Episode #315: Warren Pies & Fernando Vidal, 3Fourteen Research, “I Think That The Next 40 Years Are Unlikely To Look Like The Last 40 Years”
Guests: Warren Pies is the founder of 3Fourteen Research. Prior to founding 3Fourteen Research, Warren led Ned Davis Research’s Energy and Commodity strategy. In that role, he built the firm’s commodity-related studies, models, and unique indicators. His research combines proprietary fundamental, technical and macro indicators to identify major investment themes and market trends affecting capital markets. He earned both his Bachelor of Science and Juris Doctorate from the University of Florida. Warren is an Energy Risk Professional – Certified by the Global Association of Risk Professionals.
Fernando Vidal spent 7 years working as a quantitative analyst at Ned Davis Research’s consulting group conducting research, building and testing models and studies for institutional investors. Following this work in the investment space, he spent 6 years founding and leading a Data Science team at SauceLabs, a VC-backed fast growing market leader in software testing based out of San Francisco. At 3Fourteen, Fernando leads our model development process and brings machine learning research into our mix of qualitative analysis and quantitative rigor. He earned a Master of Science in Machine Learning from Georgia Tech and a Bachelor of Science in Finance and Economics from the University of South Florida.
Date Recorded: 4/28/2021
Sponsor: AcreTrader – AcreTrader is an investment platform that makes it simple to own shares of farmland and earn passive income, and you can start investing in just minutes online. AcreTrader provides access, transparency, and liquidity to investors, while handling all aspects of administration and property management so that you can sit back and watch your investment grow. If you’re interested in a deeper understanding, and for more information on how to become a farmland investor through their platform, please visit acretrader.com/meb.
To listen to Episode #315 on iTunes, click here
To listen to Episode #315 on Stitcher, click here
To listen to Episode #315 on Pocket Casts, click here
To listen to Episode #315 on Google, click here
To stream Episode #315, click here
Comments or suggestions? Email us Feedback@TheMebFaberShow.com or call us to leave a voicemail at 323 834 9159
Interested in sponsoring an episode? Email Justin at email@example.com
Summary: In episode 315, we welcome our guests, Warren Pies and Fernando Vidal, co-founders of 3Fourteen Research, which combines expert qualitative insights with true quantitative discipline.
In today’s episode, we take a data-driven approach to look at the markets. We start with the firm’s original story and why Warren believes real assets have a place in portfolios going forward. Then they walk us through their research process and the benefits of combining machine learning with technicals and fundamentals. Finally, we hear what their models say about inflation, commodities, oil, and Bitcoin.
Warren and Fernando were kind enough to put together some of their research for you to refer to during the episode. Click here to see their reports.
Please enjoy this episode with 3Fourteen Research’s Warren Pies and Fernando Vidal.
Links from the Episode:
- 0:40 – Sponsor: AcreTrader
- 1:32 – Intro
- 2:30 – Welcome to our guests, Warren Pies and Fernando Vidal
- 3:07 – The origin story of 3Fourteen Research
- 9:38 – How they approach markets
- 14:16 – Best practices for building and designing machine learning models
- 16:10 – Their view on oil prices and inputs for their prediction model
- 19:26 – The power of narratives when it comes to market valuation
- 22:31 – Defining parameters and levers that are built into most models
- 24:52 – Overview of their real asset allocation model
- 28:02 – Calculating historical data over a rolling multi-year window
- 28:49 – Whether or not they’re involved in the cryptocurrency space
- 31:48 – The driving force behind scaling back their equity position
- 33:02 – Thoughts on yield optimization and the energy sector
- 35:54 – Bracing for the Hangover, and their thoughts on inflation
- 40:46 – Is gold a “chameleon” asset?
- 43:04 – What the future holds for the US Dollar
- 48:03 – Stock positioning as the world gears up to reopen
- 52:15 – Their business economics, clientele, and services provided
- 54:49 – Common questions and recurring narratives amongst their clientele
- 56:17 – Thoughts as they look out to the horizon; Saving us from Ourselves
- 58:24 – Their most memorable investments
- 1:00:55- Learn more about Warren and Fernando; 3fourteenresearch.com; Twitter @3F_Research
Transcript of Episode 315:
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: Today’s episode is sponsored by AcreTrader. I’ve personally invested on AcreTrader and can say it is a very easy way to access one of my favorite investment asset classes, farmland. AcreTrader’s and investment platform that makes it simple to own shares of farmland and earn passive income. And you can start investing in just minutes online. AcreTrader provides access, transparency, and liquidity to investors while handling all aspects of administration and property management so you can sit back and watch your investment grow. We recently had the founder of the company, Carter Malloy, back on the podcast for a second time in episode 312. Make sure you check out that great conversation. And if you’re interested in deeper understanding for more information on how to become a farmland investor through their platform, please visit acretrader.com/meb. And now, back to our great episode.
Hey, friends. Great show today. We have a rare two-guest episode. Our guests are the co-founders of 3Fourteen Research, a shop that combines expert qualitative insights with true quantitative discipline. In today’s show, we take a data-driven approach to look at the markets. We start with the firm’s origin story and why they believe real assets have a place in portfolios going forward. Our guests then walk us through their research process and the benefits of combining machine learning with technicals and fundamentals. Finally, we hear what their models say today about inflation, commodities, oil, and Bitcoin. Our guests published some of my favorite research and were kind enough to let us share it with you to refer to today’s episode. Be sure to check the show notes, mebfaber.com/podcast, for their research and chart book to follow along. Please enjoy this episode with 3Fourteen Research, Warren Pies and Fernando Vidal. Warren, Fernando, welcome to the show.
Fernando: Thanks, Meb.
Warren: Appreciate you having us. Happy to be here.
Meb: We have a rare double interview. Where are you guys located? Where does it find you in late April 2021?
Fernando: I am in Sarasota, Florida on the Gulf Coast.
Warren: I’m in the San Francisco Bay Area in Moraga.
Meb: Well, gentlemen, first of all, congratulations on your new venture, 3Fourteen Research. I’m excited to follow along with your research and the work you guys have been putting out for a while. What was the inspiration? Why did you decide to go start your own shop, one of the hardest things a human being can do in the world, be an entrepreneur? What gave you that good confidence? Tell me the origin story, how you guys teamed up.
Warren: I started my career as an attorney for the natural resources industry, and then really always had a passion for the markets, didn’t have the traditional route though. And I was working in Central Florida, and had identified a few of the shops, this was more than a decade ago, that I would be interested in trying to, kind of, carve out a career in finance. And Ned Davis research was in my backyard, and I respected their work, the way they approached markets, in general. And as luck would have it, I was able to connect with the head of the commodities team over there, John LaForge, who’s now doing real assets at Wells Fargo. I hounded him for a period of time, and I think he was skeptical about bringing me on. Eventually, he gave me an interview and the rest is, kind of, history from there as far as NDR is concerned.
