The Joe Walker Podcast - Housing Bubble Week: Lessons From Amsterdam's 'Gentlemen's Canal' - Piet Eichholtz

Episode Date: May 14, 2019

We all know that real estate is the best long-term investment for capital gains...See omnystudio.com/listener for privacy information.

Hello there, ladies and gentlemen, boys and girls, swagmen and swaggads. Welcome to episode two of Housing Bubble Week and welcome back to the Jolly Swagman podcast. I'm Joe Walker, your host. Now, we all know that real estate is the best long-term investment for capital gains, and property investors should just hold through the ups and downs of the market, right? But is that really true? There are two classic mistakes people make when using past prices to predict future prices. For one, because real estate transactions are few and far between, we tend to use prices from decades ago as our

reference point. We hear stories about how our parents or grandparents bought their homes for next to nothing, and we compare the original purchase price with current prices. It can impart an impressive impression of real estate as an investment. But when we anchor to these halcyon days, we anchor to nominal prices. That is, we fail to adjust them for inflation. This is a well-documented error, and economists even have a name for it, the money illusion. The other error that plagues our thinking about where prices might go in the future is that even when we do consult real prices, we have a habit of extrapolating the future from a very brief time period. One Friday night in April 2019, I was leaving Star City Casino in Sydney with some friends

and on the escalator down, I got into a conversation with two real estate agents. Naturally, the Australian property market was discussed, but I promise this is not the only topic I talk about in social situations. They were earnest and sincere, and I asked them why, as they continued to insist, they thought the worst was over when it came to house prices falls. The one who was doing most of the talking gave me an answer that shocked me. He said, just look at what prices have been doing in the last seven years. But it's not just the literal man on the street who peddles this sort of analysis. I've been to numerous property seminars which frame the potential for future capital gains in the context of the last 30 years of price data. And I'm aware of a certain large ratings agency

that takes the same approach. Now, in fairness, the ABS only began publishing a national price index in 1986. But since then, thanks to UNSW professor Nigel Stapleton, Australia does have a long-run price index going back to 1890. So there's no genuine excuse for limiting the timeframe. To paraphrase Kenneth Rogoff and Carmen Reinhardt's book, This Time is Different, if you're looking for a 100-year flood in 33 years of data, you've a two in three chance of missing it. So to come back to the question, is real estate the best long-term investment for capital gains? To answer that, you need to avoid both of those pitfalls. You need to take a bustling, cosmopolitan entrepot.

Looking out on its port, you could have mistaken it for a forest. Such was it bristling with the masts of flutes, yachts, galleons, boats, buses, boas, barges, brigs and balanders. The 17th century was the Dutch Golden Age, as Simon Schama so beautifully described it in his book, The Embarrassment of Riches. The Dutch economy was turning like a windmill. Jan de Vries called it the first modern economy and indeed the world's first stock exchange, the Bourse, was founded in the city of Amsterdam

itself. Amsterdam was truly the centre of the action and it was not just economically but also culturally ascendant. At various points the city was home to Rembrandt and Spinoza. It was a refuge for migrants and merchants of all types. Women had greater rights than most of their European counterparts. And the threads of liberalism itself began to coalesce in the city. Amsterdam went through a period of expansion during the Dutch Golden Age. But for our purposes, we're interested in only one piece of that story, the Herengracht Canal. It was one of three canals laid out in a half circle

around the medieval city centre. The first part of it was dug in 1585, the second in 1612, and the third in 1660. As it grew, merchants flocked to establish their residences along it. And to this day, the Herengrok Canal is the ritziest canal in Amsterdam. Locals refer to it as the Gentleman's Canal. Many of the houses strung along it today are identical to what you would have seen if you had taken a stroll in 1628. But what has this got to do with housing bubbles? Fast forward 347 years. A book is published celebrating Amsterdam's 750th anniversary. Its title, translated into English, is Four Centuries Herengracht.

It contains a general history of the Herengracht, but something else, something of great value to economic historians. A complete history of all the transactions and transaction prices of all the buildings on the canal stretching back to 1628. Pete Eichholz, my guest, is the Fortis Professor of Finance and Real Estate and the Finance Department Chair at Maastricht University. More than 20 years ago, Pete had heard of this book and he'd been searching for it, but it had eluded him like a holy grail of real estate economics. Having all but given up, one day in the mid-1990s, as he was browsing a secondhand bookstore in Amsterdam, he stumbled on a dusty copy of Four Centuries Herengracht, complete with its prices data. In 1997, he published an academic article using the information in the book, and the Herengracht Index was born, the

world's longest running real house prices index. For almost a decade after that, the index remained only relevant in esoteric academic circles, until Robert Schiller praised it in the second edition of his bestseller Irrational Exuberance, published in 2005. That's how I heard about the index and I decided I had to talk to Pete. In this episode, Pete explains the index, what makes it special, how he compiled it, why we need to take a long-run view of house house prices and how he thinks about housing bubbles. I also challenge him on the external validity of the index and ask him, can we generalize its lessons to other housing markets like Australia's? So without much further ado, please enjoy my conversation with Pete Eichholz. Pete Eichholz, welcome to the podcast.

