The Joe Walker Podcast - The Modern Empiricist Proving Old Wisdom On Household Debt And Recession - Amir Sufi

Episode Date: December 19, 2019

Amir Sufi is the Bruce Lindsay Professor of Economics and Public Policy at the University of Chicago Booth School of Business.See omnystudio.com/listener for privacy information.

You're listening to the Jolly Swagman Podcast. Here's your host, Joe Walker. Hello there, ladies and gentlemen, boys and girls, swagmen and swagettes. Welcome back to the show. I am delighted to share with you this conversation with Amir Sufi. Amir is the Bruce Lindsay Professor of Economics and Public Policy at the University of Chicago Booth School of Business. Amir is a young economist. In 2017, he won the Fisher Black Prize, which is given biannually by the American Finance Association to the top financial economics scholar under the age of 40. Notwithstanding his youth, Amir has amassed a body of work that is as prolific as it is important. Much of it was summarized in his 2014 book, House of Debt, co-authored with his frequent

collaborator, Princeton economist, Artif Mann. Former US Treasury Secretary, Larry Summers, called House of Debt the most important economics book of the year in 2014. And in my humble opinion, as someone who is not an economist, it's one of the two best books to come out of the 2008 financial crisis and subsequent Great Recession. The book, as well as much of Amir's broader research, is important and groundbreaking in that, first, it places credit at the center of financial bubbles, and second, it challenges the prevailing narrative that the financial crisis and Great Recession were essentially failures of the system of financial intermediation. So what explanation does Amir offer in place of this one-dimensional banking view, as it's sometimes called? Well, this actually raises a larger and

more mysterious question, a question which lies at the heart of this podcast episode. What causes protracted economic downturns? This question is actually more spooky than you think. For example, nothing fundamentally changed in terms of the economy's ability to produce in America during the late 2000s. There were no epochal disasters, no world wars or meteor strikes that wiped out chunks of the economy. The society didn't suddenly become less productive, yet spending collapsed and millions of people lost their jobs. So what happened? How did the worst financial and economic crisis since the Great Depression seemingly fall from the sky? Amir believes the answer falls squarely on household debt, noting the two-edged sword of leverage. That is, when asset values fall,

equity collapses disproportionately. The damage this inflicts to consumer balance sheets is enough to send the economy into a tailspin. Amir was out in Sydney lecturing at the Reserve Bank of Australia, and I was lucky to catch him last Saturday. I enjoyed our time together immensely before, during, and after our podcast interview. I only wish I enjoyed our time together immensely before, during,

and after our podcast interview. I only wish we had more time for the recording, but we had a lunch to get to. Nevertheless, I think you're really going to enjoy this chat and I commend Amir's work to you. So without much further ado, please enjoy my conversation with Amir Sufi. Amir Sufi, thank you for joining me. My pleasure. It's lovely to meet you. I reached out months ago

and discovered that you'd be in Sydney. So it's fantastic to be able to meet you in person because your book with Atif Mann, House of Debt, it's one of the best books I've read on the financial crisis and the Great Recession. Oh, thank you. 12 chapters of brilliant analysis. Although for a book on debt, it falls just short of a chapter 13.

Yeah, exactly. Maybe on purpose. Yeah, the Great Recession spurred a lot of research. It was great for economists in the same way that a plague is great for undertakers. Yeah, bad for everyone else, but maybe good for economists. But your research is some of the best I've read to come out of that whole episode. I appreciate that. Yeah. And as somebody who lives in a country with the second highest household debt to GDP ratio in

the world and some infamously expensive housing stock, I thought it was important to share your research and your views with our audience. First, I'd just like to ask you about your partnership with Artif Man. How did you guys meet? You've been so productive together, but how did that begin? Yeah, we met at Chicago, but we knew each other and had exchanged emails before. Both of us got our PhDs from MIT in economics, but we did not actually overlap. So when I started at MIT, Atif had just started at the University of Chicago at the Booth School of Business in the finance group. And a mutual friend of ours, a guy named Asim Khawaja,

who is one of Atif's co-authors, and he's a good friend of mine. I've never written with him, but I've known him for a very long time. He basically put us in touch with each other because he said, oh, you know, Atif had just graduated from Booth, and he also was interested in finance, and you guys should talk. So we exchanged some emails then about who I should talk to at MIT, who should I, you should I work with. Then when I was on the job market, that would have been four years later, we met up again. And then once I got the offer at Chicago Booth, I went there and then we started working together.

So we were on the faculty at the business school at the University of Chicago together for probably five years, I would say. And then he left to Berkeley and then ultimately ended up at Princeton. But we were close collaborators during those four or five years together. And in particular, I still remember when he came into my office, and I think it was a nice combination of skills in the sense that Atif had been studying mostly emerging economies and was an expert at spotting big problems related to countries borrowing too much. And I had a background in the US and household finance. My advisor at MIT is a guy named Jim Piterba. And he does a lot of research on household finance. And I was a research advisor for him. I mean, a research assistant for him. And he came to my office and said, this was in 2005, six. He said, this thing's crazy. Like, have you seen what the household debt to GDP ratio looks like in the US?