But when I got to NDR, Fernando was there. And he can, kind of, talk about his path. He was in the custom department and I was on the commodity team. My path eventually was to take over the energy space at Ned Davis Research, oil, energy, all that stuff, the entire complex, and then to eventually take over the entire commodity team. To take a step back as far as how did Fernando and I start working together, back in 2013, I think you might recall this, Meb, you and I had met at a conference around this time, I did the first-ever report from Ned Davis Research on the Master Limited Partnerships space, you know, pipelines, MLPs, and all that. And no one had ever touched that topic at NDR before. And I needed…The data was a mess. Actually, we didn’t have the data in-house. So we had to, kind of…I wanted to do, like, a true factor analysis of MLPs in that space. And so, I went to the customer department and really the best person there was Fernando, and he helped me with that study. That was the first time we touched this space at NDR. The editor of “Barron’s” liked it a lot and ended up inviting me to the MLP roundtable just based off that one report, which was really a factor breakdown of MLP.
So that started, I’d say, Fernando and my working relationship. And we really clicked well and we worked together well from the get-go. And that was back in 2013. So a long time ago. Obviously, we were friends and we travelled along different paths, but constantly stayed in contact and worked together while we were at NDR. For me, after I took over the commodity space and had, kind of, carved out my area there, you know, the pandemic hit, and NDR, they decided that they wanted to get rid of a dedicated commodity research and separate commodity team. And that was a casualty of 2020. And by then, in a strange way, I think that was, kind of, a nice contrarian signal that, you know, as shops were chopping their commodity jeans and chopping their traditional energy coverage and things like that, for the most part, I want to follow trends and investing. But here, I wanted to make a contrarian move and say, “You know, I think that the next four years, unlikely to look like the last four years.” And I think that this real asset space is going to be a necessary component to a broad asset allocation strategy that’s going to succeed in this new era going forward.
So that was, kind of, the basis for me wanting to do this. And then Fernando brings just a totally differentiated skill set, having a background in machine learning that he can go into. And so, you know, between the idea that I think that we have an expertise in real assets, we have a background there, and we also have a differentiated ability to bring true data science into the process and build our systems, processes, and models in a rigorous way that I know for a fact, having been at Ned Davis Research and worked in independent research space, I know for a fact it’s something that’s unique and differentiated in the market. So, that’s kind of the putting 12, 13 years of background into a 5-minute synopsis from my perspective, but I’ll let Fernando give his.
Fernando: Yeah. So, just to go into my background a little bit, so I got my training undergrad in finance and economics, and then did graduate studies in computer science, specializing in machine learning. I started my career at Ned Davis Research and spent almost eight years there. And I worked in their custom research department like Warren’s talking about. So, I basically did a lot of model building for institutional client base, as well as, like Warren said, contribute on the strategy side. Whenever there was some strategy work that required some quantitative research, I’d get involved there. So, the tail end of those eight years involved a lot of collaboration working with Warren. And honestly, the germ, I think, of 3Fourteen Research came from that really good working relationship that started at Ned Davis Research. I ended up leaving to join a VC-funded tech startup in the Bay Area about six years ago, basically, built out their data science team and worked on product features that involved AI and machine learning.
So, it was kind of a different field from finance investing but I’ve always had a foot in that world and wanted to get back into it at some point. And Warren and I, kind of, had this idea in the back of our head for a while. And in 2020, the stars, kind of, aligned and everything in the universe said, “Now’s the time.” So, I’m really excited to be working with Warren again. It is a really great fusion of Warren being a great investment strategist and then having, like, the background that I have in ML and machine learning to, kind of, give him access to this toolset of data analysis that I think differentiates our research.
Meb: We definitely have a long and happy…well, it depends who you ask. If you ask the Ned Davis side, they’d probably say I’m a pain in the ass. But a long and happy history of working with Ned Davis. We’ve had a few of the folks there on the podcast, even did all the charts for our very first book over a decade ago. And if you go back far enough, I even tried to get a job there. So you guys are infinitely more qualified than I am to talk because you guys eventually made it past the screening process, whereas I did not. So, let’s hear about your new company. What is it you guys bring to the world, new, different, interesting? What are your capabilities? And let’s hear the framework for how you guys think about the world. How do you approach markets and all that jazz?
Warren: I didn’t really make it past the screening process as well. I ended up…They give you that ELPAC, the language programming app that you test when you start out at Ned Davis, and that was totally…At the time, it was foreign to me. I ended up reaching out with a test maker and posing as a consultant and asked if I could get a copy of that test. I remember being so nervous before I took it. So, yeah, I had to kind of do my own creative way to get in there as well. But as far as how we look at the world, I think the best way to, kind of, talk about it is to work through one of our models and how we built, like, we should start with the oil model, for instance. And that was the first one we released.
Meb: Can I interrupt you and just say the really only signal you need to know is buy oil when it’s minus 30.
Warren: That is back-tested really well, actually. Yeah, that’s a very high batting average there.
Meb: That’s why they pay me the big bucks. All right, keep going. Sorry.
Warren: Taking one step back, I think that’s a great point is that our first goal is that having spent a lot of time in the industry, once you start to build models and play with data, what you realize is that it’s really easy to fool yourself. We’re not in this backtest beauty contest business in this company. That’s for sure. I mean, you’ve never seen a bad backtest. You know, that’s obviously, kind of, a meme that’s out there on Fintwit right now. And there’s a reason for it. And that’s just there’s so much overfitting to history and to noise in many cases. And so, we’re trying our best to be rigorous, rigorous in the front end when we ask the questions, like what are we trying to model right now, and really define that in a discreet and precise way. So we set up our research question in a very discreet way, and then we answer it as best we can without fitting the noise.
And so oil is a good case study. That’s an area of the market that I had an extensive background in. And so we decided to, kind of, use that as our first pass. I took the areas that I found, that I knew to be important when we’re talking about crude oil. So, we look at positioning in the futures market, the physical market, crack spread differentials, things like that, technicals. So price action obviously matters. Something that I’ve always said is you want to build your conviction on fundamentals, but you want to manage risk off of price action technicals. You take these different areas of the market, and I go to Fernando and I say, “Here’s what we watched in the oil market, these different regions, and we can backtest in them and have a certain amount of logic, a qualitative understanding of the market.” But then he’s going to provide that overlay of rigor when we’re testing everything.