Thank you. You're the man behind one of my favorite pieces of academic research, the Heringracht Index, which we'll discuss... It's very nice to hear. ...in a little while, yes. And as I was telling you, I first encountered your name last year when I was reading the third edition of Robert Shiller's bestseller, Irrational Exuberance, and I became fascinated by the index and your work. But before we begin,

I'm fairly sure most of our audience won't have heard of you before. So can you please introduce yourself to us? If you're at a cocktail party and someone asks, what do you do? How do you answer that question? Well, I'm a professor of real estate and finance. And that's also what I do. I teach real estate and finance, real estate courses, but also courses in emerging market finance. And I've taught courses in fixed income and things like that. But basically the whole range of finance. But my research is all about real estate. Fantastic. And are you from Amsterdam originally? Yeah, I'm from, well, not Amsterdam itself, but sort of an outskirt of Amsterdam. And then I moved to Maastricht about 30 years ago to do my PhD here,

and then I just stayed on because I like this region a lot. But, of course, Amsterdam is sort of the center of the universe for Dutch people. So a lot of what I do, also research-wise, pertains more to Amsterdam than to Montserrat. Yeah. I heard in a YouTube video you described Amsterdam as my beloved city. Yeah, I love it. It's a great city.

What is it about it that you love so much? Well, what I really like is that it has this very international cosmopolitan atmosphere, but it's still very small. So everything is walkable. There are, of course, high-rises and stuff, but it still has this very nice old center and these neighborhoods that give you a very sort of small town feeling but it is really big in terms of how international it is and how cosmopolitan and how connected yeah and uh and it's all great although currently there is a risk of becoming a bit more like venice where uh especially in the center of the city tourism is is is so crazy that, yeah, it's not so much fun to live there anymore. But I don't know it myself because I don't live there, but friends who live in the center

of Amsterdam say, well, it's not as much fun as it used to be. But, you know, still great. I like it. Now, the Herengraakt Index, what is it? Can you give us a brief outline of what makes it so special? And maybe beginning with the story of how you got there, how did you come to put the index together? Yeah, well, it started with a big book.

And so Amsterdam celebrated its 750th anniversary in 1975 and for that occasion the the city commissioned some historians to do a book on on the here God because that was the main canal of the of the city and And in that book is a history of every house on the Herengracht, including the people who lived in it, the people who owned buildings or rented buildings, but also the prices for which these buildings were sold and bought. And sometimes also, of course, they were inherited, so they weren't bought and sold.

So you can track the whole history of every house and that book was just standing there and then I was doing some index work at the university I gave a presentation for a group of industry folks and somebody said

hey I've heard about this book on the Herengracht index there was no index yet but on the Herengracht this book on the Herengracht index, there was no index yet, but on the Herengracht, the history of the Herengracht, why don't you take a look, maybe you can do something with it. So I took a look and then I saw indeed that there was enough price material because we don't have prices for every individual transaction, but enough to estimate this index. So then I bought the book and I had a couple of student assistants to type over all this stuff.

And then I estimated this index. And at first I did it with a sort of faulty technology, but then I used repeat sales and then the index came out as you now know it. And then I've been updating it and stuff like that. But it's basically because of these historians who prepared this book that I could do it so easily. Great. So before we discuss the index in a little more depth,

can you just describe the Herengraacht? I won't try to pronounce it uh like you can but the the herring canal for us uh yeah it's regarded as one of the the most exclusive and desirable locations in amsterdam right that's right yeah still it was it was meant to be uh that and it and it still is so amsterdam has this uh half moon shaped uh center and uh the uh so it's you Amsterdam heeft een halve maan geformeerde centrum. Je kan zeggen dat het een serie van kanalen is die zijn geplaatst zijn rond de medievale stad van Amsterdam. In een halve maan vorm. Er zijn drie hoofdkanaal.

De Herengracht, de Keizersgracht en de Prinsengracht. Herengracht betekent de herenkennel, de keizerskennel en de prinsenkennel. En het interessante is dat de Herengracht van deze drie de poshste was. Wat goed is, want Emsen was al een deel van de Republiek. Dus de herenkennel en de prinsen Prince's Canal were less posh than the Gentlemen's Canal. So that's interesting. That tells you something about the importance of civil society at that point. And it was laid out in three phases.

So the beginning was in 1585, and then it was not such a beautiful canal yet and many of the buildings were built small buildings made of wood but then in two later phases the whole canal was laid out in its full half moon shape half circle shape and then the plan was really to to make big lots big houses it was the canal where the let's say, the top of the Amsterdam trading community, merchant community would live. And they did. So they built these fantastic mansions with really beautiful gardens. And most of them still exist. So if you walk the Herengracht, you don't walk a straight street, but you start in the, let's say, the northwest of the city, and then you walk the whole canal, and then you're

in the northeast of the center of the city. And it's just a really fun walk. So for everybody visiting Amsterdam, do it. It's really beautiful. And there are a number of houses that are just iconic. I have a few favorites in particular. There's number 81, which is the oldest residential house on the canal. It's step gable, so it's got the steps on top of the roof.