It's just, it's insane. And people are very clearly borrowing aggressively against their houses. Most of the existing research suggested that people don't borrow against their houses, which we thought was crazy. At least they don't consume out of that borrowing. And it started there. You know, we then just started to really investigate things. We were interested when we first started on the boom side, we were just wanting to understand why household debt went up so much. And then it took a couple years to get all the data together. And by that point, the crash was starting. So we were kind of very, you know, very well positioned to start thinking about what was causing the downturn. Going back further, what initially interested you in economics?

Why did you decide to do a PhD? That's a great question. I have always, you know, I consider myself almost a social scientist before I consider myself an economist. I think of economics as the subdiscipline within social sciences that I feel like has the best techniques and the best tools to really get at the heart of the questions in social sciences. But I'm fundamentally, I have since a pretty young age, I would say probably in high school, I really always thought just studying the way people behave and the way institutions form and the way the economy interacts with people. I've always been fascinated. I mean, even now when I come to a new country,

this is my first time in Australia, I'm just looking around, how do things work? How do people interact? What are the differences? And so in that sense, I'm very interested in sociology, anthropology, all of the social sciences. I went to Georgetown University as my undergraduate, and it was at that point that I had pretty broad exposure to history, to sociology, to psychology, to economics. And it was at that point that I really felt like economics had the hard tools, especially the use of data, that attracted me the most to that particular field within social sciences. And so it was never so much about I was super interested in stores or prices or financial markets. It was just this is the subdiscipline of the social sciences I feel like is doing the most rigorous analysis to actually find out what is going on.

And that was what attracted me to economics initially. As an outsider visiting Australia, and as an economist who experienced firsthand the housing bubble in America in the mid to late 2000s, what do you make of the current situation in Australia and the debate around our housing market? Well, the first thing I'll say is it's not so unique. It's definitely the most extreme of the advanced economies experiencing these very high household debt and house price levels. So a common measure we use to try to assess what's going on in the housing market is something like a house price to rent ratio

or a house price to income ratio. And that usually is being driven by another variable we look at, which is the household debt to income ratio. These are very closely connected because obviously people have to borrow a lot to buy homes. So if house prices are high, it generally is the case that household debt to income is high. I think when I look at Australia, you see this long run secular trend that you've seen in a lot of countries. Since

about the 1980s, there's been a secular rise in house prices and household debt burdens across a lot of these countries. Scandinavia, for example, the United States, Canada. One big difference is obviously you all did not experience such a big correction in the Great Recession, either in house prices or in household debt levels. The United States is really on a completely different trajectory now because the Great Recession was so severe. House prices fell so much. Household debt burdens fell so much. Whereas Australia, I think there's reasons for this, but we can get into those later, was able to kind of avoid this like big correction and house prices kind of continue to march upward, at least until recently in the last couple years. So in a cyclical sense, I don't have

any kind of real strong insight into what's going on in Australia, whether house prices are going to go up or down. But in a long run secular view, Australia looks similar to a lot of advanced countries. And I think there is some global factors that can help explain what's going on. Which we can talk about. Yeah, for sure. But I want to get into your research now. And the first thing I want to pick up on is to say that a lot of the papers you've published with Arteev and other collaborators as well along the way have essentially been proving out or empirically validating an older tradition in economics,

which is the Kindleberger-Minsky view of credit and bubbles. Can you just begin by outlining the Kindleberger-Minsky view? Right. So the question really starts with what causes crashes or financial crises. And I think if you read Kindleberg or Minsky, that's really where they're starting. You see the systematic pattern. We obviously have it going back a long ways in history. You know, the tulip bubble is an earlier episode, but, you know, people have argued in ancient Rome, you saw very similar things, you know, credit supply expanding and then a crash. So what is fundamentally causing that? And I think Kindleberger and Minsky both put a primary role for credit, that credit really is a financial instrument that together with maybe changes in beliefs or changes in perceptions about

what's going on, end up causing big booms in asset prices generally. Housing is oftentimes the primary asset that's affected. And that during the boom, credit essentially fuels that bubble. That credit is absolutely key. And I think that's a difference between, say, a pure behavioral view that it's just people have irrational expectations or optimism. That, of course, is part of the Kindleberger and Minsky narrative, but credit is crucial in driving that process. And the reasoning behind that is there may always be optimists out there. There may always be people who would be willing to pay an outrageous amount for a house, but the point is if they don't have any money, they can't buy. They can't set the price for the asset. And so their focus and our focus and our research is really on what might shift the willingness of the financial sector to be willing to provide loans to people

that otherwise or maybe previously could not get credit. So that's really at the essence of the Kindleberg-Rominski narrative is something leads to a situation in which banks or the financial system more generally is willing to give credit to people that previously couldn't get credit. And those optimists, those people who couldn't get credit before, bid up the value of assets until kind of everyone realizes, oh, this has gone too far. And then the crash comes. In seeking to solve that puzzle or cut that knot as to what kicks off a bubble, in this case, a housing bubble, is it animal spirits, to use Keynes' term, or is it a credit supply shock, to use your language? You did some really interesting research which exploited elastic

