Matt Larriva, Vice President of Research and Data Analytics at FCP, joined us on the podcast to talk about current migration trends and a new way he developed to forecast cap rates.
Matt Larriva leverages data science and quantitative methods to enhance asset selection, management, and disposition across real estate investments. His active research focuses on multivariate timeseries analysis as it applies to cap rate forecasting (Vector Error Correction Methods as models of US Real Estate Markets (working)), and random forest methods as they apply to submarket rent-growth in unstructured data (Semi-Greedy Construction of Oblique-Split Decision Trees (2019)). His work on domestic net migration has been featured in Bloomberg News and Fox Business News.
Prior to joining FCP, Matt worked at Green Street Advisors as the head of the US and UK Quant teams, managing the data efforts in the sell-side real estate research firm. While at Green Street, he focused on REIT trading strategies, data productization, and analytic enhancements. Matt completed his undergraduate degree in economics at the Wharton School of the University of Pennsylvania. He holds a Masters in applied statistics from UCLA and is a CFA charter holder.
FCP is a privately-held real estate investment company based in Chevy Chase, MD currently investing a $755 million commingled fund focused in the Mid-Atlantic, the Southeast US, and Texas. Since its inception in 1999, FCP has invested in or financed more than $8.3 billion in assets and currently owns and manages four funds totaling $3.6 billion in assets.
Learn More About FCP And Matt’s Research
*If you’d like to learn more about how ReAllocate + Mariner Wealth Advisors can help you build a roadmap for your real estate investments head to — BuildMyRoadmap.com.*
Subscribe To Podcast
Speaker 1 (00:00):
All opinions expressed by Adam, Tyler and podcast guests are solely their own opinions and do not reflect the opinion of RealCrowd. This podcast is for informational purposes only and should not be relied upon as a basis for investment decisions. To gain a better understanding of the risks associated with commercial real estate investing, please consult your advisors.
Matt Larriva (00:22):
Now what we know about demographic trends is that they tend to persist for a very long time. And when they slow down, they do so very, very gradually. But things like pandemics can really put a stop to them quite quickly. So our specific question was, are people still moving to the Sunbelt?
Adam Hooper (00:51):
Hey, listeners, Adam here. Have you ever wondered if you’re investing in the right real estate deals? What about if you’re making the right decisions for your overall financial health? Over the last seven years of running RealCrowd, the number one question we received from investors is, should I invest in this deal? Well, we’re excited to announce that we can now help you answer that question. Through our sister company, ReAllocate, and through ReAllocate’s partnership with Mariner Wealth Advisors, you can now have access to teams dedicated to helping you build a real estate portfolio based on your personal investment roadmap and financial goals. If you’d like to learn more about how ReAllocate can help you, head to buildmyroadmap.com. Again, that’s buildmyroadmap.com. Hey, Tyler.
Tyler Stewart (01:33):
Hey, Adam, how are you today?
Adam Hooper (01:35):
Tyler, you know it’s 2021. 2020 is officially behind us and we’re back in the studio.
Tyler Stewart (01:41):
We are back in the studio. And we have a good one. We brought on Matt Larriva, who is the vice president of research and data analytics at FCP.
Adam Hooper (01:50):
Yeah. Man, again, I think we hold ourselves to a pretty high bar for the kind of shows that we like to have on here and the guests we get on, and Matt just flew over that. This is an absolutely fascinating conversation. They’ve done some incredible research. We talk today about some of the migration research that they’ve done. And again, we’ve talked about that a lot on the show over the last year with this kind of talk about exodus from the metros and people out-migrating and where they’re ending up. So we took a deeper dive on some of those trends that they’ve seen over the last 70 years and compare and contrast what was going on pre-pandemic, what they saw during the pandemic. And then we talked about what is I think super fascinating and very early in their research, is forecasting cap rates with some new surprisingly simple metrics.
Tyler Stewart (02:39):
Yeah, yeah. And normally when we have a deep dive on an episode, we’ll say, take a couple listens, take notes. On this episode, it’s take notes and use Google as your friend to look up some of the research that Matt is talking about. This was a very, very deep dive.
Adam Hooper (02:57):
Yeah. What I thought was interesting was the persistence of some of these demographic trends and migration patterns. How we’ve seen COVID in this current crisis as an accelerant of some of those trends. And talk a little bit later on about some of the things that might start some course correction in those trends or shifts in those trends. But once those patterns are in motion, they tend to stick with that motion. And he said he’s never one to bet against the trend. So really, really interesting conversation there. We talked about some of the factors that they’re looking at, some of the different data sources that listeners can use to start paying attention to some of those factors themselves. Again, just great, great conversation and really happy to kick off the year with a conversation with Matt today.
Tyler Stewart (03:44):
Yeah, this is a great kickoff to season five.