First built in 1590, which was just after the first part of the canal trench was dug. Yeah. And that house was bought in 1625 by a guy called Peter Fransch. And I always found that interesting from a historical perspective because he was a carpenter. I just found that curious because a master carpenter would only have earned about 500 guilders a year in the 17th century.

And I always found it curious that he somehow bought a house on the Herengraat. Well, but that's also because, like I said, that first part of the Herengraat, so the low house numbers below 100 or something, they're in the westernmost part of the canal. That part of the canal was dug in 1585 and then developed in the years, let's say the decade after that. And at that time, the Herengogracht was not supposed to be

the posh canal that it turned out to be. These were still smallish houses, like I said, mostly built from wood. And so the fact that this one house was not built from wood was quite unique, and that's why it survived, because all the other houses that were built from wood did not survive, because the municipality at some, outlawed wood buildings because of fire risk. So all these buildings had to be replaced by stone buildings. So in the beginning, the Herregas was not so posh.

And the buildings were small, so carpenters could afford it. But the other parts of the canal, the parts that are more in the south and to the east, they could never be afforded by a carpenter. It's really where the bankers and the traders and all these folks lived. And there's a part known as the Golden Bend, which today is occupied, I understand, a lot by banks, even a cryptocurrency company. Yeah, yeah, that's right yeah that's uh and these used to be these mentions but then i think uh until 1900 uh they were almost exclusively residential but never fully residential what we usually what you usually had in mention like that is that the merchant lived there with his family, but then they also had their main office in that building. So

one or two of the rooms in the building would be used as an office or to entertain customers or things like that. So there was also, let's say, a bit of office use within the main residential use. But then from the late 19th century, early 20th century on, there was really a lot of conversion, full conversion of houses into offices. Because that was then the highest and best used for the space. And especially a lot of banks. But also your law firms and things like that so let's say high high value business to business services you could say but interestingly that house prices have been so high as of late that there's now sort of a reconversion into the residential so the office market in

Netherlands is not doing so great and has not been doing so great for the last 10, 20 years. And so actually offices are now being converted back into residential, also in the center of Amsterdam and also on the canals. Wow. So if you strolled down the Herengroch today, how different would the houses you see be from when they were originally built? Yeah, some, let's say if you talk about the houses, some are the same or almost the same.

And interestingly, the book that was the source of the information for the index also contains drawings of the frontage of the houses, I think for the 17th century and for the 20th century, so you can look at the differences. And some are the same, but there's also been a lot of reconstruction in the 19th century and even in the 20th century. So it's in terms of style, it's rather eclectic. You could say that time has been also the architect of this canal, but it's still kind of homogenous because in terms of scale,

even if the buildings are new, they have the same height and the same dimension, so it doesn't really strike out as oh this is a very modern building in between these antique buildings but what is really different than uh than before is that when these houses were built there were no trees uh on the canals there was just it was just houses and the street and then the water and now there's trees on both sides of the water and so it looks far more green than it used to be it used to be a bit um well a bit stony you could say and just just the water and these houses now it's more interesting and of course the big difference is that cars and so there's cars parked below these trees and so if you look at

the canals beautiful but you have these cars all over the place and of course if you look at old buildings old paintings then it's just a couple of people strolling along the canal and that's that's more beautiful so um yeah now when a lot of people talk about long-run trends in house prices normally they're referring to a few decades which can be very dangerous the herring graft index was so special because the index starts in 1628 to 29 all the way to 1975 that was the original index yeah and then you are you updated it to 2008 yeah and we updated it again yeah updated to every 17 yeah is that the longest running house prices index in the world to your knowledge yes uh it is. Yeah, yeah. Although, that's for prices, but for rents, we've recently broken the records. So now for rents, we go back to 1500. So we've now created rent indices

for seven European cities from 1500 to the present. So we have market rents annually until 1500. So that's longer. But for prices, this is the longest, I think. That's amazing. Still. Wow. And at some point, somebody will beat it.

But yeah, it just depends on finding the records. And sometimes the records survive, but sometimes they have been burned or destroyed. And then there's just not the information that you need to create the index. It's a testament to Dutch record keeping, I guess. That's right. Yeah, Amsterdam record keeping. This is from the Amsterdam City Archive, not the National Archive. Oh, okay.

Sorry, I shouldn't give too much credit to the national government. So give us a sense of the difficulties in compiling a long-running house prices index because houses can change in quality over time. And that's one of the major problems. Can you just start by explaining the nature of that problem? If you see just one transaction or a couple of transactions in a year for houses, then you see maybe a couple of really beautiful houses being sold in one year and worse houses sold the next year.

And then if you compare prices, then you would think that prices have gone down, but it's just because you have a different quality set. So you should, from year to year, compare houses of the same quality. Now, you can do it by doing some statistical analysis where you put in a lot of quality variables and then say, well, I filter out all the quality variables. And then what I've left is the pure market market index movement. But if you want to go back in time like I did, then you don't have these quality variables. You don't know how many rooms were in houses or, you know, how many fireplaces or how many windows or how big the garden was or things like that.