versus inelastic cities to try and resolve that causality right tell us what elastic versus inelastic means and what your research found so the idea is uh take an area of the country in the united states i think of like where i grew up i grew up in topeka kansas um where it's really easy to build houses and um there's a lot of land availability, and there's not too much demand to live in any particular space. We call that an elastic housing supply city. What we mean by that is if people all of a sudden want to buy more homes or get a desire to have more space in their homes, it's pretty easy for the construction sector to just build those homes. And the idea is in those kinds of environments, you shouldn't see large increases in prices, nor should you ever even expect a large increase in prices. And in the extreme case,

in areas like Kansas, in the United States or Missouri, that ended up bearing out in the bubble period from 2000 to 2007, when house prices are going up bearing out in the bubble period. From 2000 to 2007, when house prices are going up a lot in San Francisco and New York and Miami, they're not going up at all in Kansas. They're basically flat. So the idea was, let's focus on an area like Kansas where we know people can't just have irrational expectations about house prices. In some sense, prices didn't go up and it'd be kind of extreme to view that they could ever really go up. And we basically said, even in those areas, there was a really big expansion in subprime lending, in credit availability. So what we're

basically trying to say is like, look, credit expanded even in areas in which I can conceptually shut down any behavioral economics or any irrational expectations channels. And so then when we look at a place like San Francisco where both could be going on, the point is it seems like if credit supply is expanding credit in Topeka, Kansas, then it's probably also expanding credit in San Francisco. So that was the kind of conceptual idea behind that thought experiment. Got it. Now, there are two exceptions to your thought experiment. And indeed, these are two enduring mysteries in the US real estate convulsion. And that is Las Vegasenix yeah um two cities which have virtually unlimited supply yeah i remember driving i was road tripping with a mate in 2016 driving into vegas from san francisco and it's just desert yeah for miles 360 degrees around the city now somewhat bizarrely given they we would describe these as supply elastic cities yeah real home prices increased 10 in vegas in 2003

and then 49 in 2004 in phoenix it was 10 in 2004 and 43 in 2005 yeah how do you make sense of those extreme run-ups in two cities with virtually unlimited supply so i I think there's two points. First, it matters how much you believe these measures of housing supply elasticity. I think your perception of those places that has unlimited housing supply, at least in the measures that are used in the literature, is not quite accurate.

Now, of course, they're more elastic than, say, San Francisco or New York, for sure. But I think people who do these measures, they use these geographical maps, they think about water supply, they think about how hard it is to build too far out. They would say Las Vegas and Phoenix are less elastic than, say, Kansas. Now, your broader point is, of course, correct. They are more elastic than San Francisco, and yet they experienced, if anything, even higher house price growth. So that doesn't take away from your central point. In our more recent research, we really focused on these mysterious, what are often called sand state, or sand cities. You know, these are areas that tend to be in the southwest in the U.S. or in the

sunshine state in Florida. These are the real, basically Nevada, Arizona, and Florida are really the three states that fit this point, which is they seem to have reasonably elastic housing supply and house prices rose a lot. It turns out that those states were by far the ones most exposed to the sudden expansion of private label securitization driven credit in 2003 and four. Let me put it differently. There are certain banks, I'm sure you have the analogous banks here in Australia. In the United States, we had a couple banks, one's called Countrywide, one's called Washington Mutual. These guys expanded credit so aggressively starting in the middle of 2003 through 2004, 2005. And they, even 10 years prior to that, they happened to be located primarily in these three states. And so whatever is going on, I'm not going to try to deny that some behavioral phenomenon may

have driven what happened in those states. Our central point of some of our more recent research is that credit looks like it was an incredibly important part. In the extreme, maybe you could argue that credit availability, you essentially can go out so far on the optimist curve of buyers that you can even get people to buy in an elastic housing supply area where you think there's no reason you should be buying here. Eventually, house prices almost have to crash. So I don't want to set up as if I don't believe the behavioral story. I do believe that there's important behavioral biases. I just want to say that I really do believe that credit is a necessary condition. It might not be sufficient. Maybe you need credit and crazy beliefs, but I really do

think credit is a necessary condition to get these dynamics. Let's riff on that for a moment. In your book, you cite research by John Giannacopoulos, a Harvard economist, has a great paper called The Leverage Cycle. And you use this for the basis of a causal explanation as to how credit actually transfers pricing power from pessimists to optimists. Yeah. Can you break down that explanation for us? I think it's really important. Yeah, this is, the John Giannacopoulos work, I think, has one of the most interesting theoretical insights that I've seen in finance in the past 20, 30 years.

It's just a fascinating insight, which is the basic structure of the argument is, suppose the population contains a group of people, let's call them pessimists, and a group of people called optimists. And what we mean by pessimists and optimists is really beliefs about where the future asset price is going to go. So let's think house prices. Some people in Sydney think house prices are going to go up 10% per year. Some people in Sydney think they're going to go up 3% per year. The question is who on the margin is setting prices in the market?

And the John Acopolis insight is let's suppose that the 10% people, the optimists, they buy up as many houses as they can, but they don't have that much money. I mean, how much capital does any group have, right? So at some point, they run out of houses to buy. And even though they have these very optimistic expectations, they don't set prices in the market because they've already bought houses. And so the marginal guy buying houses is a 3% guy. And so the price is going to be kind of low that's transacted. And then he says something so brilliant, I think he says, but let's suppose we introduce a financial instrument called credit, in which we allow the pessimistic guy, instead of buying a house, to lend to the optimistic guy, right?