Adam Hooper (03:46):
Absolutely., I think that’s probably enough of us talking about it. We’ll just maybe get to it. Listeners, as always, as we’re here in season five, we’ve been doing this for a while now, we’d love to hear your feedback on things that we might do differently, what you’d like to hear us cover or any new ideas that we can introduce into the show, or if you just want to send us a note and tell us that you love what we’re up to, that’s okay as well. If you want to do that, send us a note to firstname.lastname@example.org. With that, Tyler, let’s get to it.
Adam Hooper (04:19):
Well, Matt, thank you so much for coming on the show today. This is the first recording we’re doing here in 2021. So, we’re excited to dig in on some of your recent research and hear what you guys have been up to.
Matt Larriva (04:31):
I’m really excited to be here. I’m excited that you selected my team and myself to launch 2021. I’m excited to tell you about what we’ve been working on.
Adam Hooper (04:39):
We’ve got a high bar set for you, so, have a goo chat today. So before we jump in, tell us a little bit about your background, your history in the real estate space. How did you get into the industry, what excited you about it, and a little bit about what you’re up to with FCP?
Matt Larriva (04:57):
Yeah. My history has always been in real estate and basically real estate finance. So, I went to Wharton and I knew that I wanted to focus on real estate. So I concentrated in that. Unbeknownst to me, I was graduating into what would in retrospect become the worst period in the history of real estate, which was 2009. So, obviously, it was a little bit tough to jump straight into real estate, but I was a pretty competitive poker player and all the trading shops at the time thought that poker players would make excellent traders. And in large part they did. So, I prop traded for a few years to bide my time until real estate kind of unlocked. At which point I moved back to Southern California and found Green Street Advisors, who does a lot of sell side RIET research and now consulting and some private market research as well. Really intelligent place to work and cut your teeth in the real estate industry. Basically, they call it getting a real estate PhD.
Matt Larriva (05:59):
So, I went to work for Green Street Advisors in their quant team, and just was fascinated with the intersection of quantitative financial work and real estate, partially because real estate by and large is such a fundamental asset class. So people price things based on comps that they’ve seen, feelings that they get about a neighborhood. We all know that local player who’s just been in the neighborhood forever and knows how to get zoning done. And that’s all real boots on the ground type work, which is still important. But largely real estate hadn’t had that kind of renaissance in terms of quantitative analytics that something like a bond market had. So, it was a really exciting time to be in the quant space in real estate.
Matt Larriva (06:44):
And when at Green Street, I completed my CFA and then went to UCLA to get my master’s in applied statistics. So, after that, I found FCP, so I shifted to the buy side of real estate. And I’m now with FCP. They brought me on to basically build out a quantitative analytics team for them, which is basically every data scientist’s dream to take somebody from zero to one in a space that’s also sort of coming of age. So, it’s been a really fruitful relationship and I’ve done a bunch of interesting studies since I’ve been there. FCP is a owner-operator and a private equity fund, investing throughout the Sunbelt mainly in multifamily, but also doing some ground up residential development as well as some adaptive reuse office and ground up office.
Adam Hooper (07:39):
I think we’ll talk in a little bit about why the focus on the Sunbelt. You’ve done a lot of research, two topics that we’re going to cover today, again, migration patterns and then some research you did with Peter Lindemann on forecasting cap rates. We had Peter on the show earlier last year. Again, another fascinating conversation. For those that are listening, I would suggest going to check that one out. That was kind of an early take on where the industry was at in the pandemic times. Now in early 2021, vaccines are starting to come out, we’ve got some better understanding maybe of where the industry stands relative to early pandemic days. Before we get into the research, maybe, what have you seen I guess towards the latter part of last year, and what are you looking for this year as it relates to this health crisis’ impact on real estate generally as your outlook at FCP?
Matt Larriva (08:37):
It’s a fascinating time for the economy as a whole, but especially for the real estate industry, this tragedy brought about a real paradigm shift for us. When you think about who was talking about work from home prior to this, it was on the margins. It was people who couldn’t access an office. But this really gave us the ability to work from home full time. It also sort of changed the demand pattern of what we needed out of a residence. When you think about now trying to shop for an apartment or a home, you think about it and then you have to add one or two bedrooms per person because everyone needs an office, right? It’s also got implications for how we’re going to travel. It’s got huge retail implications.
Matt Larriva (09:27):
So you’ve got the sort of winners and losers of all of these paradigm shifts. On one hand, you’ve got the winners in terms of square footage needs that are going to be residential. And then you’ve got, it might be too hard of a word to say losers, but you’ve certainly got the people who have a tougher road to hoe and that’s places like office who are going to need less square footage. Then you’ve got on the other hand retail that’s really taking it on the chin, but who’s really benefiting from that? E-tailors, which means industrial cold storage, who’s going to do really well.