And so, for instance, what we do is different and he can get into this stitching together, all these components, putting together…When he does the backtest with these different components, for instance, instead of just backtesting and saying, “How did these signals fit to the full history?” he’ll do cross-validation, he’ll do out-of-sample testing, and things like that. And he has access to different models and algorithms that we never had access to. You know, we’re doing sticks and stones at Ned Davis Research in comparison to what we can do here. So, having the power to do that but while being cognizant of what really matters and having a background in that space is, I think, a really unique combination.
Fernando: Jumping into that, it’s really important when you’re building a model and researching indicators, that you set up the process so that it’s possible to fail given a particular input. Historically, like, I’ve been part of some model building processes where the goal is to build a model and it’s a foregone conclusion that a model will emerge from the process. So it’s really important and, like, we take that really seriously. I come to it from a perspective of I’ve got all these different machine learning models that I can fit to data. I want to set up a framework so that Warren can feed me these ideas. And, like, the most important part of machine learning is feature engineering, which essentially, the best place to go to figure out how to engineer good features for a model is to domain experts. So with this oil model, Warren’s playing the role of, here’s the set of indicators that are useful. Here’s how I think they work. And that’s another thing where ML is important. So, ML gives you a huge zoo of potential functional forms for how you map inputs to outputs.
Meb: What are some best practices when you’re talking about, like, this whole process? Because it’s so seductive to get drawn into the output and the optimization to where you end up in this fantasy land of the fully optimized model. Anything come to mind as things these are the best practices? These are things we really want to think about when we’re building these models?
Fernando: Your scheme for how you’re going to do out-of-sample testing is probably like the first step. And that’s before you even have an idea of what inputs are going to help you predict an output. How are you going to cross-validate things? And you also have to think through what kind of model are you looking for? Are you looking for a model that is trying to discover a truth about how markets work that exists consistently across an entire history or are you looking for a kind of model that picks up on trends that exist maybe in the last 3 years and didn’t exist 15 years ago? Because you have all this historical data but if you come to the problem saying, “Yeah, this thing that’s happening in the last three years, I’m going to model that,” then the historical data is no good to you. And the shorter-lived the phenomenon you’re trying to build a model on is, the less data you have to prove that that theory is true.
When it comes to ML, broadly speaking, you want to be looking for truths or theories about markets that persist through time because those are going to be the ones that you can have the most confidence in. I think coming into the model-building process with an opinion on those things is really important, and it informs things like how am I going to do out-of-sample testing? There’s no point in doing out-of-sample testing if you know the phenomenon you’re trying to train on exists in one slice of time and that has never existed before because you know what the answer is going to be.
Meb: All right, so oil is going to the moon, back to the days, 100, 200, back down to the 0, negatives? Talk to us a little more case study. What are the inputs and what does it look like?
Warren: To wrap up the oil model, we have inventories, technicals, positioning, and physical market. So those are four big components that we’re looking at, more or less, build a model for each one of those four components and then stitch them all together. Right? And currently, we’re in this kind of…It’s not a great marketing angle but the model is a neutral. And it’s been neutral for the majority of this big rally here at the beginning part of the year. And as I’ve said when I talk to other folks on podcasts and interviews is that…and I think oil is important to understand for a lot of reasons and it’s important to get into, that’s been okay with me. You know, it’s kind of how I’ve seen the market. While we have, kind of, the illusion of tightness, when you look at the markets. So, inventories are drawing, right, and that’s read through the model as bullish and technicals look positive. We’re looking for uptrends and pullbacks, and we’ve been getting those on our signals and the physical market has been, like, off and on looking good. And the Saudis, they’ve, kind of, stopped that up with their unilateral cuts, right?
So you have these components that look good. On the other hand, we see futures positioning or is, kind of, extended. We’ve seen a lot of optimism in the market. Our model likes to fade that. So when you net all this out, it’s more or less a neutral signal. And that’s how I would see the market. Really the big overhang for crude oil, a really important takeaway and really the hardest thing to kind of, I think, handicap when you’re looking at the market at present is the massive OPEC bear capacity. So OPEC sitting on record spare capacity. And like I said, the cuts out of Saudi Arabia have really been the driver of the market. So here we are, we had Powell on the TV just like a few minutes ago today. Everybody on Twitter, the consensus right now in my view is that inflation is here, and that we’re all experiencing inflation, lumber prices, used car prices, oil prices, right?
And I think if you understand the oil market and understand the amount of spare capacity, and the reason why the market is rallying, which is really a supply-side issue, the Saudis and OPEC removing supply from the market and being ultra disciplined, I think it may give you a different perspective of that inflation or it’s not so much that we are in a really tight market where demand is outstripping capacity. What you’re seeing is reduced capacity through OPEC. And quite honestly, it’s not a sustainable posture for OPEC to hold this supply off the market for a sustained period of time. So, how that oil comes back on the market is that X-factor when you’re trying to create a view for oil. And so, we’re neutral right now. And I’d say that the model has no way of understanding that, which is understandable when you’re thinking through just the quantitative model. But it’s difficult from a human perspective as well to see how does this reopening and what’s going on in India, for instance, how does that collide with all this OPEC spare capacity? So I think it’s…you know, neutral’s a fair position. That’s where the model is and that’s honestly where I’d be.
Meb: It’s interesting because you guys had a piece…I like your pieces because you have some great quotes in the beginning and one of the oil ones from last year, you’re talking about narratives. The narrative, as you mentioned, certainly, is what you were talking about, lumber, inflation, everything else but it’s always fun to look at the actual components of a model. In looking at kind of y’all’s oil story, you have the curve indicator trading strategy that looks at are the futures in backwardation? Is it flattening contango? Is it steepening contango? And then also, you mentioned the role of CTA is where you fade them at extremes saying the position is too high. So it, sort of, extracts the media what you hear all day versus actually some of the things that are going on behind the scenes and putting the weights on those as needed to come up with a signal. And as you mentioned, equally important not to have a position sometimes than it is to just have one for the sake of talking about it, but it’s a fun model, certainly, a big one, big dude, oil.
Warren: And that’s the tricky part in the environment we’re in is you see backwardation, steep backwardation a lot of commodities in oil, in particular, which when you get that backwardation signal, it’s telling you that there’s a deficit in the near term present on market conditions, you have a deficit in the market. So, backwardation is going to call oil out of storage onto the market. It’s a telltale signal, historically, that it’s a tight and bullish signal for prices. But when you look just beyond the backwardation in the physical market, and that’s the physical market component in our model, you see this massive glut of spare capacity sitting there in OPEC. And to be honest, like, to put it back to what we do and why I think our process is the right way to approach a market like this is ultimately like you said, we’re not going to get caught in the narratives. We’re going to control our emotions and quantify these things that we know historically work. And then ultimately, we’ll discount them where there’s some kind of X-factor sitting there and we’ll discount that as well.