And that kind of you can see right away what's going to happen is that now more credit goes to the optimistic guy. The optimistic guy then bids up the asset price, right? Now, of course, the thing most of your listeners may be thinking in their head is, well, wait, why would the pessimist ever lend to the optimist if they kind of think the optimist is a little crazy? And that's really the brilliant insight is that it's a debt instrument. And the way debt works is that you get the senior claim on the asset. So it's basically you're thinking, okay, I'm going to lend to you. I think you're probably a little bit crazy, but I know that if you default, I at least get the home, which I

think is why you'll still lend. Yeah. So brilliant insight. So it's a way in which if all of a sudden the bottom, the pessimists decide that they're willing to lend a little bit more to the, that you can get a big movement in asset prices. So small movements and credit availability end up having big effects on prices. And I think that's the insight that I think is really, really quite valuable. For most people, a bubble entails a crash. And insofar as that is our definition, is there any way we can identify housing bubbles ex ante? That's a great question.

I think where finance has gone, not just with housing, but with all assets is to think very carefully about valuation metrics. So price to rent ratio is the one that people typically focus on. So this is the equivalent in a housing market of a PE ratio. Yeah, exactly. And it builds in both those components. We've got the market price and then the intrinsic value. Exactly. Exactly. And the idea is that, you know, the rent is ultimately the cashflow you get. And so why in the world would you ever see a rise in the price to rent ratio?

There's basically two explanations on why the price to rent ratio can go up in a rational model. In a rational model, the only way a price to rent ratio can go up is if you expect higher cash flows rationally. So you expect rents to go up over time, right? Or, because remember,

it's always the current price to the current rent ratio. So if you think rents are going to go up, then you're willing to pay a higher price because you're going to be paid off in the future. The other one is that you rationally expect that returns are going to be low, that you kind of, now, why would you do that? Maybe you're just, you know, you're a little bit like to take a little more risk or whatever. I mean, I've never really thought that explanations for that in the rational model are so good. So I can't explain why that would happen. But that's how the rational model views it. Now, what do we know from the data? We know from the data that systematically high price to rent ratios tend to predict at a two to three year horizon, a decline in returns, meaning, no, I don't want

to say crashes. Crashes is maybe a little bit too strong, but at least lower returns than were experienced prior to the rise in the price to rent ratio. Now you're speaking about the United States. This has been shown across a number of asset classes across a number of countries. I don't know specifically in any given country, but this is, you know, I remember this very fascinating research going back all the way to the 17th century using Dutch and English stock prices. And you see this pattern in a lot of environments going back in history. Now, this ability for high price to rent ratios to predict lower returns, when I first learned this, and I'll be frank, even though I'm a finance guy, I didn't really fully appreciate this until a few years after my PhD. proponent of more rational type models, where he shows this very clearly, that high price to rent ratios or price dividend ratios in stocks is what he was focused on, tend to predict lower returns,

not higher dividends. When I heard it, I immediately became more of a behavioral economist. I mean, I just, for me, it's, it just sounds like a behavioral point. How can you, how can you know when high price to rent ratios are there? Why would you willingly pay that high price if you rationally knew returns would be low? So that should be your base rate. Exactly. People act as if they've neglected or forgotten that base rate. Exactly.

And so why would they do that? The behavioral guys come in and they say, because people have irrational expectations on rental growth. They think the rents are going to go up or they think the prices are going to continue to rise. Irrespective of the rents. Exactly. And I think the survey evidence is very consistent with the irrational view. And I'm thinking especially of Andre Schleifer's work. He works with one of my colleagues, a woman named Yaron Ma.

And they've just shown in so many environments at the peak of these price to rent or price to dividend episodes, if you just ask people, what do you think your returns are going to be? They say, I think they're going to be high, not low. In the data, we know they're low. We know going back using so many samples that they're actually going to be low. If you buy the peak it's going to be a bad thing but people don't see that so i think one of the most striking survey results i've seen are the case schiller surveys of people in certain cities going back to i think they started them in 88 didn't they right um and and home buyers recent home buyers in los angeles in 88 you know, towards the peak of the mini housing bubble in the late 80s. And again, in the 2000s, think that their house prices are going to appreciate about 14% per year on average for the next 10 years. And at the same time,

most of them say there's little to no risk in buying at the moment. Yeah. What's amazing is about Andre Schleifer's work is that maybe, you know, your listeners are thinking, yeah, there's probably some irrational people in Los Angeles. They show that even CEOs of major companies make these same mistakes. So that if you actually look at what CEOs are saying about projecting their own cash flows, their own returns, they also show the same behavioral bias, which is when my stock price is high relative to my dividend, I somehow think for sure my dividend growth is going to be high or my price is going to still go up. Even though statistically, if I look at the past data, it's very clear that you should, you know, this is a very common notion in finance, right? You should not be buying when the price is high, right? You want to buy when the price is low.

And yet somehow people convince themselves to continue buying when the price is high, right? You want to buy when the price is low. And yet somehow people convince themselves to continue buying when the price is high because they have these expectations of cash for growth or price appreciation. Now, Schleifer's explanation, which I think is so neat, relies on Danny Kahneman's representativeness heuristic, which people might remember if they've read his popular book, Thinking Fast and Slow.