So so you can use that trick. And the other trick you can use is what's called a repeat sales index and that's what we use there what i use here is where you say well we compare transactions of the same house so there's a house that was bought for example in 1650 and then it was sold in 1675 and we compare these two transactions with each other. Now, of course, the house can have been renovated between these two periods or have deteriorated a lot. That's why it's so important to know something about the history of these homes

because very often we also know that there was a big renovation or the house was in ruins or something like that, and then we forget that pair of transactions because then obviously we're not comparing like for like in terms of quality. But if we don't have information about renovation then we assume there was no big renovation but just the normal upkeep and then the house stayed the same in quality and then we compare you know for one house uh we compare one house to itself that's what you do with repeat sales index and then you do

for all the houses and based on that you can then get a good let's say quality adjusted uh idea of what the what the what the what the market has done. And that's how we've done it. And that's how it's commonly done. That method was pioneered by Bob Schiller and we read with Chip Case together. They wrote a couple of classic papers on it. And actually before them, it was pioneered by bailey

muth and norse in already already way earlier but they took that up and then popularized it and it's basically the same method that i use so the repeat sales index it's very important because you're comparing apples with apples you're controlling our confounding factors you're just looking at the same houses being bought and sold over time and that's right that's why the herring gracked index is so special roughly how many how many houses and how many transactions were you dealing with between 16 28 29 29 and modern times? Yeah, so in total, the number of houses on the Herengracht is 628

or something. I don't know exactly, but 630, but something like that. Now, 614 houses on the Herengracht. So that's the total sample in terms of houses. And we observe about 5,500 transactions. So that's the total sample in terms of houses and we observe about five and a half thousand

transactions. And that's enough to create an index. So we just have two years for which no transaction data is available so there we interpolate but for the other years we have transactions so we have, yeah, we have we have 4200 pairs so that this it's enough it's not a lot but it's enough to create a good index without too much estimation error yeah yeah now when you talk about a quality adjusted approach are you talking about the hedonic approach there yeah you can we could have used the hedonic approach. And in fact, the first time I estimated the index, I used the hedonic approach because I had some quality information, but obviously not enough

to make a really good index because it was very volatile. Not so much because there was real underlying volatility, but because it was mostly estimation error. And then I switched over to repeat sales, which was clearly the better index method here. And I think for all long historic studies of house prices, repeat sales are the best. Now, there are many studies of long-term house prices that just use an average index,

that just say I use average sales prices that I observe. But then you don't have quality adjustment and you really do need quality adjustment. Otherwise you're telling nonsense really. And so it's very crucial to do it. Now you can use more sophisticated techniques, but my experience with this data is that the correlation between the very sophisticated repeat sales models and the simple repeat sales models is 0.99999. So, you know, I use Occam's razor and just keep it simple. Yeah. Now, I'm going to ask you to describe what the chart would look like if we plotted real prices over time.

But first, I think just for the people who aren't necessarily financially literate in the audience, it might be beneficial to explain the difference between nominal prices and real prices. Could you just give us a definition there? Yeah. So normal prices are the prices as you see them in the newspaper. So this is the price that does not adjust for inflation. But if you want to really make a good long-term comparison of these prices, then it's better to take inflation out because otherwise you would see that these normal prices go up, but it's not so much the housing market that has gone up, but just the general price level and that gets reflected into house prices.

So what you do to get from a normal price index to a real price index is to divide that index by the index for the general consumer by the consumer price index and then you get the real index and that gives you more meaningful information about the housing market rather than the housing market and the general pricing and that's of course the way you should make these intertemporal comparisons. Yeah. So that distinction should be, I guess, born in mind when we talk about the index, the difference between nominal and real prices. Now, what does the chart of real prices over time actually look like for the Herengraacht from 1628-29 to today? Yeah, well, it looks surprisingly stationary.

I wouldn't say it looks surprisingly flat because it's not flat. It's really moving around a lot. But it's stationary in the sense that it doesn't really go up. And so if you talk to people currently, people think that prices in tend to go up. Sometimes there's a bit of a crisis like what you have in Australia. Now, we had one in the Netherlands a couple of years ago, but on the whole house prices go up. But if you take out inflation of of that curve, then you see that they don't at least you don't you see it. The house prices stay more or less flat. So over the whole period that we looked at,

the house price increase in real terms was just 0.2%. And house prices just doubled over these almost four centuries. And that is something completely different than what people experience. People think that if you want to get rich, buy a house. And there is something to say about that because the mortgage that you take out stays the same and your house price goes up. But the price increase is mostly inflation rather than the house price per se. That's just crazy to me. And I remember reading in your original paper, I think if you'd chosen 16, 32 and 33 as the base period, there would have been virtually no

increase in real house prices. Right. Yeah. Yeah. Just crazy. Because a lot of people buy into the idea that housing is the best long-term investment you can possibly find. Yeah. But the Herengraacht Index shows something very, very different, which is that the values of the houses are preserved, but they took 350 years and they barely doubled. That's right. That is prices. But, of course, what the Herenggracht Index doesn't show is rents. So if you would look at this from an investment point of view, you would get a price that is, you could say, inflation stationary,