But, like, at the end of the day, we’re just going to add up our components and see what comes out. For now, the model is neutral and it makes sense to me when you see what’s happening in the market. So, I don’t see this as the beginning of an oil supercycle. I think that we could get one down the road. I definitely see the seeds being planted for that. But you’re jumping the gun, I think, if you are calling for an oil supercycle when there’s almost 10 million barrels of spare capacity in OPEC sitting on the sidelines.
Fernando: In this model, you know, anytime you throw together features that don’t have any required correlation between them, the model is going to be neutral when all the evidence is in favor of a certain position. But, like, one of the things that we try to do with 3Fourteen is model explainability. We don’t want black boxes. That’s why we show the model in terms of its components. So if, for example, all you cared about was technicals, the model has had a really strong bullish technical reading for many, many months now, which is kind of the model’s way of saying, “Hey, you know, the price action is looking really good but there’s other things that go into the model.” And it’s all about building conviction with these automated systems to the point where lots of things have to be in your favor for the model to make a correct call.
Meb: Traditionally, what are the defining parameters? Is it price? Does it tend to be sentiment? Does it tend to be fundamentals? Does it tend to be positioning, flows? Is there any, sort of, main lever that, sort of, has its threads throughout most of your models across assets?
Warren: Always price. So, one of the few, like, rules that I’ve developed over my career just playing with data and doing analysis, one of the main rules is that if you’re doing cross-sectional analysis for different securities and definitely for different assets, you should lean heavily on price and not on fundamentals. And, you know, if you think through doing, like, factor analysis, or the Russell 3000, or S&P 1500 or whatever, and you wanted to look at PE ratio or something like that, or price to book, or something like that, you have such, like, a heterogeneous group of stocks that, you know, these different business models, you know, are by their nature going to, kind of, shake out in a certain way when you apply those fundamentals. And so, just like the value factor, when we talk about value stocks and really what that’s done over time, it’s become more or less like a financial synergy, you know, type of sector bundle versus a true analysis of what’s value now because we have these intangible assets and things like that, that don’t make it into that equation.
So, when you’re looking across heterogeneous groups of stocks, heterogeneous asset classes, the one factor that you can’t get that’s always there and I think is always a potential signal for you is price action. So, we always have trend analysis in all of our models. The other big one that we have is the real asset allocation rule. So this is a 17 asset or high-level allocation model. And because of what I just laid out, we do not look at fundamentals in this model. We don’t try and say, for instance, you have a goal model and the goal model looks at real interest rates and things like that, but we don’t look at real interest rates because we’re trying to compare gold in this case to gold versus value stocks versus tech stocks versus commodities versus reads. You know, there are just too many diverse assets here. And so, we stick to our proprietary trend analysis, in that case, and in all cases we definitely have as a component.
Meb: Perfect. Let’s hop over to another one. Give me a preview what another model you guys have constructed is. Feel free to pick and choose or we could even hop right in the asset location. Any one pop into mind?
Warren: I think the real asset allocation model is a good one to talk about and, kind of, give that idea of what we’re talking about with disregarding fundamentals when you’re talking about heterogeneous assets. And so we have, like I said, 17 assets in this model. It really comes down to we’re going to decompose it, trend, correlation, and volatility. And those are the three things we’re looking at when we compare these assets. The highest level first pass is just a trend analysis. And we have a proprietary way of looking at trend. We can talk about in a bit, it’s called Trend Breadth. But it’s basically our way of judging trend across a multitude of timeframes. And that’s a core component of almost every one of our systems. It’s really our take on momentum and I think a more robust way to measure momentum ultimately. So that’s our first pass in our asset allocation model, then we apply hierarchical risk parity, which Fernando can get into, which is a concept coming out of ML, which more or less is a portfolio optimization tactic, which I think really is a differentiator for this model. I don’t know if, Fernando, you want to get into that a little bit?
Fernando: Hierarchal risk parity is the approach that we use for the portfolio optimization side in the real asset allocation model. And it addresses the issues with mean-variance optimization because it trusts the correlation estimates less, essentially, than mean-variance optimization. So instead of using the correlation matrix directly to figure out a weighting scheme, it first clusters all the assets into different clusters, where similar assets are in one cluster and they compete for capital allocation only with assets that are similar to them. And the model basically has this top-down view, kind of the way that a traditional portfolio manager would look at things that got equity bonds, within equity has got large-cap, small-cap, within large-cap, I have different sectors, etc. Basically, the hierarchal risk parity will build its own tree structure on the assets from the correlation matrix and then assign weights so that risk is equally allocated across those different assets.
So the correlation matrix is just used at that first step to figure out the correlation structure of the markets, and then it’s thrown away and you do your risk parity approach to efficiently allocate capital. And what you end up with is much more highly diversified portfolios, and more importantly, like, an output that even if you tweak the correlation matrix pretty substantially, you end up with very similar outputs, which is very different from traditional mean-variance where you tweak that correlation matrix just a little bit and another asset will just pop up to a crazy high allocation because it’s so sensitive to what is very hard to estimate.
Meb: But how much history are you looking at? Is this something that does, kind of, like, a rolling shorter period or is it trying to ingest, like, 100 years of history? What do you feed into this?
Fernando: We do rolling estimates because we’re confident that we don’t, you know, rely on a correlation matrix that much, we want the best estimates from the recent history in the model. So we actually roll a multi-year window and estimate correlations, you know, from the recent data, which is cool because you have a model that is adapting if one asset starts behaving like another asset. And you can imagine, like, we have both Bitcoin and gold in the model. So, it’s keeping track of the recent evidence about how assets move together in order to figure out how it’s going to allocate risk in the portfolio.
Meb: What are they saying now? What are you guys all in on Dogecoin or what?
Warren: Yeah. We haven’t gotten that far into the crypto space. I’d say the most controversial aspect of it has been an 8% Bitcoin position, which it’s held for, really, since inception. And this has been…I remember when we launched the company and had the initial…We revealed the model. I have a buddy of mine who is a little bit older, who ran a hedge fund for many years, kind of, traditional Wall Street guy. And he really pushed back on that and he’s like, you know, “I think that’s not going to fly with most people who have an 8% Bitcoin position.” Again, we were like, “Hey, this is how the model has come down.” And so it’s a kind of, we’re not really Bitcoin apologists, and I don’t want to get into that almost religious debate around Bitcoin. But it’s an uncorrelated emerging asset where there’s a lot of money coming into it right now. That’s the basic facts on the ground as I see them.