And the representativeness heuristic is basically about we judge frequency or likelihood based on similarity. So the question people ask themselves when they get a piece of good news, be it about a particular stock or the housing market, is not, you know, what's the base rate here? But it's, is this representative of a future state in which there's a boom exactly this feels

like a boom and they kind of you know fling their expectations forward like that yeah um i find that fascinating yeah were you persuaded by that yeah it's a really nice piece of research i mean because it takes the best behavioral economics research take behavioral biases that have been documented and proven in the lab by psychologists or behavioral economists and then try to extrapolate that well-disciplined behavioral bias in this in this case the heuristic bias and then put it into an economic model and make sense of what you should expect. And what you just said is exactly what you should expect. And in some sense, people overweight what they see recently. It becomes more representative.

And then they think even though they should be looking at the full historical data, they don't. And as a result, you get this kind of momentum where like prices have just gone up. That makes me think they're going to go up even more even though that's not what the historical data suggests yeah now the mainstream neoclassical consensus in australia and i've spoken with with many neoclassical leaning economists about this is don't worry about high price to rent ratios historically high price to rent ratios because interest rates have been low and declining since the late 1980s when they were incredibly high and if your real interest rates are lower the discount rate by which you're capitalizing rents into prices is lower therefore we should expect to see higher

prices and higher price to rent ratios my response is that relies on a rational expectations assumption and i'm not entirely convinced that the present value relation is the model of the economy in the public's mind their response in turn is just to say but people act as if it is i'm not convinced because i think i know from survey evidence for example um dallas rogers an academic at the University of Sydney, has an interesting survey he ran in November 2015 where they actually asked a group of 899 Sydneysiders what they thought drove house prices. The number one factor people selected, it received 64.4% of respondents, was foreign buyers, which is outsized in relation to the actual impact foreign buyers have had if you trust Treasury and RBA analysis. To me, that seems like a narrative driving people's belief and rationalizing the elevated house prices. I think if you asked most Australians what a real

interest rate was, they wouldn't be sure what you were asking. How do you think about the role interest rates play in housing bubbles? So, you know, my whole career, I've always been fascinated by the fact that a lot of the people who are more proponents of kind of pure rationality and financial markets, and I'm at the University of Chicago, so obviously I have a lot of the people who are more proponents of kind of pure rationality and financial markets, and I'm at the University of Chicago, so obviously I have a lot of those people around me on a day-to-day basis, how heavily they rely on the present value relationship to kind of prove their point. And just for your listeners to make sure, because I think it's such an important equation, and it's such an important way of thinking about some of these issues about how people are willing

to pay high prices, the present value relationship just says that any price of an asset is the expected cash flows in the numerator, like how many payments you respect to receive, whether they be rents or dividends or whatever. And then you discount them back with a discount rate. That's in the denominator. And the idea is that if the denominator, if the discount rate is lower, meaning low risk-free rates like your rational proponent friends, then that boosts the price of the asset. And the rational guys are right in one sense. This is a mechanical accounting identity. It almost has to be the case that the price is a function of expected cash flows discounted. But that does not in any way get you out of behavioral biases.

You just have to discipline the behavioral biases. What are the behavioral biases that could lead to wrong prices? It tells you either you have expected cash flows that are incorrect, you think they're going to keep going up, or the actual interest rate that you're paying is artificially low relative to the true discount rate and that could be for example because banks are willing to lend to you because they have bad beliefs so in some sense this present value relationship doesn't get you out of behavioral

biases so in the context of your specific question you can easily tell a narrative story which is for some reason maybe it's chinese homebuyers, or if it's foreigners buying homes in Australia, whatever it is, people have bad or incorrect expectations of future cash flows, which then boost up price to rent ratios in the present value formula. But it doesn't, you know, that's perfectly consistent with the fact that we should expect a big crash in Australia. I'm not saying that's the case. I'm just saying it could very well be that people have incorrect expectations of cash flows, and that's why they're willing to pay such high prices. But in some fundamental sense, it's a bubble because those expected cash flows are incorrect. To your specific question, I do

have some sympathy for the view in a long run secular sense that it is the case that real interest rates, risk-free rates in particular, have come down since the 1980s and that therefore all assets in the economy will have higher price to dividend or price to rent ratios. That comes straight from the present value formula. That's an accurate statement. But there's a quantitative statement that these people are making, which is it's only the decline in the risk-free rate or the interest rate that explains why the price-to-rent ratio has gone up. And I find I'm

usually skeptical of that view. It's usually more than just that that's leading to high prices. It could be things like you said, people have irrational expectations on cash flow growth, for example. So far, we've discussed the two necessary ingredients for a housing bubble. We have credit and we have crazy beliefs. Your contribution has been to say that we shouldn't treat credit as a sideshow and crazy beliefs as the main game. Credit plays maybe a more important role than some traditions have given weight to. If we accept that credit supply shocks are what kickstart housing bubbles, what causes the credit supply shock?

