I was using the same data I always used, and it was already in the model in some way or another. Having those 125 little ratios kind of eats away a lot of degrees of freedom. Whatever I came up with was already in some combination there.
Then life takes me to Hong Kong. I was running this company for this billionaire, the guy who bought AIG asset management. It was a crazy one-year stint there, but shifting the perspective and thinking about everything out of Asia, I start to sit back and ask myself the big question, where do I truly believe companies value comes from? And with that distance, it was right away kind of pops into my head that it’s the intangible assets. And then I look around my friends and companies I’m buying from, and this balance sheet in my head totally flips from tangible economy to intangibles. Customer satisfaction, brand reputation, leadership, leadership. My first boss, not the head of the whole department, but another lady who’s in between. She was a pure diehard PhD quant kind of from abstract quant side, and she used to make fun of the management in a way that, well management doesn’t matter, it’s a random factor.
You can replace one CO. She taught me a lot of good stuff about so much in quant, very grateful to both her and her boss. But that one opinion stuck with me and I was like, really? I don’t know. I think management does matter. And then looking back at it, I’m like, of course it matters. It’s so important. Then when I quit my Hong Kong gig, I went traveling again and I was like, all right, I’m on my own now. Let me try to build a new kind of factor model where I’m not stuck in the traditional data, the traditional factors. Let’s see what I can do with this intangibles. And I started looking up, this is around 2011, there was some brand values floating around the internet and there was some customer satisfaction surveys and there’s some employee engagement surveys. And then the more I dig, the more I realize, hey, there’s actually dozens and dozens and dozens of these ugly looking unstructured things, ugly from a quant point of view.
There’s no unique identifier. Who knows if it’s point in time, you got to map names and what’s the underlying process anyways, it’s just a random online list or is it a rigorous process that generates 10,000 surveys and they get systematically every year aggregated to the score. And that’s the kind of digging, I said I moved to a Los Angeles for two years and ran this startup, which was really just me glitching out on alternative data by myself called Okta Quant. And I hired a bunch of people in India and all over Asia online to help me type up all this stuff from fortune lists and Forbes and whatever. Upwork. Upwork is great tool to hire cheap labor if you’re by yourself, or there’s very talented people there. Anyway, so I ended up collecting over 120, depends how you measure it, different small, very narrow subsets of data and then based on the underlying processes of how the data was generated, I selected the top seven brand reputation, customer satisfaction, employee engagement and leadership. Those were my initial intangible assets. And I first built the brand back test, brand yield. Brand equity divided by market cap and then a change in brand, year-over-year change in brand value. Value momentum, combine the two things together, it was off the chart. I was like, yes.
Meb:
How often does that change though? Does that actually change? Does that change that much over time or over short periods or is it more just the stock price cratering?
Mikhail:
Yeah. Marketing caps change, but year over year brand changes, they do change. They definitely change, reputation changes. You see AIG end up in the bottom. It’s slow moving. It’s definitely slower moving than some faster factors like value and quality and momentum, but it moves enough to have a 12-month return that’s very strong. I’m talking about generation one of intangibles data. Now generation two is where I’m at now is a lot more AI, natural language processing, it’s a lot faster, but generation one was slower moving.
Meb:
Can you give us a little insight in the culture line? Is that a part of this idea and process and AI or is this something a little different?
Mikhail:
Yeah. Culture line came out of my collaboration with a really talented ex-quant also from a big American century quant guy who we collaborated on a consultant project that I was working for Voya Financial, helping them build an innovative ESG model. As part of Two Centuries, I was working with Voya and that was a year kind of and a half long project. And Taal Asani, who’s my partner in Culture Line, did a great job there and I already had culture as part of my intangible asset I was working on and I knew I could never get Taal to work for me full-time because I just couldn’t afford it, his rates, and he wanted to be independent. I decided to hey, in the spirit of collaboration and innovation, which happens in many industries except ours, which is very siloed and seems to be everybody protects everything, I thought that the research will go way further if we start this data company where we can actually offer these insights to other investors and I’ll, from Two Centuries, benefit from it and it’ll go a long way to collaborate like this.