The thing that gets, I think, us interested as asset allocators and, kind of, quants is when you run through the little limited history we have, look at how broad portfolio interests, in this case, our model portfolio performed in 2018 when Bitcoin declined by 75%, 80%, and the model portfolio fell by 5.5%. That was the year where we came into that year with what I would say is a max Bitcoin position at 8%. And so, the model was able to…it took its licks, but it scaled the position down because we have the trend component in the model. And then Bitcoin’s lack of connection can fall apart, at least back in 2018 it could, without impacting all these other assets, we could switch over into a more attractive asset mix and sidestep most of that carnage. So, you know, we had an 80% decline, peaked to trough in Bitcoin back in that 2018 period, yet the model was only about 5.5%, 60/40 was down roughly 3% that year.
So, you know, we barely lost to 60/40 in this approach. And to me, that’s the most powerful argument for Bitcoin, this kind of new diversified asset, which, you know, allocators and quants are always looking for. And so, that’s one area of the model likes. We’ve been overweight equities really since the end of last year, through the middle of last year and into this year. And we’ve scaled that position back some. The average equity weighting in the model’s around 38% historically. We’re down to like 43%, came into the year at like 51%, 50% equity. So we’ve scaled that back. Real assets still has a big weighting. The model likes Bitcoin, like I said, likes commodities, and it’s still, kind of, shunning bonds, the bond position’s down at like 23%.
Meb: Is the equities scaling back, is that due to equities looking worse, or simply other things looking better, or what’s the, kind of, driving force behind that?
Warren: I think it’s been a little bit of rotation within the equity component. So we were really overweight small caps coming into the year and then you saw, kind of, this really powerful small-cap rally. And then that, kind of, stalled out. So you saw rotation out of small caps, value got a little bit but mostly it came out of small caps and, this might disappoint you, came out of emerging markets as well in the model into, really, commodities and real estate were the two spots that picked up the flows primarily in cash. There’s a bit of a cash bill in the model as well, which is, again, something we’ve pointed out before is that our cash component, if you’re building a model, historically, even Ned Davis, like if you build a model, you have a switch where you’ll be switching into cash. If you switch into T-bills, and you’re backtesting against the falling interest rate environment, you’re going to get this massive tailwind. I don’t think people realize the kind of tailwind you get on your backtests switching into T bills and a falling interest rate environment.
Well, that’s not…Again, this is an example of the past isn’t elevating the future. We’re not getting that tailwind moving forward. So when we build our models and you’re switching to cash in our model, it’s a static ones position. So, you know, we’re getting no uplift. It’s a defensive risk management position that has been a place where the model has, kind of, moved some of those small cap and emerging market equity allocations into through the first four or five months of this year.
Meb: Should we go yield optimizer or do you want to talk about should we just all buy a bunch of beaten-down energy bonds or what? What should we do?
Warren: Yield optimizer is, like, yeah, again, this is our take on how do you deal with the obvious dilemma that you have no yield in the current environment? And so, we took 13 different assets, all income-producing assets. And this, kind of, can fall into a few of our views on the market, which I think are unique. So we took 13 different income-producing assets, including this, would be things that might not be traditionally thought of as income producers, like the energy sector. We pulled them into the same framework as we have for the real asset allocation model, Trend Breadth overlay, and then using our same optimizer. And we rotate around within that income-producing assets to try and generate some yield. And so this is, kind of, our solution, again, at trying to…There’s all these clients and investors out there that want to generate income. And so this is our one way to solve that problem or get at solving that problem.
I think that the interesting part when we came up with that model for a lot of folks is to see energy in there as an income asset. And this is something that, back in April of 2019, the energy sector actually became the highest yielding sector in the market for the first time ever. It stayed there really until present day. It still about 100 basis points above the utility sector on yield. So this has become a high yield sector. And this is something that’s, kind of, focused on commodity and energy space that I was watching prior to the pandemic. My interpretation is that the energy sector has become what I call a short-duration investment or short-duration equity sector. So, investors were tired of giving their money to management teams and trusting them to drill new holes in fund CapEx. Instead, they’re saying, “We want that money back now or as soon as possible in the form of dividends.”
So it’s forcing this discipline onto the energy space. And it’s creating this, I think, interesting pocket when you look out across all the different segments of the market, whether you’re trying to create income or just find a diversified portfolio, you now have, kind of, a short-duration option there in the energy space. You can get duration…there’s plenty of ways to get duration or long-duration exposure in the market, whether it’s big tech, or bonds, anything with interest, really high interest-rate sensitivity, it’s easy to get that long duration. But short duration is, kind of, a little more tricky. So energy, I think, it’s giving you that yield, and that yield is reflected in the yield optimizer. And it’s also giving you diversification. And it’s got a different set of drivers than what’s happening in other areas of the market like big tech and in long-duration types of sectors.
Meb: At the beginning of the piece, you had a great quote from Cormac McCarthy’s, “The Road.” And I just finished “Blood Meridian.” “The Road” is one of the few books I’ve ever read cover to cover in one sitting. It says, “Man, we’re starving now.” And the boy said, “You said we weren’t.” The man said, “I said we weren’t dying. I didn’t say we weren’t starving.” And that’s what it feels like. I mean, I know a lot of people, interest rates to them, I mean, obviously, I think it pays to think in the real terms. But with junk bonds and some of these hitting lowest levels ever, you start to get a little nervous. It’s fun to look through your charts. And hopefully, we get to post a bunch of these to the show notes, listeners. So, definitely check out mebfaber.com/podcast for some of the charts on some of these because they are really insightful. What should we hop over to next?
Warren: Inflation. As I see it, I think that’s, like, the hot topic in the markets right now. Is this inflation or are we jumping the gun? And our view is that it’s not inflation, basically.
Meb: Interesting. What is it then?
Warren: We are, kind of, trend followers by nature. We’re not trying to be contrarians, just for the sake of being…And I feel like that’s, like, some kind of, like, media-contrarian thing to admit that we’re not going to be contrarians because everyone wants to be a contrarian. Everyone wants to buy the thing that’s down and sell the thing that’s up. It’s like that human nature is the way I see it at least. And a part of that or at least some aspect of that, I think, is this desire to see inflation for everybody else and to, kind of, pull anecdotes out of the world and say, “Aha, here’s inflation.” And then also, of course, criticize the Federal Reserve. You get a lot of brownie points for that these days, it seems like, or at least get a lot of retweets, and likes, and things like that. So, those things, kind of, all have conspired in my view to make…That’s really the consensus view or at least in some circles that inflation is here. It’s not something to worry about in the future. It’s like we’re seeing lumber prices. We’re seeing semiconductor shortages. Used car prices are going up. There’s a lot of different examples that you can point to, oil prices, breakevens, whatever.