So what we focused on, that's a really difficult question, by the way. That's definitely what we've been focusing on most recently in our research, especially since the book. I think that we view the most likely explanation when we look at the historical record is some shock or some change in the economy that leads to a rapid influx of savings into the financial system. So let me be concrete. The most obvious example of this, which has been quite well proposed and defended by Ben Bernanke, and I believe, is the global saving glut view. Okay, so the global saving glut view is what we tell in the book, is that basically around 1997, 1998, it's actually a fascinating narrative, because it shows how bad political decisions

can kind of screw things up badly. A lot of the East Asian countries get in big trouble in the late 1990s. And the narrative is that the IMF and some of the other international organizations treated these countries very unfairly when they came in and imposed discipline on them to help them bail out of their problems. So the narrative goes that as a result of that, these East Asian countries, and China in particular, saw what happened and said, we're never going to get ourselves in this problem again. And as a result, they started to stockpile dollar-denominated assets. They started to buy huge amounts of U.S. treasuries. They started to buy huge amounts of bonds from a lot of advanced countries. What does that mean? That means a huge amount of money is coming into the financial system that wasn't there before. The idea is like a very old one where you have a bunch of money sitting around,

you're probably going to spend it more, right? I mean, you know, you go out to the bar on a Saturday night with a wallet full of money, it's probably going to be all gone. And that's the same idea that the financial sector sees all this capital coming in, they're going to try to find something to do with that capital. We all hope that there's some real productive investment that you can do, do whatever, you know, you fund a new business or you fund bridges, whatever you need. But it seems, unfortunately, that in a lot of advanced countries, instead, it just ends up being funneled to households. It can be funneled to households through mortgage lending, it can be funneled to households through auto lending. But that's a concrete example of what I think of as the kind of shock that leads to the financial system being

more willing to lend to people that previously couldn't get credit. The global saving glut is one view. If you want another one, which I find fascinating, is the Latin American debt crisis of the 1980s. I give credit to Michael Pettis, who's a scholar who works in China. This is where I first heard this, and I did some investigation, and I think he's absolutely right. The argument is, in the 1970s, we had the OPEC oil shocks that led to huge rises in oil prices. The oil market is one in which dollars are used to transact, U.S. dollars. That led to a situation in which a bunch of oil producing countries, think Abu Dhabi, Dubai, Iran, Saudi Arabia, they end up having huge new cash flows because they're just basically making so much money on the same unit of oil they were selling.

The argument goes, what did they do with those dollars? They put them in international global banks, Citibank in particular is an example. All of that money comes into these international banks. It's all a bunch of dollars. What are those banks going to do with that money? And the argument that Michael Pettis gives, and I think the historical record shows this, is they turn to Latin American countries and they start lending like crazy to Latin American countries both sovereign debt and corporate debt that leads to a boom in Latin America huge amounts of increase in leverage then of course Paul Volcker raises interest rates hugely in the 1980s and the whole game ends and you see this massive crash so that's another concrete example of you know why was it

it was this OPEC shock that led to huge amounts of money going to these financial systems and they turn around and start lending it very aggressively. Does understanding how housing bubbles begin give us any insight into how they end? Like, what pricks housing bubbles? That's probably hard to answer in the abstract, but historically, how have housing bubbles ended? You know, I can speak about the US experience in the Great Recession, because that's what I know best. And I can tell you that what it looks like, and this is very consistent with Kindleberger and Minsky, by the way, this is what they would say, is that essentially you stretch so far that you get to such a low credit quality credit

that essentially you give this credit and within a month, the person's defaulting on it. Now, that doesn't necessarily answer the question because, well, if prices kept going up, presumably they could work their way out of that default. They could refinance. So it's a hard question to answer, but the credit boom gone bust narrative is that you consistently see the same pattern, which is the last guy to get credit is the first guy to default.

And that's when the game's over. You've run out of marginal buys. Exactly. You've kind of hit to a point where you've gotten to the most optimistic guy. He tries to sell to someone even more optimistic and everyone says, you're crazy. And then you kind of look down and you realize there's only people below you in that hierarchy. Yeah. It's not a very great answer. I'll admit that I do think that if I were to say where I look to behavioral economics the most is in trying to answer questions like this,

which is what really defines these turning points, both on the boom and the bust. The credit supply view that I've been giving, I think of it as less able to answer that question. Because whatever is driving the banks to pull back, I mean, the banks pull back when they start seeing defaults. It's very clear. They start seeing initial defaults in the middle of 2006, late 2006 in the US. They start pulling

back as early as December 2006. You see it. And defaults start to rise as early as December 2006, January 2007. People find that surprising because they say, oh, the Great Recession was night 2008, Lehman Brothers. The housing market was collapsing starting in the beginning of 2007 in the US. And that's why we always place that as the central part of what caused the Great Recession, not Lehman Brothers and not these other arguments. What I find intriguing is if you ask Bob Schiller how the US housing bubble ended, he'd say it was a swing in public opinion and the narratives changed. I'm not sure I find that so persuasive, but it seems like that would be sufficient

to end a housing bubble. Yeah, I totally agree. Sufficient, but it's not so obvious that... Yeah, when you talk about it very high frequency, what exactly is the trigger becomes very tricky. There's all these things happening at once. Now, there is an interesting piece of international evidence,

which I think supports the, the credit supply narrative as far as what initiates the housing bust, not, not just begins the bubble. I don't know if you're aware of Sergi Basco, a Spanish economist. He did his PhD at MIT and his thesis was on housing bubbles. He has an interesting paper where he,

he almost thinks that he can pinpoint the moment the Spanish housing bubble ended. And the day was the 9th of August, 2007. Okay. And what happened on the 9th of August, 2007 was BNP Paribas announced that it was freezing three of its funds because of the securitization market. Yeah.