And we’ve been making really exciting process on it compared to the first generation models. These are the latest large language models can really dissect every word, every piece of meaning that exists and you can then build it up and aggregate into a lot of really cool features and culture specifically and human capital is just such a rich from dataset point of view. There’s a lot of data with employee reviews and other human capital data, but you can also glean that from the 10K reports and the conference calls. And we were personally very motivated also, we worked with some very great people and some very challenging people in our careers. And so you know how it feels inside a great culture or toxic culture. And so, we have a lot of contextual experience in that topic, how bureaucracy feels, how burnout feels, and through language you can now have all this granular understanding of the different types of positive and negative cultures, what moderates people internally versus just salaries.
Are they tap dancing to work or are they burned out and just quietly quitted or actually going to quit in the next 12 months? All of that is being built up right now and cooked up in our kitchen. We have some prototypes, we have some early clients testing it, but we’re going to be rolling it out next year as a full-fledged alternative data culture-based rankings.
Meb:
And if you had to guess, do you think the best use of this will be standalone or inserted into a traditional factor framework or are you just going to wait to see what it spits out?
Mikhail:
We’ll see. I think our ideal case users in the beginning are fundamental folks, actually, who have culture as part of their investment process. The beauty of these models is that you can really customize. Let’s say you have a Warren Buffett kind of culture preference, we can quickly convert and combine features to give that kind of culture tilt. If you’re more of an innovative culture shop, we can focus on that. I don’t think saying there’s one right culture for winning is really our main goal. Something might show up like that in the back test and we’ll show that whatever the back test history is not going to be too long. Although who knows, maybe one day I’ll come up with a 100 year culture back test, which I think might be possible, but obviously simplified version. We want to kind of work in this way where these scores are a bit more customized into the process that investor is already comfortable, already has insight in and that’s our target plays because the bigger quants are already using some of this data. They’re already building a lot of these scores and we don’t want to compete on the strength of a back test to do this. We want to compete on the quality of their underlying measurements, the insights, and then if they match the demand for that quality, that’s where it would feel the best from the business point of view, and there’s less crowding in that and there’s more differentiation. Those features are nice.
Meb:
I want to hear the Steve Jobs story. What’s the Steve Jobs 10K story?
Mikhail:
That was one of my aha moments. I think about creativity and how it’s such an important part of culture and I’m like, how am I going to ever measure it? And then of course language comes to mind and then I think, well, who’s the most creative person I can think of? Well, Steve Jobs, no-brainer. And then it’s one of these investigative moments, I’m like, well, let’s look at his 10K when he returns back to Apple in ’97 from Pixar, very creative firm. And that, just breathtaking. If you look at the 10K between 1996 and 1997, ’97, he’s back as a CEO. There’s one word that changes in their opening sentence, and he adds the word creative to describe their target customer. The other words that describe the customer are education, consumer, I forget the exact words. It doesn’t have the word creative next to the customer and then this word appears. I’m like, wow, it’s so cool. First of all, it tells you that sentence, he looked at it, he read it. It’s not just written by lawyers and marketers.
He took time to edit it and he puts the one essential adjective that tells you the direction of the future culture. And then the second breathtaking moment was I then of course fast-forward to when he sadly passes in 2011, and that annual report comes out a month after he dies in October 2011 because they’re off the typical annual reporting cycle. And if you compare 2010, the first sentence grew bigger and bigger, so it was a lot less punchier, but the creativity word was still there before he died. And then in 2011 they took it out. The one word difference, they take it out, Tim Cook takes it out. I was like, oh my god. And first I thought, wow, that sucks because creativity is so cool and do they not get it? The whole thing was about creativity and then this is my thinking when I discovered it in 2015 or so. But then time goes by and I’m watching Warren Buffett and you recently wrote a blog on it by Apple and it’s his best investment in dollar terms ever, like over $120 billion he made.
And I started looking at Apple’s culture through our culture line scores and it evolves from innovation. You do see after Steve Jobs dies, innovation scores on the culture start to go down. And before they were crushing Microsoft on innovation, and then it took 10 years, it kind of converged just with Microsoft on innovation, but what I didn’t get right away, now I can see it. And what Buffet did get right is that they bet on quality, and they had a high quality brand, high quality product. Tim Cook, his DNA is quality, execution, production. Again, to my point, there’s many ways to win, not just innovation. Quality is another one. And their quality culture continued a huge edge versus Microsoft and Buffet bought it as a quality consumer, not an innovative tech firm, more of in his wheelhouse as a quality brand. But by the way, Buffet talks about intangibles so much and he kind of spelled it out. He’s like, look at the quality of management, look at employee engagement, brand, et cetera, modes.