Matt Larriva (10:14):
Then you have to think beyond that, as we’re all sort of living a more wired life because we to an even greater extent, can’t see each other in person, we’re going to rely on the obscure sort of sectors of real estate, like tower reads, data centers. So, it’s a pretty fascinating time.
Matt Larriva (10:33):
That said, we have always known it to be the case that industries don’t categorically and uniformly go up or down. So when we talk about the forecast that office is going to have a demand contraction of about 10 to 15%, and we’re citing Green Street in that, you have to realize that that’s going to hit different tiers of office completely differently. It’s going to hit different segments of the market completely differently, and different locations.
Matt Larriva (11:06):
So you imagine the difference between a high rise where everyone’s in an open floor plan, which is sort of the epitome of density in terms of workers per square foot. It’s the epitome in terms of transmission of potential viruses. And then you think about a very broad, sprawling, one story, maybe two-story but with stairs office space out in the suburbs, or a redevelopment from what used to be a factory. And then you’re thinking in terms of cars that can drive straight up and park without having to take an elevator, people who can have enough space for offices. And a lot of natural light, and a really easy ventilation system to deal with because you’re sort of only one or two stories. So, I use that as an example because it really highlights that for the “winners and losers” in this, it’s going to affect all of them very differently.
Adam Hooper (12:02):
I put you on the spot there, it’s a pretty broad brush to paint the whole industry, right? But yeah, I agree, I think there are areas that have been more challenged, and a theme that we’ve talked about on the show quite a bit over the last year is COVID as an accelerant, this crisis as an accelerant of some of those trends that were already in place. Office, again, I still think is one of those wildcards. The balance of working from home and remote work paired with the need to expand beyond the 125, 150 square foot per employee that we were densifying into, those opposing requirements will probably balance somewhere out in the middle, but still kind of to be determined on that one.
Matt Larriva (12:44):
I think that’s totally true, and a point that isn’t made often enough, perhaps, is that going into this, we were at the highest densification level in the history of recordings in the US. So people were just crammed into their office space. And this concept of open floor plans was really kind of used to make the average employee take up even less space. So, we had one end of the pendulum going into this. Now we’ve sort of swung to the opposite side, where not only are we de-densifying, but we sort of need to from a health perspective.
Adam Hooper (13:19):
Yeah. So, one of the trends that, again, that we talked about a fair bit, and I know your research goes back pre-COVID, quite a ways back pre-COVID, is the paper that you guys put out around migration. So why don’t you tell us a little bit set up what that study was, what you guys were looking for, and then we can dig into some of the findings on this migration study that you guys performed?
Matt Larriva (13:43):
Yeah. So what we wanted to do was figure out if the trends that had existed for the last 70 years were continuing, or if this paradigm shift was going to upend those trends. Now what we know about demographic trends is that they tend to persist for a very long time, and when they slow down, they do so very, very gradually. But things like pandemics can really put a stop to them quite quickly.
Matt Larriva (14:12):
So our specific question was, are people still moving to the Sunbelt? Now, this was a phenomenon that existed since the 1950s. If you ever look at density graphs or graphs of where the people in the US are living or the so called halfway points, they were all up in the northeast until about the 1950s, right around 1950 exactly. And out of curiosity, do you guys have any theory as to why that shifted right after 1950?
Adam Hooper (14:40):
I cheated, I read the paper, which we’ll have in the show notes here for everybody that wants to read along. But yes, why don’t you let us know.
Matt Larriva (14:50):
Exactly. So it was basically just the advent of air conditioning.
Adam Hooper (14:53):
Which is crazy that that would have such a huge impact on migration. I doubt many people would normally go to that as their first cause for why that migration occurred.
Matt Larriva (15:03):
Yeah, it’s totally crazy. But yeah, the creation of the air conditioning, which happened right around the same time that the ice machine was invented, sort of opened up an entire area of the country that people weren’t, not allowed to access, but wouldn’t comfortably access to be sure. After that point, you see this consistent move to the Sunbelt, as we call it, or to the southwest, more specifically, in the south, properly. Now, that existed since the 1950s. It really started to kick into full gear around the 70s, and then it tipped around the 90s. And so, the South now is grossly the largest part of the country if you break it into quadrants, or even into three parts.
Matt Larriva (15:53):
So, people are moving down to the south. They’re doing it for a number of reasons. It’s got more space. Generally, it tends to be a little bit more worker friendly, a little bit more livable in terms of cost of living, and then a little bit more business friendly. So we’ve known these trends to be in existence for a very long time. But like I said, you get a pandemic and that can really force people to reevaluate their preferences. So we had to make sure that our investment thesis was still intact basically.