That dried up liquidity to the Spanish banks. And from that moment, you can see the average mortgage size in euros start to deteriorate. So very, I don't know about who had the earlier day, maybe BNP Paribas, but that was August, 2007. I can still remember the summer for academics

is the best time of year. We just sit around talking about our research and we were all sitting around the lunch table at Chicago. And that was around that time that Bear Stearns first announced that they were seeing some losses on some of their credit instruments that were supposed to be all invested in AAA security. So there were two or three weeks in August 2007, where you could really see something's very wrong. So I think that's fascinating. I would say that, in my view, it's a little late. Like, I think something happened earlier that, you know, that again, if you want to talk about the exact peak

of the credit cycle, I think it ended probably three to five months earlier than August, 2007. For the U S I, yeah, for the U S I'm not sure about for Spain, but for sure this event in August, 2007 was when all economists, I would say, at least those interested in finance, our ears kind of like, you know, he started looking around saying, okay, something's different. Something's going wrong. Now, in favor of Andre Schleifer's appreciated, is the fact that the credit side of the economy, as early as August 2007, becomes incredibly worried about what's going on. From August 2007 until basically September 2008, most credit instruments in the economy had shown heightened levels of risk awareness, a dramatic re-evaluation of what's going to happen in the

housing market. And yet equity markets in the United States continued to rise during that period. And I still remember, and that's why I'm an academic, I don't have the guts to be like a hedge fund manager, but I still remember being a credit guy myself thinking, this is crazy. How does the equity market not see what the credit side sees? And there's big problems coming up. I think Kindleberger referred to this period as a period of stringency. Exactly. Sort of at the top of the bubble where things are continuing, but there's growing anxiety. And then people overreact to the downside on that same mechanism of representativeness if you take schleifer's research yeah

no it's really interesting over the last to make it a little bit more recent you know we have seen episodes like this recently so for example for sure in the fall and winter uh the fall of 2015 slash winter 2016 there was this big scare china basically left their peg and you know you saw some pretty serious capital outflows there was a rise scare. China basically left their peg and you saw some pretty serious capital outflows. There was a rise in every measure of risk premium that you can see in the US during this time period. I remember in January, February of 2016, a lot of people were predicting, okay, now here's the correction we've been waiting for. And this is going to throw a wrench. This is not something I'd really plan on talking with you about. But I do think that the Fed, like the Fed stepped in quite aggressively during that time period, January, February, March, and in some ways seemed to stave off that correction.

You know, I mean, and that's been the game that seems to have been being played. It's played out recently as well. So there's an interesting question when you're really talking about how does the bubble end, you know, on what does the monetary policymaker do? What does the monetary authority do? And do they have some ability to keep the game going when otherwise it would have ended? So we began this conversation by speaking about the Kindleberg and Minsky view of bubbles and credit. There's another even older economic tradition, which your program of research has lent empirical support to, and that's Irving Fisher's debt deflation theory of recessions.

Outline that theory for us. Yeah, so Irving Fisher, it's funny, as many of your listeners may know, is oftentimes maligned because he made some very bad predictions about where the stock market would go in 1933. But I think to his credit, he wrote, in my opinion, people cite Keynes all the time, but I think the debt deflation paper that was written, I think, very early in the Depression, 33, 34 or something, it's in Econometrica, is one of the most foundational papers written in economics. I mean, everyone should know this argument. And it's very simple at the end of the day. It is that elevated debt levels combined with some kind of correction, whether it be a house price decline or a broader economic decline,

exacerbates economic downturns because the people who have debt and therefore need to make payments and have lower cash flows to make those payments have the highest propensity to cut their spending when this negative shock occurs. That's pernicious. Yeah. And I mean, it's such a powerful insight. And the nice thing about it is it almost has to be correct, right? I mean, Bill Gates, who has tons of money, if there's a negative shock to the economy, Bill Gates' consumption is not going to change, right? Whereas, you know, some of my neighbors on the south side of Chicago who are leveraged to the tilt, who have a lot of credit card debt, who

have a lot of mortgage debt, if they lose their jobs, what other margins they have except to massively cut back on spending. And I think that debt deflation idea, his idea goes even further, which is, this is a bit more technical, and I'm not sure it's so relevant today, but it certainly was in the Great Depression, that that pullback in spending then leads to a decline in the nominal price level. And debt is written in nominal terms. So the real burden of debt has actually risen sharply. And so you get this vicious cycle, which is like debtors are already constrained, they pull back their spending, that leads to deflation, that leads to even a higher debt burden, and thus the term debt deflation. Fortunately for the United States in the Great