Yeah. I think the kind of touching, and Steve Jobs rarely spoke on conference calls actually. I try to see if I can get a lot of his language. He rarely, rarely spoke on conference calls. It’s hard to get too much insight. Of course he has famous talks, but his footprint in the 10K was really fun and I felt like I was on the right track tracking language as a place to see what leadership is doing in the subtle way where we’re taking the company.
Meb:
I love it. I signed up on Culture Line, so I’ll get all your updates. One of the things, and again, listeners, you got to download the papers because there’s so much goodness in these, but talking about, I think I originally might’ve reached out to you, I was like, man, I love this paper talking about asset allocation strategies. And most investors they think about how to put it all together, but one of the things that we saw a lot is people have a strategy, and this applies to individual strategies, but also entire ways of thinking about allocation and it may not do well for a while and they kind of move on to something else. Certainly periods where various parts of the portfolios underperformed or the strategies underperformed, but I remember getting to page 40 and on in the appendix of this paper and there’s so much rich resources as far as data sources, models, how to build and kind of replication of certain ideas and asset classes for a long time.
It’s worth the download alone, but tell us a little bit about a century of asset allocation crash risks for those who are now ready to put it all together. Any main takeaways, insights you got from studying and putting this paper together that you think would be particularly interesting?
Mikhail:
Yeah. This is sort of the after nine years of doing individual extensions of value momentum, I then took it to asset allocation. Asset allocation, I think, is one of the biggest unsolved puzzles in finance. All these smart finance departments around the world haven’t answered the question, what portfolio should an investor hold? They themselves don’t really hold onto it. The reason I’m saying this is because as we all know, dollar weighted returns are very different than time weighted returns, meaning people do not end up holding onto what you just said several times do not end up holding to their portfolio over the long run to actually get a time weighted return. There’s a lot of in and out depending on what’s happening. And so I started really zoning in onto this and then using long-term history to help me answer my favorite questions, how much does something crash?
Because that’s one way I’m going to get out of something personally and professionally. And then the other reason I usually get out of something when I have fear of missing out the upside, I think you asked somebody recently also this on your podcast, is that it’s true if investors underearning, the death by a thousand cuts, they end up ditching their boring strategy and going into a different one that recently performs better. And that generates a dollar weighted return gap, which is around between one and 3% a year, but it feels a lot worse when you personally sell at the bottom and don’t earn upside. Meanwhile, this idea for this paper came out when I was working for a Wharton professor, so we were doing a lot of academic style consulting and building different types of portfolios. And the traditional 60/40, obviously I think hopefully most of us know it crashes a lot because of equity risk in it.
Famously, 90% of that portfolio is driven by the equity risk, even though it’s only 60% inequities, which means in Great Depression it crashes 63%. Now before 2008, we might’ve ignored Great Depression just like with factors, but once 2008 rolls around, 60/40 crashes 33%, which its worst crash ever since the Great Depression. Unless you looked at the Great Depression, you would’ve thought this is the end of the world and you weren’t prepared for that crash. By the way, 33%, 63% range is way beyond a moderate risk investor. 60/40 is like a moderate risk investor, but three to six standard deviation events are not moderate risk. Unfortunately, equity drawdowns are not normally distributed, as we know. There’s these tails. Normality doesn’t hold in these 2% of cases and that’s where investment, either the actual drawdown or the fear of a drawdown like that ruins it for most investors to hold onto.
60/40 is hard to hold onto once you look at the evidence that it could be a 63% crash. Then diversification keeps going and risk parity gets invented somewhere in the nineties but becomes really popular in 2011, which it’s relative to 60/40. It was crushing it on a trailing basis by 2011. A lot of institutional money flows in. You could see it in the pension funds statements that are public. They started allocating to risk parity around 2011. Of course, unfortunately, that’s just the peak of risk parity and it starts to mean revert, and by 2018, they’re all selling it. Now all you had to do, which we did, was zoom out and look at a hundred years of risk parity versus 60/40 and its zigs and zags and mean reverts, and it’s pretty much the same crash risk, same average return, but you’ve got to use leverage, commodities.