Matt Larriva (16:23):
So, I set about trying to find good migration data. And the number one source for this is the US government. They track migration outstandingly well. County to county, in-migration, out-migration, international migration, domestic net migration. But they do it with about a two year lag, and that’s understandable, it takes a long time to compile that data. But we can’t wait two years, so we had to look for another source.
Matt Larriva (16:52):
So what you saw was a lot of companies realizing that their data was valuable, namely U-Haul, who said, oh, well, we can sell you the data of where a person started, where they picked up their van and where they dropped it off. And that’s cool. Maybe we’ve all moved with a U-Haul once, but it doesn’t really track everyone. And then you had united Van Lines doing the same thing. But again, that’s a different slice of the population maybe that rents an 18 wheeler hires a professional moving company. You’ve got Zillow and Redfin trying to do the same based on either A, where people are looking for houses, or B, where people transacted houses. But basically one of my favorite activities is just to go on Zillow and look for houses. So that’s not really indicative that I’m actually going to move. So, none of those usually reliable sources would do for this.
Matt Larriva (17:43):
So, I had previously worked with this company called Orbital Insight. And it occurred to me that they might have what we’re looking for. Now, if you’re unfamiliar, Orbital does low Earth orbit satellites, and they track visual data to track logistics. So basically, everything from what does the supply chain of palm oil look like in this area of the world to how is the crop yield looking to how much water is still in this reservoir. And then you can try and infer what the crop yield is going to be able to do because of the water that’s available. So, really fascinating applications.
Matt Larriva (18:26):
Now, why that’s interesting to us is because they also do cell phone data tracking. It’s all completely anonymized, so they don’t have any ties to an individual, they just have ties to a ping. So, a lot of companies use this to answer the question, what side of the building are people entering on? Or interestingly, if you can’t go to Chick-fil-a on a Sunday, where do you end up going instead? So, you can answer questions like that when you have cell phone pings, and you can look at them over a long enough time.
Matt Larriva (18:58):
What I wanted to do was look at it on a national scale. So I reached out to my contact there, and he had never had a request like that before. They weren’t sure they could do it on that broad of a scale because they’re used to doing it down to like the foot. I wanted to do it to obviously the 1000s of square miles. But they huddled up and then they realized they probably could do it. So, we worked on a pattern that we would want to see, and then they delivered the data to us, and we got back this really interesting subset that showed about three million cell phones. So, 1% of the population, and you’re thinking, wow, that’s not that much, but 1% of a random sample is actually really indicative, that’s really powerful in terms of statistical inference.
Matt Larriva (19:46):
So, we got that data from them, and we actually found that all the trends that existed pre-COVID were in fact accelerated during COVID. We had people moving even during lockdowns, even during the periods where you weren’t allowed to go outside, they were moving. We had people immediately after the lockdowns that were moving. And in almost all cases, they were repeating the patterns that had been established previously. And that kind of makes sense. You can think of people who were saying to themselves, when I get the chance, I really want to move to Texas. I can’t deal with California traffic anymore. When I get the chance, I really want to move out of New York down to Florida. My friend lives there, it’s got better weather, it’s cheaper cost of living. And then COVID hit and they all took that as an opportunity to do so.
Matt Larriva (20:36):
So, what we were able to see from that is that our investment thesis was really actually intact, and we were able to continue our investment process in a time when a lot of other people had that kind of pause and check the gauges and make sure everything was still heading the direction they thought.
Adam Hooper (20:54):
Yeah, which is fascinating, again, you talked about the persistence of these demographic trends. What have you seen in terms of the profile of people that are moving? Did you see any changes in demographics by age, by any other demographic markers that you guys look at in terms of who was moving or different demographic moved to different parts of the south or in the Sunbelt? How did that look?
Matt Larriva (21:21):
It’s such a good question. Because the data was anonymized, we didn’t have ability to say, okay, this is a young person. But what we do know, historically, is that young people tend to move much more than older people, and the less wealthy tend to move much more than the wealthy. And finally, that people tend to relocate to the suburbs. So, what we can kind of do is a qualitative extrapolation. If what the trends showed before COVID persisted after COVID, which is what we saw in our data, then it’s probably safe to say that also those demographic trends held up as well.
Matt Larriva (22:01):
So if you think of your average mover, you’re thinking of somebody from, say, Los Angeles, who is making a moderate income, who is younger. Probably somebody who’s going to be a renter. Now, we can also probably say that they’re going to move towards the middle of the country and probably a little bit further south, and they’re probably going to end up renting somewhere on the outskirts of a major city.
Adam Hooper (22:28):
And so, this is, again, what you guys dubbed in the paper this swarm to warm concept. So, tell us a little bit about what that means, and then also, after that, this kind of out migration from metros into suburban and south. You also mentioned there’s a kind of a whipsaw with some of these metros too. So I’m curious how that plays in, what we’re seeing now as kind of that reaction to the exodus, if you will, or the big migration out, what that looks like on the swing coming back into those metros.