And I think ultimately, those are all supply chain issues or mainly supply chain issues. You can trace almost every one of those instances back to something on the supply side, some disruption on the supply side. And quite frankly, the way we look at it is not to…We don’t win an award for finding the most interesting anecdote that proves inflation. Like, we are going to stick to the metrics that the Federal Reserve is watching because, ultimately, that’s what matters is that the Fed moves quicker than just consensus. So we look at the CPI, we break the CPI down. And what we see is this massive dispersion within the CPI. The median correlation of more than 200 CPI component parts to the CPI is now negative for the first time ever going back to the 1950s.
So that tells you that we’re going through this weird period of the massive dispersion within the economy where you’re seeing pockets of inflation, living alongside pockets of deflation, or at least ostensible inflation and deflation. And really when you see that dispersion and then you see…You can, kind of, look at the slack in the economy, the output gap, labor participation rate, unemployment rate, all these things, and then to go back to the first part of the conversation, where we talked about 9 to 10 million barrels of spare capacity in the oil market, what you see really when you start stitching this all together is a world that actually the Fed is correct, I believe, that this is a supply chain issue. There is a lot of spare capacity out in the system. And the bottom line is adjusting monetary policy at this point to try and stamp out these anecdotal inflation years would be premature and probably ineffective in my view.
And I don’t think that that’s a real popular viewpoint as much as everybody is…It would be a lot easier and a lot cooler for me to come on the Meb Faber podcast and, you know, just, like, lay into the Fed and get brownie points from everybody for that. But that’s just not how we see it. We see it as a supply chain issue with a lot of slack in the economy at this point still. As investors, that tells us that the Fed is going to stay on the sidelines for a while. And by our estimations, given the fact that they said they’re going to let inflation run hot for a while, you know, and bottom line, as you look at their 2% target, we’ve done some work on where would breakevens have to go. If you saw breakevens, like 5-year breakevens go to 3% to 4%, then you might start worrying. But we’re still 100 basis points away in our estimation, whether you’re looking at 2-year, 5-year breakevens before the Fed starts getting nervous on structural inflation. So all those things come together and we say, “Look, the Fed’s not going to end this party anytime soon.”
Meb: So a lot of people as part of this, you hear a lot a surprise, the shiny metal is not at 5,000 or 10,000 or something right now, gold. How do you guys think about gold? You call it the chameleon asset. What’s the approach and how do you guys think about it right now?
Warren: It’s very similar to how we did the oil model where we have a four-component model, first off, that sits on the…this is the sterling answer. You know, we have a four-component model, that’s four components, our trend positioning, real interest rates, and asset allocation, attractiveness. We wrote an entire pub on that, a report on that, where we outline each one of those factors and, like, kind of, how we stitch them together. And that’s, again, kind of, relying on Fernando’s expertise and our ability to cross-validate all those components. But the reason we call it the chameleon asset is because when you really start testing gold, it goes through different regimes where different things matter and it’s really hard to find a stable set of indicators across a long arc of history that call it the gold market.
And so that brings us to today. Like, why is gold now reacting to everything? And, you know, I could point to, in our model, would say real interest rates. But basically, we’ve had CPI inflation measures, more or less pinned. I mean, that’s going to change. So, we’ll see how that reads through the model. But they’ve been pinned, whereas nominal rates have just skyrocketed straight up from August 4th onward. But an interesting thing, when you look at gold, it topped out on August 6th. Gold tops out August 6th, rates bottom August 4th. It’s pretty easy to me, you don’t want to overthink it, that gold is reacting to this rise in interest rates. And, again, it brings in this duration argument. Gold, in my view, is an infinite duration asset. So, it is ultra interest rate sensitive. And that I think is the most powerful force operating in this market right now is as the reopening comes closer, I think it’s just going to be a force we’ve never seen before. It is setting the table across assets and how assets are performing relative to each other. And so, gold is caught up in that. it’s an infinite duration asset. And I think it’s having a hard time moving higher with rates skyrocketing like this.
Meb: More than any of the assets, excluding crypto, probably elicits a binary response. People are either total gold bugs or absolutely hate it. Usually, there’s not a whole lot in the middle. Other asset classes, I don’t feel like really generate that sort of vitriol, but crypto seems to be in the same ballpark. Is this value trade going to continue? What about the U.S. dollar? The U.S. dollar seems to be plumbing the lows from 2018 and 2021. Are you guys pointing towards further weakness there or is that story played out?
Warren: Again, our model for the dollar is highly reliant on Trend Breadth. And so, that’s a concept we haven’t talked about but it’s a proprietary technical indicator we’ve created. And it lives in a lot of the indicators we’ve made. So, the bottom line, I give you the…I don’t want to bury the lead on the dollar. The dollar model went from a buy to a neutral, you know, recently. So it’s another neutral model for us. I think we’re in this transition point. And I think, as you point out, where the dollar is close to long-term support. And so, if we break that 89, 90 level, I think that’s going to be a pretty important tell for how things go. And the fact that if you ask me just without looking at the model what I think is going to happen, I think we end up breaking those levels and the dollar is weaker over the longer term. And that’s a result of the fact that government spending as a percent of GDP has increased by 50%. So we went from basically a long-term spend of 20% of GDP up to about a third of GDP. And that’s going to be, I think, a structural shift higher. It’s one of the themes that we’ve, kind of, laid out for our clients is how to play this structural shift higher in U.S. federal government spending.
And we’ve outspent almost every country combined in stimulus for the pandemic. So, I do think the dollar is going to suffer ultimately from that kind of fiscal spending. But at this point in time, from a timing perspective, our model is still neutral right now. So we’re kind of in this transition point in the dollar here and now. Again, the big driver of that model and a lot of ours is something called Trend Breadth. And I’ll give you a quick rundown but I think Fernando can probably explain how we’re able to crunch big data problems in a way that other firms aren’t. And this Trend Breadth is an example of that.
So what I did in NDR for a long time was I would create these different indicators. I looked at…run regression trend lines through time series and measured the slope and the change in the slope of these regression trend lines and drive different signals from those. And I always wanted to do, like, a whole look at it from so many different timeframes. So when we got together, Fernando and I came up with this Trend Breadth concept, where we were running linear regressions across dozens and dozens of timeframes for each asset, and then deriving information for each one of those regression trend lines so we can look at is the slope positive or negative? What’s the residual for each one of those regressions? Percentage of time frames that we are positive or negative and how that number is changing, so a second derivative. So, there’s tons of information we can derive from just that one indicator.