so you don't lose money over the long term because of inflation. And then you get the rent as your return. So there is a positive return to it. And also very importantly, if you think about the way you finance your house and you finance it with a mortgage and the mortgage is usually a fixed rate mortgage and that just stays the same in nominal value but in real value the value of the mortgage goes down because of inflation if you if you look at the whole package you have a house that stays the same in real value you have a mortgage that goes down in real value

and you have an interest rate income so that the whole package can still generate a very attractive investment picture, but you shouldn't do it because house prices always go up. Because in real terms, they don't. You should look at that whole package. Yeah, that's exactly what I mean. So, I mean, property, real estate is an attractive investment if you're generating revenue or dividends through the rents. But people who think that in terms of the capital appreciation, that it's the most fail-safe long-term bet, history seems to say otherwise. Yeah, it's true.

So you described the graph and it sort of undulates over the centuries. Now, I want to ask you how much of that can be explained by the vicissitudes of history and i thought i mean maybe you could do it i could do it but to give a quick abridged history of amsterdam from when the index first begins so interestingly there were uh still in the 17th century some pest epidemics in Amsterdam. And you really see that reflected in the index. So these were like plagues.

They were plagues, yeah. And in these plagues, people, you know, Amsterdam would lose 10% to 20% of the population in a year. So it was really, really huge. Wow. But interestingly, cities before 1900 always had a higher death rate than birth rate. So cities always needed more people, fresh blood, you could say, for the jobs. So cities were very open open and what you see is that the house prices go down quickly in a play but then they revert really really quickly back up because city is open and people come back in and demand for real estate

stays the same well then there was a lot of volatility in 17th century there were a couple of wars with the bridge about who would be the dominant partner and it's in trading in the seas and there were the Nordic Wars the first English Dutch were the second English were always at sea but it did disrupt trading. So it did affect Amsterdam's income and therefore also house prices. So you see that. Then in 1683, there was an invasion of the Netherlands. That's called in the Netherlands the disaster year where the French army invaded the Netherlands and almost took Amsterdam, just not.

And then you also see a huge drop in real estate prices. You see some, let's say, financial crisis in the 18th century reflected. You see a lot. So throughout all this, mostly the Amsterdam population grew or stayed stable, but then there was really one period in which the population really trended down for a longer period. Of course, during these pests, population went down, but then it reverted. But during just that period before Napoleon and after, so let's say 1795 until 1814, population really declined by 1% per year.

And there you really see prices go down a lot. Prices go down, I think, 50% or something. Well, there was the invasion of Napoleon, and there was also what's called the continental system of Napoleon meant that it really was bad for Amsterdam trading. So and then the population of Amsterdam went down and it really declined with in total about 20 percent or something. And that really you saw that really reflected in house prices. That was the worst long-term crisis in the city of Amsterdam in terms of prices but also rents and we also have a rent index you see the same thing.

That was the worst period in the Amsterdam housing market. Then in the 19th century when Amsterdam started gradually to industrialize, population grew and then house prices grew. And then, yeah, so, yeah. really take general lessons from the Herengracht Index because it's only valid for this particular canal in this particular city, which happened to be in a very turbulent historical context? Well, it's always, of course, dangerous to just generalize. And I think it's important to also create indices for other cities to see whether this is generalizable. But I think the important thing is that Amsterdam has been sort of, you could say, a constant quality city.

Amsterdam was a quite rich city in the 17th century, in the 18th century, the 19th century, the 20th century. It hasn't been like a boomtown that boomed and then faded away or it was always the richest city in the world or something. So it has been kind of in terms of its economic position, you could say kind of stationary. And within Amsterdam, this canal has been kind of stationary too in terms of quality. This is not a neighborhood that was good and then bad and then gentrified, sort of constant quality in terms of neighborhood and the houses were also constant quality and therefore I think

you could generalize some of this to do and tell a broader picture Also because like I just said many of the fluctuations that that I observe Can be explained by underlying fundamentals and are just not some flukes of And of course the underlying fundamentals can be unique I don't think that Sydney is going to have a pest epidemic real soon or that the French army is going to invade Sydney real soon. So, of course, these were historical

well, unique things. But let's say Napoleon invading and then the continental system that did affect the Amsterdam economy and the Amsterdam population numbers. And Sydney will also have will be affected by events in terms of trading and population numbers. And there you can draw real lessons. So that's one thing. And the second thing is that the fact that the prices in real terms don't go up to me is evidence that this market works.

And because a market where prices go up like crazy is a market that doesn't work well. That's not in equilibrium. But if prices adjust and stay sort of the same in real terms then this is a market that works well and so if and this was a market that was for a long time in its history really free market without a lot of government interference and it seems to have worked so yeah so you said you update the index every year. Where does it currently stand from when you first measured how much your prices increased in real terms? I don't know. I should have, of course, prepared that for this talk, but I don't know.