Matt Larriva (22:57):
Yeah, yeah. So that’s a really interesting phenomenon. So, this swarm to warm is a term that we use because it rhymes, but basically just suggests that, that’s how you can track where people are going to move. And that’s not just based on our research, that’s based on a study that showed since 1950 through 2000, the primary determinant of population growth is the mean temperature of a city in January. And they are directly correlated, meaning the higher the temperature in January, the more the annual population growth.
Matt Larriva (23:33):
So, it’s a sort of a quick way to give yourself a demographic check. What do you think a city is going to do at least in terms of domestic net migration, it’s probably going to either gain if it’s warm or lose if it’s cold. Now, a lot of news outlets look at our research and said, oh, wow, then New York is dead. Or they took a look at our research on the second pass because we did it in two phases. One was March through June, then June through September. And in the June through September segment, Chicago and New York actually were among the cities that people were relocating to. And so, they thought, oh, well, the big city is in full swing again. It’s not really the case in either way.
Matt Larriva (24:25):
Now, what happened was that those cities both have a high population period, and they have a high population of students. So what you can imagine is that they sort of emptied out in the early phases, and then slowly people started trickling back in over time. But I do still think it’s the case that you’re going to see this warm to warm continue. I think a good way to think about the sort of winners and losers in this is that the colder cities or the denser cities, they’re not really going to lose population. People love to think in terms of what cities are going to just dominate, what cities are going to become the next Detroit. I only used Detroit as an example because they had decreasing population for quite a bit.
Matt Larriva (25:17):
I don’t think that’s a good way to think about it. In most cases, what we see is this sort of like Markov dynamic, where people flow through a place, and what we see is that international immigration tends to go to LA, New York, Chicago. That keeps those cities kind of stoked. But domestic migration tends to draw from the cities and fuel the Sunbelt region.
Adam Hooper (25:42):
And so, what are some of the cities that you’ve seen or geographic areas that have benefited the most from the, I guess, both this historical longer term look, and then more recently, since beginning of 2020?
Matt Larriva (25:54):
So historically, you’re thinking places like Houston, and basically the bigger cities in Florida. Those just did really well as a result of the swarm to warm. More recently, Phoenix joined that list, started doing really well. And then even more recently, Las Vegas is on that list and doing really well. Las Vegas is a really interesting area for a variety of reasons, demographic trends that we can get into if you want, but Vegas is an interesting spot to keep an eye on.
Adam Hooper (26:33):
How much of this population is coming from the metros? You mentioned the dynamic of you have international immigration that kind of keeps the equilibrium I guess of the domestic that’s moving out. What geographic locations are the losers of this? Where is the bulk of this pattern coming from, and are we seeing net population decline in specific areas, or is there enough international immigration to keep it kind of at base levels?
Matt Larriva (27:02):
There’s some international, rather, there’s some actually declining cities, but they’re very few and far between in terms of big cities in the US. Generally, what we’re seeing is this drive towards the suburbs. That’s been in effect for quite some time. I expect that to continue, for any number of reasons, you can kind of play out. But the suburbs just sort of give people more space to live a life that they wanted. So, that’s one trend that I’d keep an eye out on. I’m going to be curious to see what immigration does under a new policy and how that’s going to impact the sort of holding capacity of these cities, if they are actually going to really surge forward. New York City is kind of on a precipice right now, and really could go either way. So it’ll be interesting to watch as time carries on.
Adam Hooper (27:56):
I guess for the listeners of the show here, what are some accessible metrics or indicators that you guys are looking at that maybe they can start to pay attention to or keep an eye on where they don’t have to, they don’t have access to the three million cell phone user data from Orbital Insight? What are some more accessible metrics that maybe people can start to watch and keep an eye? What are you guys paying attention to that might be some indicators of some of these trends or changes that might be coming?
Matt Larriva (28:27):
Yeah. That’s a really good question. So, what always amazes me is how good a job the US government does at tracking data like this and releasing it. So, FRED has a wealth of data. And they have free APIs. So you can download almost anything off of FRED, specifically, population in these big MSAs. And that’s always a great place to start. Look at who’s gaining and losing in population. Another really good place to look is the employment figures by job type, which the BLS has, and they have on a monthly basis, it’s really good data. If you see jobs increasing in an MSA, especially if you see them increasing in industries that lead to so called agglomeration economies, then that’s a really good indication that that MSA is on the brink or at least has a really stable base to grow on.
Matt Larriva (29:23):
So, if I were starting from scratch and I had zero money for data, I would just tap the APIs of FRED and BLS and start there. Now, if you’re looking at a longer term play, I would say look at the county to county migration data, and that’s available from the Census Bureau, but also government dat, that’s released and the prelims just came out, so you can get some really good reads off that too. County to county is really powerful stuff. You can see some really cool patterns there.