I mean to do risk parity extension, by the way, we had to have the commodity futures extended back to 1927. That took the nine years just to extend that piece of it. And risk parity without commodities doesn’t hold the full benefit. Then there’s the endowment model argument. A lot of people are arguing for endowments and of course there’s beautiful track records by Yale and some others with great access, great managers. There’s a lot of alpha there. But as a kind of beta portfolio construction idea, we tried our best to extend endowment 100 years. That was the hardest one because we don’t have hedge funds and private equity, but we use factors to extend it. We look at factor-based asset allocation, which is where I spend this five years kind of trying to see is it possible to have a portfolio from asset allocation point of view, not from an alpha point of view where factors sit on top of a benchmark, but where factors are sitting next to traditional factors like equity premium, fixed income premium, commodities, and then you have value, momentum, 30, 20% allocations across the portfolio.
That was a theoretical limit of where we try to push factor investing until it was really just anomalies, in my opinion, and they’re flattening out so they’re not reliable from the return point of view, in my opinion. If you’re going to allocate 30, 40% next to equity beta, they might be reliable in an alpha sense or some other ways. Depends, again, how much you innovate inside of them. The takeaways from the studies, the drawdowns are similar for the first ones I mentioned. 60/40 diversified all the pie charts that you can build from 23’s to 30 different SBAA class, we extended REITs, we extended all sorts of things, growth value, et cetera. All the traditional asset allocation stuff like that crashes too much for moderate risk investor to hold onto. Then you get into risk parity endowment, same thing. Then you get a factor based.
There you do get a pickup on drawdowns from 60 to 40% because now you just added a whole bunch of uncorrelated return. Unfortunately, again, how do you think about it going forward on a premium basis? But as a question there, but at least it does improve drawdowns. And then the best one, you start doing it dynamically and that’s your work has really pioneered, I think on your papers, but dynamic asset allocation where you have these divergent, especially signals like trend, the cap, the downside, volatility targeting, long-term bonds, the hedge equities, you plug in that system, and not too sophisticated and just even plain vanilla kind of for a quant. Not to, in my opinion, to harvest any alpha, but just to kind of harvest the same betas that exist there. With this couple factors like trend and volatility targeting, you kind of reshift the risk. Those drawdowns get significantly improved in long-term history to become easier to hold onto.
The big caveat is that paper ends in 2020 data, December 2020, and the one thing that history teaches me is that always be ready for surprise and uncomfortable stuff. No matter how long you look at it, there’s still surprises. And 2022 was pretty brutal for dynamic asset allocation because both stocks and bonds had a drawdown that was inflation driven, not growth driven. In hindsight, everything’s so obvious. And the long-term yields didn’t protect either. We had a similar drawdown as a 60/40 would have. In some cases, dynamic is clearly better. In other cases it might not be better, but at least, well again, so history teaches you a lot and the main goal of that for me was just building up resilience to be able to hold on to whatever you pick for longer. If you pick risk parity in 2011, please don’t sell in 2018, just hold onto it.
All the zig and zagging will wash away. And if you made that choice based on your theoretical studies of how you want your distribution and returns to look like, more balanced across environments and growth and inflation, then stick with that bet or innovate within that, but don’t drastically jump out when it does the worst. Same with value investing, same with dynamic asset allocation. I’m holding onto that one in my main multi-asset portfolios, but I’m also open to innovating and building resilient other types of portfolios, but always look back at history to inform myself how bad can it get? Because inevitably things will come close, at least in the imagination of investors when you read all the news. That’s dealing with that uncomfortable distribution of returns is what our paper was about, and that showing investors longer history helps them hold on, no matter what portfolio they pick. And doing it dynamically helps in many traditional growth driven market crashes.