But it's actually a lot higher now. Right. So there's a lot of price increase over the last, let's say, until 2007. Then prices dropped about 20% in the two, three years after that. But since then, they've regained about 50. So prices are at record highs right now. And so MSTM is more expensive currently than it has ever been. So a general lesson we can take from the index is that prices overshoot and undershoot over

time but ultimately they remain somewhat flat. Well overshoot and undershoot, so the second part of that sentence I totally agree with. Over time they remain somewhat flat. But overshoot and undershoot really sort of suggest that they're kind of bubbly and then they bear out. But that's not what we see. We really see that fundamentals play a role. So, for example, right now, Amsterdam house prices are at record high. But if you would ask me, is this a bubble?

I would say no. It's just because we stopped building, not fully stopped, but we built far too little. So it's more a supply demand kind of situation where supply hasn't been able to adjust for demand properly and therefore prices go up. So that's not really a bubble. It's really a fundamental. So I would say prices indeed fluctuate, but they don't go up. They don't trend up over the very long term. Okay. Or down.

So why, as a matter of economic theory, should we expect real house price increases to be incredibly modest over time like they are for the Herringgracht. Maybe a good way to think about that question is to compare real estate to stocks. Like, does the S&P 500 go up in real terms much more attractively compared to housing? And what's different about housing? Well, I don't know if stocks are the right thing to compare it with. Maybe we should compare it with peanut butter.

Okay. And because with the stock market, supply and demand is kind of a difficult concept. But with housing and peanut butter, it's simpler. And with peanut butter, the supply of peanut butter can react almost instantly to increasing demand. If Australians start to like peanut butter far more than they now do, then, you know, some traders will say, hey, there's more demand for peanut butter, and we're going to supply the supermarkets with more peanut butter, and then there's more peanut butter, and it means that the price of peanut butter won't go up.

Maybe in the first couple of weeks, when there's all of a sudden a run for peanut butter in the Sydney supermarkets, then maybe the price goes up a bit, but before you know it, the peanut butter suppliers will be on it and prices will stay as they were. So if supply can adjust to demand in a very elastic way, then prices don't have to move. And in that sense, a flat price is a sign of a well-functioning market, if you look at the peanut butter example. And for houses, that's, of course, more difficult

because to increase the supply of houses, that can be a very long-term process. Now, if you go to the outback in Australia, then you can build whatever you want. So if there's more demand for houses in the outback, you build a couple of houses. So no need for price to fluctuate there. But in Sydney, I guess, or in Amsterdam, there's a lot of regulation about zoning and where you can build and where you cannot build and when you can build. built so before you have let's say the time it takes from a first plan of a housing development to actually the realization of the housing development

can take 10 15 years and so it's very difficult in the housing market for supply to react instantly to demand and therefore you have these fluctuations. But in the long run, of course, supply will react to demand because if you look at the maybe it takes 10 years, maybe it takes 15 years, maybe it takes 20 years. But if you're looking at a time period of a couple of hundred years, supply will react to demand just like in peanut butter. So then there is a limit to how far prices can go up. And if the supply in the housing market would be as elastic as in the peanut butter market, we wouldn't see fluctuations at all.

Prices would just stay flat. Now we do see some fluctuations, but in the long run, prices are flat, just like in peanut butter. And it makes total sense because supply reacts to demand. So I want to talk about bubbles for a little bit now and the concept of a bubble.

When people speak about bubbles, usually they're referring to a large discrepancy between an asset's intrinsic value and its market price. Before we discuss this question in relation to housing, just tell us what are the key fundamentals that contribute to a house's intrinsic value? Well, you can look at this from two ways. You can really look at, let's say, the economic fundamentals underlying the housing market overall and say, well, it's population numbers, it's the purchasing power of all these people,

so it's the economic GDP in the city or in the country or things like that. So that's one way to look at it. You can also say, well, the two fundamentals of the house price are just the income you could generate with a house and the discount rate you need to calculate the present value of that income, right? That's the approach that we've taken in this house price fundamentals paper with Brent Ambrose and Thies Lindenthal. That's also the fundamental of a lot of the work

of Bob Shiller, where he just says, so in terms of stock volatility and stuff, he just says, okay, what's the income, what's the discount rate, and can we square that? If we look at the income and the fluctuations in that income and take a discount rate that is in line with the risk of that income, then do we get a value and a development of value that is in line with what we actually see in value? So you could say we can predict a value development based on the income and the discount

rate. And if that prediction is in line with what we see, then the market works fine. And then you see that the market far too abruptly to risk. So it seems that the risk premium inherent in the discount rate goes up and down far too much, far more than is justified by volatility in the income. And the income is what ultimately is the basis of the risk. What you would expect then is that what we find in the price of the buildings on the canal is that over the long term, it doesn't really move a lot. And we see exactly the same in the rent.

And so the rental values in Amsterdam also, they move up and down, but in the long run, they don't move like this. They don't go to the sky or something. They're just stationary movements without a trend up or down. Which would make sense because landlords can only charge in rent what people earn in income, and income is usually stable or gradually increasing over time. Absolutely, yeah.