Adam Hooper (29:57):
Perfect. Are there any things in your forecast or things that you’ve seen maybe starting afoot that will, if COVID wasn’t the catalyst to change some of these 70 year old patterns, what would be? Is there anything in your forecast that might be something to kind of start changing some of these trends or do you anticipate that these will continue on for decades to come?
Matt Larriva (30:27):
My general philosophy is don’t bet against the trend. So if I had to say, which way is it going to go, I’d say it’s probably going to continue. Now, a few things could make this abruptly stop. And that would be something like a really sweeping rent growth legislation. That could put some pressure in one direction or the other. So, if you see rent growth legislation, I’d say that would be a reason to maybe tread lightly.
Matt Larriva (31:00):
Another thing that is unknowable right now would be the climate change impacts. It’s not that climate change is unknowable, it’s that the impacts are intractable right now, at least with our current computing power. So, whether or not it’s going to make certain areas more prone to natural disasters, whether or not insurance companies are going to just say, I can’t write a policy for that building anymore, that’ll be something to keep an eye on as well. I would say those are probably the two biggest things. But on the other hand, you’ve got these agglomeration economies that once seated take a really, really long time to reverse. You can look at LA, look at San Francisco, these places that just grew way beyond their bounds way too quickly, but that arguably show no signs of stopping.
Adam Hooper (31:55):
Perfect. I think that’s a good overview on the migration research. Again, we’ll have a link to that report in the show notes. Absolutely, highly recommend reading it, fascinating stuff. But I also want to cover, which I’m super excited to get a little bit nerdy on, is the work that you did with Peter Lindemann, we talked about a little bit earlier, forecasting cap rates, which is huge. That hasn’t really been anything that has gone outside of a gut feel really historically in our industry. So, what did you do with Dr. Lindemann, and tell us a little bit about that research and then we can dig in a bit?
Matt Larriva (32:33):
Sure. So Dr. Lindemann was at a conference with FCP. He and I were talking, he was obviously a very reputable Wharton professor, he’s now really strong in the space of consulting, has the Lindemann Letter, which is really high quality research. And we were just sort of talking about cap rate directions and stuff like that. And I quoted the off to use phrase, well, interest rates are low, I’m sure cap rates will follow. I thought I was being very intelligent, very insightful. And he said, no, that’s actually not true. Not one to disagree with people who are far smarter than I, inquired as to why he thought so.
Matt Larriva (33:18):
And so, some of his research, he said earlier on, noted that interest rates actually weren’t a prime driver of cap rates. Actually, something else was. And it was a variable that he had named funds flows. So meaning how much money is going after real estate normalized? So, I was a little bit incredulous. I went back to my office a little bit later, tried to recreate the study. And sure enough, he was absolutely right. I looked at his work, I ran the numbers using a different statistical technique, and it came out the same way. So, we found based on his work, that you could actually do a much better job of forecasting cap rates if you used funds flows. And what we used was the mortgage debt outstanding, again, government data, divided by the GDP. You can do it in nominal, you can do it in real dollars. But that figure moves, interestingly. And it actually turns out that you can use that to chart where cap rate should go. Go ahead.
Adam Hooper (34:22):
I was just going to say, that sounds so simple.
Matt Larriva (34:25):
Yeah, it really does. But that’s sort of the crux of it. It’s this deceptively simple and complicated space. Previous to that, a lot of people were looking at interest rates as determinants of cap rates. And arguably, that’s simple too, that’s just playing on the discount rate. But it’s a really hard space, and incremental gains are really hard to come by. What we found in the early 90s, and I say we collectively as an industry, is that you could forecast cap rates pretty decently using those metrics that everybody cites. Interest rates, future growth expectations, return expectations, and then a risk premium measure. So, that was sort of where we were as an industry for a long time. And the research was pretty good, but it always failed you in these periods where things whipsawed really quickly.
Matt Larriva (35:29):
So what we did was we looked at this funds flow figure, and then we added in the unemployment rate and the cap rates themselves. So you get into an auto-regression space. And what we found is that we can actually do a, as good, if not better, in many cases job than all of these other variables that you have to sort of forecast themselves in forecasting cap rates. And so, why is that interesting? Well, because when you’re in a space like we are right now, where real rates are pushing their zero lower bound, rather nominal rates are pushing their zero lower bound, real rates are lower, in some cases. They cease to be a reliable metric, right? You can’t look to these figures that are just held low. So you need other figures. And the fact that you’re able to forecast cap rates so well using alternate data really has a couple interesting, overarching points to it.