Meb:
Yeah. Thinking about drawdowns, I mean almost all investors underestimate individual asset class drawdowns. I love doing the polls on Twitter, talking about bond drawdowns. So many people think they’re zero to 5%, although they’re learning very quickly now about how big bond drawdowns can be, particularly real ones. But even with asset allocation portfolios, I’ve said on Twitter a number of times over the years, I said it’s almost impossible to come up with a portfolio that over time doesn’t decline by at least a quarter and more likely probably a third to half. The longer you go back to, and I was laughing, because I love to poke CalPERS, and CalPERS had a piece out when they do their quarterly meetings or their presentation and they were talking about maximum drawdown risks. And in their portfolios they were showing 20 to 25%. And I was like, you can go ahead and double that, all right? There’s no scenario you have a, and I think if you marked assets to market, probably would’ve already hit it within one year. It is like, boom right after that. But I think that’s dangerous because unless you think in terms of the worst case scenario is when something bad happens, you’re probably going to react emotionally and usually we know that that ends up being kind of a rough place to make money decisions.
Mikhail:
Absolutely. Endowments was another fund. I took actual endowment returns, which are annually reported June to June, and then you extrapolate them with the monthly. When you’re using the asset allocation, you plug in the indices, but then you make sure the return ends up at the same level so you’re not changing the average return, but you fill in the missing data for the monthly data and that draw down in 2008 was 30 to 40% for the endowments. And they’re super diversified, a lot of alternatives. A lot of the volatility might wash away if you use, again, private equity. Oh yeah. We would unsmooth private equity to get there as well. That was a big caveat for endowments. You had to do that as well to get the monthly private equity returns from quarterly. Again, you don’t change the returns, but you add back the actual volatility that listed equities experiences and you see a lot of risk out there everywhere.
Now behaviorally private equity might have an advantage because they lock you up and also they do the calls. And I never thought I would be saying that as a listed guy always, it is fun to make fun of private equity because you say, well, it’s levered beta with a lot of fees ignoring the alpha argument, whereas do they have it or not? There is an argument to be made about that dollar weighted versus time weighted where as a listed equity investor, you don’t have control over client’s dollar weighted return. You can try to influence it, by the end of the day they decide when to give you money or pull it out. You can just do the coaching and try to do your best, but then you are responsible for time weighted return. In a private equity world, they call you, you give them the money, they give it back, they give it back.
They actually have the dollar weighted return under their control, which is behaviorally definitely a plus for that asset class. Of course, you give up liquidity and then many fees and there’s all the other problems we can talk about, over smoothing returns, et cetera, market to market. But yeah, that part exists there. But in general, no matter what you’re doing that’s sort of upscaling maybe the risk, smoothing it out, but underlying volatility, yeah, as you said, I haven’t seen a portfolio that is not just pure alpha driven from, again, we can pick on some best hedge fund managers, maybe they can get 10% with zero beta, but they’re closed to new investors and the capacity issues are there. But as a general for the public, for the advisors to be able to have asset allocation, you got to be ready for 30% drawdowns if it’s a moderate risk portfolio. Or more really, I mean 30%, you throw out that number, I like to say 63, but then nobody would invest. I do say 63 in Great Depression.
Meb:
Yeah, no, exactly. I try to err on the side of it being palatable. If you say something worse, people just think you’re crazy. We’ve talked about a handful here, but we like to ask guests two questions now. One is what’s something you believe the vast majority of your professional peers don’t believe? You mentioned you used to reach your head over the cubes and make some ascertains, but what’s something now, 2023, that if you were to say in a crowded room of pros or at the bar with a bunch of your buddies, they would shake their head and say, I don’t agree with that. Anything come to mind?
Mikhail:
For the quants out there, I still think that the most unpopular belief is that type two error is way worse and more dangerous to our industry, to their office, to their careers. And type one error, most people get totally into type one error. Hypothesis first, test it once, don’t data mine. And that just completely shrinks innovation and I think that’s why we don’t have much alpha out there. That’s for the quants. For the fundamental and asset allocation people, I think the very idea of saying stuff that’s unpopular and hard to say, that’s the area where there could be big return moves. And the stuff that’s easy to say and everybody shakes hands, it’s unlikely you have a big return move or positive return at all. If I look at it today and I just stood up and said, hey, I think market is going to double over the next couple of years.