Right, so in that sense, the intrinsic value of the property will always be sort of grounded to people's income. Yeah, absolutely. That's true. So there's a sense in which, quote-unquote, bubble is the wrong metaphor for house prices because the idea, the image of a bubble implies a sudden and calamitous crash, an end where the bubble bursts or it pops.

But as we know from history, house prices tend to rise and fall in long cascades. And even in the famous crashes in Ireland, the US and Spain last decade, it took about four to six years from peak to trough for those housing markets to quote unquote crash now something interesting you picked up on in your analysis of house prices is that that when prices move away from equilibrium in what might look on the surface to be bubble conditions in many cases that's just followed by a slow decades-long return back down to equilibrium. So we don't even see the four to six-year crash. It might take decades. Yeah.

Now, I just want to clarify, is your conclusion, therefore, that bubbles don't exist or can't be predicted in housing markets or do you just say that it's more difficult than people think well i would never say that bubbles don't exist although it's really difficult to predict them and to even even uh let's say diagnose them when they are there. Even Bob Shiller wasn't so great at it because I know a paper that Chip Case and he wrote just before the bubble burst in the United States where they said, well, there's not really a bubble because if you also look at what wages have done

over the past decade, then the current house price level is totally explainable. So even the man who was supposed to be the guy who called the bubble was not so sure that there was a bubble. So it's very difficult to say when you were in a bubble or not. But also very difficult to say that bubbles do or don't exist. So I'm not going to say they don't exist. I think they do. I think

there is something like irrational exuberance and housing markets can get over the top. And there's sort of a feeding frenzy that people just want to put their money into housing. You see, you've seen it also in the Middle East in places like Dubai. I think you've seen it in China where if you look at house prices and house price to income ratios, it's completely crazy what has happened there and that's clearly not sustainable. But on the other hand, most of the movements up and down that I've seen in Amsterdam are pretty well explainable. And then the question is, are we talking about the bubble or is the housing market just responding to a fundamental? And the fundamental may go away, so the prices may fall. But then you're really not talking about a bubble.

So if I look at this housing market study that I've done for Amsterdam, I haven't really been able to identify clear bubbles. Yeah. So, for example, there was one big run up of house prices in the late 1970s that turned into sort of a crash in the early 80s. But it was completely, let's say, explainable from if you look at mortgage rates and the way mortgage deduction works and stuff like that. So house prices went up because there was this huge mortgage subsidy or let's say a tax subsidy. And then that reverted in the 80s so yeah it crashed yeah so there wasn't really a bubble it was really more response to a policy fundamental

so i want to put something to you i think you have to be really careful about how you define a housing bubble so my definition for a housing bubble has three parts. Firstly, a very large run-up in house prices. Secondly, a very high price to rent ratio. And thirdly, all of that is driven by speculation as to capital gains. And I think it's that third part, which is quite important that you might like a lot of those run-ups in house prices or the changes in rent to price ratios that you saw in amsterdam over the centuries may not have been bubbles because you need that third element of the speculation as to capital appreciation for it to really be for it to really be a bubble because it's the speculation that implies a fall or a price decline because it's got the nature of a Ponzi scheme.

That's right. Yeah, and so there you would really see – so you're right that you would normally see in a bubble a very high price-to- rent ratio or a very low rent to price ratio. But of course the interest rate plays a role there as well because currently mortgage interest rates in the Netherlands are 2-3% or something. And if you have a very low interest rate then you know normal discounted cash flow thinking would tell you that you have a high price to rent ratio. Because values are just really high to rent because of a very low discount rate. And then if the discount rate goes back up, yeah, then prices will fall.

But then again, is that a bubble? No, probably not. It's just a reflection of discounted cash flow thinking. And it's not even enough just to rely on the price-to-rent ratio. I talk about price-to-rent. In your paper, you talk about rent-to-price. Same thing, just a different way of expressing it. Sure.

But in the later paper you referred to where you're looking at fundamentals, I think you and your co-authors wrote that while the rent price ratio is a measure of house prices relative to fundamentals, it does not give a complete picture of the housing market. For example, in the period 1781 through 1815 saw a dramatic rise in the rent price ratio from 4.6% to 11.4%, suggesting that prices decreased relative to rents. However, during this period, rents and prices both declined at relatively similar rates, negative 3.9% and negative 1.9% respectively. Thus, even with this small difference in relative declines, the rent price ratio changed substantially. So it's not just enough to look at the rent price or the price rent.