Adam Hooper (36:29):
Going back there, so funds flow added in unemployment, added in actual cap rates using concrete numbers, whereas you said before, you were maybe forecasting off of forecasts. So you had to forecast some of that underlying data and then you had to use that in your analysis. So are you able to use actual real nominal data to come up with these cap rate forecasts rather than forecasting off of the forecast now?
Matt Larriva (36:53):
Exactly. So, we’re using, for example, what was the cap rate last quarter, the quarter before, and the quarter for that? What is unemployment right now? And then what is funds flow as a percent of GDP? Now in reality, and I think it’s important to note, we’re using these and we’re using their difference and the values themselves. Now, what a lot of finance gets wrong or what they just sort of hand wave is the fact that time series is really gnarly data. Once you get past one variable in time series, you get into this esoteric space of needing really advanced models with a bunch of correction terms to make sure the residuals don’t auto-correlate themselves.
Matt Larriva (37:36):
So all of that is sort of a fancy way of saying, the best way to predict a stock price is to predict what it was yesterday. So, we have to use changes of all these variables to get reliable statistical models. So, back to your original question, yeah, we just used the three actual numbers and we were able to forecast cap rates, which is in juxtaposition to most models which use return expectations to forecast cap rates. So then you’re basically forecasting on forecasts.
Adam Hooper (38:04):
And so, how far are you able to look in the future with this, and how long have you been working on this methodology to be able to test this?
Matt Larriva (38:14):
So, we back-tested it all really thoroughly, it back test really nicely. We did it in sample and out of sample, and it works really nicely there too. But yeah, disappointingly, the most you can do is about four quarters into the future. And if you’re thinking to yourself, wow, that’s really not that great, you’re absolutely right. But that’s kind of where we are in the industry right now. That is, surprisingly, state of the art. You get four quarters ahead with any sort of dependable r squared with any sort of statistically robust model. Why is that? Because cap rates are a very, very impregnated variable. And what I mean by that is there is so much held in a cap rate. It’s past pricing, expectation pricing, it’s rates, it’s your fundamentals, all wrapped up in this one little figure. So it’s not surprising< and you shouldn’t be able to forecast it much further.
Adam Hooper (39:08):
And how sensitive is that to events like what we saw here in the last 12 months?
Matt Larriva (39:14):
Yeah, I mean, that’s a great question. So, what we saw was that the GDP basically collapsed. And what that meant was that there was actually proportionally more money chasing real estate as a percent of GDP than there was prior, which ultimately leads us to the conclusion that cap rates should be stable, if not go down. And that was helpful as we were investors in this crisis looking for indications of what we wanted to do with our dry powder. Did we need to pause or could we proceed? Could we take advantage of some of these deals or did we miss it by the sidelines. So it’s good to have models like this in those cases. And that’s why when you try and evaluate what use is a model that only forecasts one to four quarters in the future, well, it’s really useful if you’re going through a financial crisis, or if you are a fund manager, trying to think about, do I liquidate now or should I wait a year?
Adam Hooper (40:14):
Yeah, the hold versus sell analysis. Something that can forecast four quarters out is better than something that can’t.
Matt Larriva (40:20):
Sure. Exactly. Yeah.
Adam Hooper (40:22):
That’s a huge step. A lot of listeners of the show are investors, passive investors in real estate deals with other third party managers, we have a number of managers themselves that listen to the show. How is this helpful for them? Are there any practical day to day applications of this work or this methodology? Are you guys going to be looking to commercialize this going forward? What is the plan now that you’ve come across this methodology? What now?
Matt Larriva (40:56):
I’m a big supporter of transparency, especially in the space of academic research. Some of the code that I’ve written, I’ve done talks on helping other people to use it. Some of the research that I’ve done, I’ve obviously published broadly and tried to get peer reviewed. The full excerpt or the full study here is in peer review right now. So, assuming full passage there, that’ll be online if anybody wants to look at the full methodology and the math behind it, and check me and tell me why I’m wrong. But otherwise, if you want to keep it simple, you can, again, download all this data publicly. It’s all FRED data. So you just look at the, something that’s effective like the statement of accounts, but you can find it on FRED. You just look at mortgage debt outstanding and divided by GDP. See how that term is changing and how that term is held up and try and look at that against prior periods. That’ll give you a good indication of what cap rates are likely to do in the near term.
Adam Hooper (41:56):
And what are they likely to do in the near term?
Matt Larriva (41:59):
By our research, they should stay stable, if not continue to decrease.
Adam Hooper (42:04):
Yeah. And so, it’s again, I’m agreeing with you, but in the back of my mind, I’m going, it’s an interest rate thing. That’s how the industry has just, that is the baseline, cap rates move in relation to interest rates period. So this is just a very, very different way, again, not arguably that complex of a way of looking at it that, again, sounds like the research I’m very much looking forward to reading that when you guys do publish that. Sounds like this is just a much simpler way to look at how to forecast those.