And then when we do cross-sectional comparisons, it allows for some…I think, a really fine detailed look at the trend of an asset in comparison to another one. And so, if you’re comparing it through traditional momentum where you do, like, 1 in, like, a 12-month momentum with minus 1-month momentum, even if you’re blending that with a 6-month momentum, you only are looking at, like, 4 data points, ultimately, to derive your signal there, where Trend Breadth is looking at using every single data point for the last whatever your lookback period is as you go back to three years and be calculating trend lines all the way back there. And to me, it’s much more sensitive at scales in and out positions much better than your traditional momentum factor while still picking up on some of the same dynamics.
Fernando: And this is an example I think too where the toolset for machine learning, kind of, helps in developing new indicator ideas. It’s like when Warren talked to me about, “I find trend lines across certain time periods to be useful for its strategy work.” And then we were talking about, “Okay, what do we want to do here?” And it’s like, well, let’s try medium-term and longer-term. And then we were, like, “Maybe we should incorporate some mean reversion and look at the shorter term as well.”
And basically, as we were going through this exercise, what we were doing resonated with this algorithm for machine learning world called random sample consensus, which is all about taking subsets of your data set, and running regressions against the subsets, and then making a conclusion about the percentage of those subsets that coincide with a particular conclusion. So, like, in our case, Trend Breadth is basically saying, “Let’s run a ton of trend lines against a bunch of different slices of historical data.” And what percentage of those would say that we are currently in an uptrend in this asset? And that’s essentially in a nutshell with this Trend Breadth concept is when Warren talks to me about strategy work, certain things vibrate in ML world that can come in and help us come up with some novel techniques.
Meb: I’ve always been a big fan and proponent of interest in the breadth world. I find that part fascinating. It often in my mind, kind of, signals, perhaps some areas that may be overlooked or it’s hard to certainly argue the results that it’s spitting out. You guys do talk about stocks individually a little bit as far as lists. I saw some for talking about reopening and some ideas. How do you think about individual securities as a part of your overall offerings?
Warren: Broadly speaking, again, we’ve been bullish, and like we said, we, kind of, dialed back our equity exposure but we’re still overweight equities and we’re still more or less bullish. Within our recommendations, the big trades we saw unfolding from a macro perspective last year was the reopening trade. Obviously, that’s kind of consensus. And we were recommending stocks that I would say were the tip of the demand sphere. So we had some airline plays, mainly domestic base routes, hotels, and oil refiners. And so, for the most part, we close those out here in April and most of those positions have run pretty well. And I think that really what you’re…Not that the reopening is not going to be a huge boom but I mean, these airlines are now…Their enterprise value is 25% above what it was when we were recommending it last year. So, a lot to be done. So the expectations are baked in.
So then the big theme we’ve been digging into outside of our quant models, our macro theme would be entrenched interest. And so, again, going back to the government spending that we’ve seen jump up here, there’s a lot of nodes and it’s, kind of, traditional Wall Street to go through and look at the infrastructure spending bill, for instance, or the stimulus package and trying look at how we’re backwards from how will this bill help certain companies? You know, like, where will the money go? And then trace those bouncy balls to…And that way…And I think that’s a valid way to do it, but it’s so early to us. I think a better way to do it right now and to get positioned and probably just as effective even at the end of it is to just find those companies that are already doing business with the federal government.
So you have this belief on our side, our thesis is that government spending is going up structurally in the next few years. And so, how do you play that? And our recommendation is entrenched interests. And this was a list of stocks that we came up with using, kind of, two different routes to get to this answer. So the first route we did is we went through the largest contractors, publicly traded companies that contracted within the last five years with the federal government. We compiled that list. And then I turned to Fernando, and I had him basically use some of his ML techniques to work through the company filings and find companies that were highly exposed to the federal government, renewing contracts, stimulus, things like that. So, I don’t know if, Fernando, you want to describe that side of entrenched interest analysis.
Fernando: It’s just a cool example of the kind of stuff that natural language processing if it’s in your toolkit, we can do something like scan through thousands of annual filings and come up with a set of phrases that we’re going to look for that are basically hallmarks of companies that do business with the federal government and do a factor analysis that is pretty unique. We’ve talked in the past about wanting to do more along these lines. But imagine if you could come up and do factor analysis on particular phrases that appear in annual filings or in earnings calls, and you can build a portfolio of sentences and do analysis on that. It’s a really cool area of research that we, kind of, want to explore. And this is, kind of, like, the most basic implementation of it, just looking for companies that if you look at their Salesforce leads database, the federal government is on there. They’ll hit them up when they see that they’re flush with cash. It’s a standard startup thing to do as well. You watch for who just raised funding and then you go and try to congratulate them on their funding and then try to sell them your software. So, I have a feeling that these companies, you know, are going to be, oh, you guys want to spend on infrastructure? Well, we’ve got some infrastructure ready for you here, you know.
Meb: So how do you guys work with clients? Tell us the business model or are you targeting investment advisors, individuals, institutions? Is it a subscription fee? Do you guys do custom work? What does the business look like?
Warren: We are an institutional research provider. And so, we work with institutions, asset managers, RIAs, hedge funds, family offices. And you can, more or less, subscribe and get our research. You can get our core models. And we update daily on our website, all of our library of charts on our website. And we publish weekly. So you get a weekly report in different formats. And then we do have everything from fully custom models to what model portfolios which are a tweak here or there of our current model offering. And so, we have clients who are following and creating their own version of the real asset allocation model. So they can put their own stamp on what we’re doing. So we give them this framework, and then they can go in there, and pull the levers that they want to, kind of, make it fit their benchmark. So we have clients that…we want infrastructure in there. We want certain healthcare components in there to be represented, or not represented. ESG mandates have changed things in ways that you wouldn’t really even imagine at a certain time. So they might want something changed on that front. And we can present that for them and build that, kind of, custom version of what we’ve already, kind of, demonstrated to the market. So, that’s, kind of, the basic business model.
Meb: Is it mostly advisors or is it big institutions? Is it a mix?