You also need to have the run-up in house prices, and that all needs to be driven by speculation. That's right. So with that, what do you think, Pete? What do you think was going on in the world in the 2000s? We had these price run-ups in countries like Australia, the UK, US, Ireland, Spain, South Africa, New Zealand, Iceland, Latvia, but in particular in the sexy global superstar cities, the Hong Kongs, London, Sydney, LAs, New Yorks, even Amsterdam. What was going on? Well, I think there was a lot of speculative frenzy there, and it was fueled by very abundant credit, not even really cheap credit because credit is cheaper now than it was

then, but it was abundant. You could get a price range of, you could get loan to value ratios that were very high, didn't need to put in a lot of money of your own to speculate on the housing market, which is by the way also where housing is really different than the stock market. If you go to a bank and you say, I'm going to buy stocks and I want to borrow 90% of the money that I put in the stocks, the bank will say no go. But if it's a house, no problem. So it seems to be far easier to get credit to buy a house and even to

speculate in housing than in other assets and that I think clearly played a role. So I think there were really bubble situations there. But other markets, for example, like Amsterdam, I don't think the, so what our own paper said then, we had this fundamentals paper, was that in 2007 or 2008, the Amsterdam housing market was sort of an equilibrium if you looked at the price relative to the rent levels

and the interest rate. Yet, prices fell 19% afterwards. I think it was mostly due to government regulation. So the government really when there was already a sort of

lack of confidence in what the housing market would do, the government all of a sudden lowered the maximum loan to value, decreased the mortgage interest rate deduction and did all sorts of contractual things at the same time and that i think uh brought the housing market down so i think that was a government-induced crash

although again you know exactly like you say a 19 decline of prices over three years can you call that a crash yes yeah i don't know. I don't think so. In that list of countries I mentioned, one of them is particularly interesting and that's South Africa because they didn't have a crash. Their price is stabilized. Do you know what happened there? I do not. Yeah. South Africa I think was the biggest, had the biggest price appreciation of them all. And then the market didn't crash. I don't know what happened there or how that worked.

Yeah. So, no, I can't tell you. But Australia was pretty unique too in its never-ending economic boom, you could say. Well, our government stepped nimbly in and resuscitated the housing bubble as it was breathing its last breaths of air. We re-inflated it. We had a first homebuyers boost, and the Reserve Bank, the central bank started slashing interest rates in earnest from about 2012.

I only have two more questions, Pete. Sure. I asked Dean Baker, the economist who's credited as picking the US housing bubble and Great Recession, I asked him the same questions. So I'm very interested to hear your answers. The first question is, between historians and mathematicians, who do you think makes the better economic predictions and why? Interesting question. I think that depends completely on what you want to predict. So let's say if you want to predict how consumers are going to respond to an advertisement or

something, which is also an economic prediction, then I think the mathematicians are better because that is just sort of a big data analysis type thing. And the mathematicians are clearly better at that than the historians. But if you're talking about longer economic, let's say more macroeconomic predictions, then maybe historians are better because they've, because if you look back in the long run, you see the same, the same phenomena, the same fundamentals underlying these private movements, no matter where we go, big 50 years or 100 years or 200 years, and maybe historians are better at understanding that. So I think the long-run macro thing, it's more the historians,

and the short-run stuff, maybe more the mathematicians. That makes a lot of sense. Finally, have you heard what's happening in australia's housing market yeah i've heard that it's cooling off it's a correction and soft landing yeah yeah yeah but there's nothing to be worried about what i see right it's just a bit just a bit of a price correction. Yeah, it's a once-in-a-generation opportunity to buy the dip. Yeah, that's good.

Well, Peter, it's been great talking to you. I thought maybe just to finish, you know, at the end of your original paper on the Herengraacht index, you said that, quote, from 1946, the annual increase in value for houses on the Herengraacht averaged 11.6%. In real terms, the annual increase was still 3.2%, which is obviously much more attractive compared to the 0.2% over time. So compared to the historical returns reported in this article, these averages are very high.

This indicates that data from the post-World War II era is very likely to give an overstated impression of the performance of real estate. This is a problem since most performance studies of real estate are based on data from that period, end quote. What would your final message be to people, particularly Australians, who make sweeping conclusions from data with very short time horizons? I would say look at the long run.

Yeah. Housing is really a long-term investment because let's say you're talking about your own house then you're typically you're entering the housing market when you're 30 or something and you're leaving the housing market when you die it's like 85 so this is a this is a 50-year time horizon of course you're not living in the same house but you're going from house to house as it were but basically exchanging houses in a certain market situation at the same value as you go on. So it doesn't really matter where it's a series of houses, just one.

It's really a long term investment. And if you want to evaluate it, you have to use long term data. Fantastic. Well, thank you so much for joining us. It's been a pleasure. Yeah. Okay, Joe. Thanks so much for joining us. It's been a pleasure. Yeah. Okay, Joe. Thanks so much.

Thanks so much for listening. I hope you enjoyed that as much as I did. Before you go, I have a quick favor to ask. If you liked this episode, please rate and review the podcast on iTunes. I know everyone asks, but it really makes a difference. I make these podcasts for free. They are bloody time consuming, but they're important and I couldn't do it without you.

Finally, for show notes and links to everything discussed in that conversation, you can find them on my website, josephnoelwalker.com. That's my full name, J-O-S-E-P-H-N-O-E-L-W-A-L-K-E-R.com. You can also get in touch with me there or on Twitter. My Twitter handle is at Joseph N. Walker. Thanks again for listening. Talk to you soon.

Ciao. I'm forever blowing bubbles Pretty bubbles in the air