I don’t know fundamentally, I don’t predict fundamentals to that precision or to that level, but if I were to bet between somebody saying, be cautious, there’s a big recession versus things are going to double, even though the doubling thing sounds crazy and I have all the data to show that it’s recession, I would be careful which way I would bet. And I think people, like last September, I was very comfortable to be very bearish and I was telling clients, look, I don’t know if we’re at the bottom. I have no idea. I do know that it’s much more uncomfortable to be bullish now, so if there is a return, it’s much likely to be in the bullish upside direction than in the downside because downside is really everywhere. Everybody’s comfortable with it unless it’s some really extreme downside that would make me even uncomfortable to say it out loud, people think I’m crazy.
Otherwise, if you’re just bearish, it’s already out there and it’s shifting. Now soft lending is getting comfortable. Bear market is also comfortable. The market is somewhere between the bear market and recession soft lending. Nobody is really comfortable being bullish. Again, it’s not about the fundamental correctness. I’ve read this somewhere from Silicon Valley that there’s two things. There’s a two by two matrix being right about the future in terms of fundamentals and actually what’s going to happen and then being popular versus unpopular or crowded or not. I carry that over to investing and I always do this gut check with myself, am I comfortable or not? And if I’m not comfortable, that’s a good thing, even though it’s brutal to live with some of that, but you kind of get used to it.
Meb:
And so what’s been your most memorable investment? Anything come to mind, good, bad, in between?
Mikhail:
I have a lot of humble lessons where all the memorable investments I haven’t made. I’ve read Peter Lynch’s book early on in my career and man, I wish I just followed that. If I’m buying a product, just buy the damn stock with the same amount. I bought my Mac and I was a film editor, and so I would use the university’s computers. And then finally I saved up for my own big G, whatever it was, three or four in 2004, and I spent like 4,000 bucks on it. I barely used it to edit anything. At that point, I was a senior and partying. And if I just put that 4,000 bucks in Apple stock. And then the same thing happened many, many times with Whole Foods. I would start shopping there or Lululemon, my wife would wear, or Tesla in LA when we lived and my neighbor friends started driving it.
I was like, that’s an amazing car. All those stocks I didn’t own from beginning. Google IPO report, I read. I loved it. I used it and didn’t buy the stock. I’m not very good at fundamental stock picking. As a quant, I have a really solid track record. The original model I described here a couple of times continues to outperform. It’s part of now PineBridge Investments and by new models that are built on intangibles since inception have positive alpha, so in dollar weighted impact and time weighted, that’s my strength here. But ironically, the best dollar weighted return I’ve compounded personally is in my 401K portfolio, which is the classic advice of just set it and forget it. And as a young analyst, I just open my 401K accounts, read the book, random Walk down Wall Street and start putting money into S&P 500 systematic, well rules based every paycheck. Now it’s managed by my strategies, that equity strategy I run. But that’s been my best dollar weighted investment, not surprisingly, but also kind of surprisingly that it does work.
Meb:
Speaking of Peter Lynch, we found an old video from 1997. We’ll put the show note links that was a lecture on his 10 most dangerous mistakes investors make. It’s on C-SPAN two, but it’s a fun, grainy old video. I was smiling as you’re talking about the missed opportunities Peter Lynch style. When my father passed away, I found an old postcard from 1989 that was talking about Coke, Anheuser-Busch, McDonald’s and Disney. And if I just put $100, I think, into those at that point and put them away in a lockbox forever, that would’ve been probably better than all the other alpha it ever created in my career. Some lessons learned about the old Lynch style. Peter, if you’re listening, come join us on the show sometime. We’ll talk about it. Mikhail, this has been a blast. Where do people find your very voluminous work and great studies? Is there a best place to sign up to track what you’re doing?
Mikhail:
Yeah, twocenturies.com. It has a lot of my blogs out there. It has a page with all the talks and papers I’ve written. And for the culture insights, sign up for cultureline.ai and stay tuned. I’m also on Twitter and LinkedIn.
Meb:
Awesome. Mikhail, thanks so much for joining us today.
Mikhail:
Thanks, man. I really enjoyed it. Thank you.
Meb:
Podcast listeners, we’ll post show notes to today’s conversation at mebfaber.com/podcast. If you love the show, if you hate it, shoot us an email at [email protected]. We love to read the reviews. Please review us on iTunes and subscribe to the show anywhere good podcasts are found. Thanks for listening, friends, and good investing.