Warren: It is right now. So everybody likes to have, like, what did you learn from starting a business type of advice. So, five months into this business, what have you learned? And the idea, and what I sold Fernando on, is that I think this RIA side of the market is kind of underserved. And there’s money gravitating in that space. And so we want to price this in a way that they can utilize it but obviously, we want to have high level, really high-level institutional research. And when we launched really, I guess, maybe our research resonated more with the high-level institutions less than the RIAs. So we had a lot of bigger hedge funds and asset managers sign up. And we are working with RIAs, we have some, but I expected them to just be coming in droves. And what I’m learning is that’s a set of different clients than a hedge fund, for instance. And so, there’s trust that needs to be built and a little bit more explanation of what you do and how you fit into the market. So, it’s an education process really. I would certainly hope to have a lot of RIA clients in the years ahead.
Meb: What’s the dialogue been like with the clients? Is the back and forth conversation, anything in particular on their mind that surprised you or you’re like, “Dude, everyone will not stop asking you about ESG or cannabis or Dogecoin?” Whatever it is. Is there a narrative that everyone’s interested about or worried about?
Warren: Bitcoin is like…It blew me away the amount of Bitcoin interest early on that, like, even just saying we’re starting the firm, “Are you going to be covering Bitcoin?” And I do think there was this…We’ve passed this point of this phenomenon where, you know, it was career risk to consider having Bitcoin exposure. And I do believe we’ve passed into where we flipped it. So it’s now, kind of, career risk if you don’t have Bitcoin exposure. And that nervous energy is coming back from a lot of folks who really neglected that space for many years and are now, kind of, trying to play catch up and figure out what’s going on. There is a lot of energy, unsurprisingly, around Bitcoin and cryptocurrencies. And we’re working on a standalone Bitcoin model right now. Again, we’re taking our time because as Fernando said, we’re not just going to throw something out there. We’re taking our time and trying to build it right. And this is a pretty young market. So you want to do it right. So Bitcoin is big. And in the inflation debate, and oil and energy is always top of mind. I think folks are wondering, are we going to have a huge spike in oil prices? And that’s come back as well.
Meb: As you look to the horizon, as you build out your business, anything on the brain that you guys are working on? I mean, these reports are so thorough. It’s a lot of fun to read. What are you guys diving into? Can you give us any peeks behind the curtain that you haven’t published yet?
Fernando: I think you can expect a lot more work on our stock selection system. Like, our latest report, we just put out basically what we consider version one, which is taking Trend Breadth, kind of, as we’ve talked about it and used in our other models, and just done a typical, like, factor backtest of the Trend Breadth system, and found some encouraging results. But we think there’s a ton of value to be mined in digging deeper into that, doing some, like, portfolio optimization on our stock portfolio to come up with some interesting techniques for figuring out how to improve on an equal weighting of stocks that rank high on Trend Breadth. So that’s kind of an open, exciting area for us that we want to dive into more.
Warren: I echo that. I mean, that was going to be one of my top of mind areas that we’re digging into. We’re working on a Bitcoin model, kind of, in tandem with that. To me, as I think about it, strategically, this regime definition, this problem is something that we’re doing some more and we’re trying to apply machine learning to it. And it sounds, kind of, opaque. But, like, you have this, kind of…Everyone wants to fit each period that you’re into, into like, kind of, a simple, clean-cut regime. We’re in reflation right now. We’re in deflation, or stagflation, or, you know, different, kind of, simple quadrants of the business model. And so, I don’t want to get too far into it because we haven’t published the work yet but we’re trying to put our own spin on and define those different environments in maybe a different, more flexible way and, hopefully, more accurate, and something that fits on to it I think is what happening right now. Because I know this is such a novel time that you can get your signals crossed really easily if you’re trying to just fit this period of time into those, kind of, historic analogues without thinking it through very deeply. So, that’s something I’m thinking through as a strategist.
Meb: Always uncharted waters, right? Like, that’s the excitement of our business is every day brings something, well, in the last two years, weirder, I would say, and new as well. To the extent, you guys have done some on your own. What’s been your most memorable investments? You guys get to pick it goes first. Good, bad, in between. Anything coming to mind?
Warren: Okay. I’m going to give my bad one because I’m not going to call the person out because somehow I always remember my bad investments better than my good ones. So, when I first started investing, this little, he was called Nastech. It was supposed to be a nasal delivery system for insulin, I believe. Someone was pumping it that’s still around in the financial media at the time. And I was just, like, in law school and I threw a few bucks in and it went all the way to zero. And that’s the only time I’ve ever rode an investment to zero, so it sticks out in my mind. But it was a good lesson and it scared me off of biotechs for the most part.
Meb: I mean, that went back to the ’90s, Inhale Therapeutics. I remember, like, one that was doing that, God, 20 years ago, at least talking about it. I have no idea what happened to them, listeners. It may bring back some memories. Scars usually leave a mark. That’s for sure.
Warren: Yeah. You learn a lot more from your mistakes than your victories. And that was…Like, for me, I didn’t do any, kind of, traditional business school or anything like that. So, really just losing money when it was, man, enough to me but it wasn’t really any money. It was a good lesson. For a good investment, I think it’s nice to be right for the right reasons. I bought a lot of Williams companies last year at the lows, and that thing’s got up quite a bit. And so, I was right for the right reasons. I just knew that that was a screaming cheap stock. And so, I still own really most of it. So, riding that thing, it’s like a 7% dividend yield. That’s been a good investment, I guess.
Meb: Seven percent dividend yield, gone are those days. That’s pretty rarity in this world. I think most people are happy with about 2%. S&P is not even 2% anymore. I don’t think, I think it’s 1% something.
Warren: That’s right. Those are two that pop up. And hopefully, we’ll be talking in, you know, three, five years, and the best investment will be starring 3Fourteen Research.
Fernando: I’ve got a good and a bad one just in recent history. Last year during the turmoil, I dramatically increased equity exposure in my discretionary portfolio about a week-and-a-half before the bottom. So that’s, like, the good one, right, ramped it right back down towards the end of the summer thinking that, “Okay, this is enough.” So that’s, kind of, the good and the bad all in one year.
Meb: Got to pay more attention to your breadth rankings. That’s for sure.
Fernando: It’s good to have a model. I mean, that’s actually the moral of the story. It’s good to have a model, not get skittish.
Meb: What’s the best place to find you guys? How do we keep up with what you’re doing? If somebody wants to check out some reports, maybe sign up, where do they go? How do they find you? What do they do?
Warren: 3fourteenresearch.com is the site. You can enter your email and get the trial access. You can reach out to me on Twitter or to 3Fourteen Research on Twitter. We’re pretty easy to find. So, if you reach out, we’ll be responsive.
Meb: Awesome. Gentlemen, Fernando, Warren, thanks so much for joining us today.
Warren: Thank you, Meb. I really appreciate it.
Fernando: Yeah, thanks for having us on.
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 feedback at firstname.lastname@example.org. 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.