Matt Larriva (42:43):
Yeah, it’s true. You’re totally right in your intuition. You’re right to say, no, this can’t be no, it’s got to be an interest rate thing, because if you just graph 10 year yields versus cap rates over the last 20 years, you get this really nice graph that graphs up into the right. And it’s this really beautiful scatterplot with a .5 r square.
Matt Larriva (43:04):
The problem with that is that you can’t look at time series data that way for a few reasons. But moreover, you quickly run into this thing called Simpsons paradox, meaning, if you group a bunch of different things together, you can get the wrong impression. So if you look at these cap rate and yield series separately, look at them from 2000 to 2002, look at 2003 to 2011, look at 2012 to present, all of the regressions point negative. And so, you’re saying, well, sure, you can massage any data, but it’s not true. We didn’t just arbitrarily slice it. There’s these big gaps in the data that are showing you there’s a lurking variable there. So, take a look at the graph that we have on the paper, and I think it’s pretty illuminating.
Adam Hooper (43:54):
Yeah, we’ll definitely have links in the show notes for all that for everybody that wants to dig in a little bit more. We talked before the show not getting too crazily academic, but being able to dig in a little bit more. So I think we’re at a good spot there.
Matt Larriva (44:07):
Adam Hooper (44:08):
Is there anything else that you guys are looking at right now in terms of research or areas that you’re excited about that you can kind of give us a little preview or teaser on?
Matt Larriva (44:17):
Yeah. One of the key ones is portfolio optimization. So, there’s a really cool phenomenon going on in portfolio optimization as it relates to geographic diversification. Basic theory is that real estate doesn’t behave in the traditional modern portfolio theory framework. So, it’s not just about maximizing [inaudible 00:44:43] ratios and putting those together and trying to get on the efficient frontier. There’s actually something else going on there that I’m trying to dig into. I’ve synthesized a couple other holding mechanisms that grossly outperform. So, looking to publish on that and hoping for some interesting reviews there.
Matt Larriva (45:00):
And then, I’m fascinated by demographic shifts. I’m really interested to see how holding capacities of cities in the US manifests itself. Why we can’t figure out if cities seem to be growing or dying, what seems to be the driving force behind that? I think there’s something interesting going on there. I think it might be a Markov process but I’m gonna dig into that. And hopefully, if any of that’s fruitful, we can talk about it.
Adam Hooper (45:25):
Absolutely fascinating. And yes, there’s a lot more to talk about. Before we let you go, though, some kind of standard questions we’ve been asking. What’s keeping you up at night, and then, what are you optimistic about here in 2021 as that continues to roll out?
Matt Larriva (45:41):
Those are great questions. What keeps me up at night is wonky economics. So, I’m worried about how inflation manifests, how governments slide into modern monetary theory, and what the ramifications and consequences are of just having massive debt loads. I think Dr. Lacy Hunt is a genius and he publishes all his stuff on Hoisington Investment Management. He’s got a lot of data and a lot of historical evidence to show that when you take on this much debt, you kind of wedge yourself into a disinflationary stance that kind of becomes really hard to get out of because of the marginal productivity of debt. So, that’s what keeps me up at night. He’s a former economist for the Dallas Fed. That’s what keeps you up at night.
Matt Larriva (46:38):
What I’m optimistic about is, briefly, the night is always darkest before the dawn, right? If you look at when really positive paradigm shifts have occurred, in many cases, it’s right on the heels of some pretty intense turmoil. For example, you’re going to get a lot of students who have sort of divorced themselves from the education process because of this gap that they’re forced to take. Well, that’s terrible that we have students that are losing their interest in education. But maybe this forces us as a society to open up our views on alternate paths to success that don’t include college. Maybe that becomes a way that we as a country start to legitimize and bolster the role of trade schools, which have such a prominent position in other developed nations but seem to have sort of fallen by the wayside for us.
Adam Hooper (47:33):
Yeah. What does, and that’s something, again, we’ve tried to figure out, what are the long term changes that persist through this experiment that we’ve been running the last 12 months? How will this new way, again, whether it is remote work, whether it is, again, completely rethinking the education system, some of these norms that we had that we stuck to that have been completely shaken up, how do we rebuild from there? How do those look when they come out better and stronger? It’s going to be a very interesting time here as we head into 2021.
Matt Larriva (48:07):
Adam Hooper (48:09):
Very, very much appreciate you coming on the show today, Matt. Fascinating conversation. We’ll have lots of links in the show notes for everybody. So thank you for spending your time with us today.
Matt Larriva (48:18):
Thank you very much.
Adam Hooper (48:19):
All right, listeners, that’s all we’ve got for today. First episode of 2021 in the books. As always, if you have any comments or feedback, please send us an email to email@example.com. And with that, we’ll catch you on the next one.