Recession, and I think Ben Bernanke and the monetary policymakers generally deserve a huge amount of credit for having prevented a deflationary cycle. That would have been devastating and made the Great Recession much worse. So this is what's at stake. And this, I think, also indicates why not all bubbles are created equal. The most damaging type of bubble is surely a leveraged housing bubble. Yeah, for sure. And it's precisely this distributional point, which is, and we make this argument in the book a stock market bubble and i oftentimes joke with my colleagues about this who i think get excessively worried about bubbles um you know if it's a bunch of rich people trading

money with each other yeah and then they all and half of them lose some money half of them don't i mean what are really the macro implications like the the dot-com bust of 2000 was very mild. I aware, but there is a big rise in corporate debt, especially driven by what are called CLOs, collateralized loan obligations, which are securitized instruments that look a lot like subprime mortgage securitization. So a lot of people see the analogy though, oh, this is going to end really badly. It's going to cause a huge recession. And I always say, look, first you have to ask yourself, what are the businesses doing with that credit? Are they distorting the real economy in any way? If you look at corporate investment, it's at an all-time low. Employment

is recovered, but it's not super strong. So what are they doing? They're just buying back equity. So they're raising debt. They're buying back equity. Let's suppose that market collapses. Let's suppose that corporate debt collapses. Who loses money in that market? Who actually holds corporate bonds? It's basically the top five to 10% of the income distribution in the US. And so that's another analogy that's a little bit more controversial, which is, I'm not saying it wouldn't lead to a downturn or maybe a minor recession, but it's not like what happened prior to the Great Recession. It's a whole different ballgame when people in the 50th and 40th percentile of the income distribution are borrowing versus people in the 95th percentile. A lot of people like to talk about the wealth effect and the flip side, the negative wealth

effect when it comes to the effect that house prices have on consumption. You have a really interesting take on this. Can you explain what that is yeah so you know the data suggests and this is very related to irving fisher's argument yeah that not everybody reacts to changes in house prices equally that it's in particular lower income and when i say lower income i want to be clear given the massive amounts of inequality in most advanced countries, including Australia, when I say low income, I mean people in the 80th percentile and below. So I'm not talking about people in the 10th percentile. I'm talking about pretty much most people except for the top of the income distribution. They tend to react to house

prices and their consumption behavior much more aggressively. So if house prices go up, they spend more. If house prices fall, they cut spending by more. And so a pure housing wealth effect story has a hard time explaining that kind of variation. Yeah. Like, why shouldn't everyone equally spend out of a rise in house prices? And our reasoning is, again, credit, you know, that the people in the bottom 80% are relying on credit. When their house price goes up, they're able to borrow more, they spend more. When the house price goes down,

they have to make their debt obligation payment. And so they have to cut spending. Yeah. For me, I thought the most important page or the kind of engine room of your argument in your book occurred on page 51, which is that when debt concentrates losses on indebted households, they stop spending. They tighten their belts. That swings an axe through aggregate demand. Exactly. And that's where we see these really nasty protracted recessions it seems kind of obvious but why why why does someone tighten their their spending when their net wealth has

has been destroyed like yeah so there could be two reasons i think one is the kind of more standard explanation which is just in real terms you you're poor. You used to have this claim on this house. Now, let's suppose you essentially, the price drop has led to a situation in which you have no equity in the home. Effectively, the bank owns the home. You may be living there. At that point, you just don't have as many resources going forward to consume, and so you got to start cutting back. That's kind of, I think, the more standard explanation. The other explanation is a more credit-focused explanation,

which is during the time in which prices were going up, you were using your home to consume. You were borrowing against it to buy a new TV or to redo your kitchen. And so then if you think about growth rates, you know, prior to the housing price correction, the growth rate was really high in your consumption. And then just naturally, when the house price starts to fall, you no longer have that source of financing. And so you have to, you know, it's almost mechanical it's like you redid your kitchen you bought a new tv so your consumption was high now your house price goes

down you're not going to be able to buy an even nicer tv you're not going to be able to redo your bathroom as well and so consumption has to fall and i actually think the second explanation while perhaps less interesting is probably correct you know i think anytime you get a boom where people are borrowing against their homes and consuming if that that bust happens, it's going to be those exact same individuals that were borrowing so much that are going to cut their consumption the most. I mean, this is bad enough, you know, the economic effects and the social costs, the human suffering wreaked by these balance sheet recessions. What about the political aftershocks?

Yeah. You know, that's something we did a lot of research on in the years after and showing that these financial crises driven by balance sheet issues tend to lead to heightened political polarization. It tends to lead to difficulty in decision-making, building consensus. And there's a natural reason why these balance sheet recessions very naturally pit people against each other, because there's a creditor and there's a debtor and there's not enough money to pay both. Right. So you can almost draw the division along those lines, debtors and creditors. Yeah, exactly. I mean, I think that's what happens in a lot of these episodes. It's kind of amazing, actually, when you look at the history of, for example, the Great Depression, and we all know

what happened in Europe and the horrible and terrible things that happened. But if you look back at that, a lot of that was viewed, obviously there was a tremendous amount of anti-Semitism and racism involved in all of that, but a lot of it was pitched as a debtor-creditor issue, which was debtors saying, these creditors took all our money, and now we're suffering. And so there's been a lot of historical research basically documenting that, that that was one of the narratives that was out there, was that these creditors took all our money and now we're poor and so we need to blame them. And so you get that kind of narrative in a lot of different circumstances. In the U.S., obviously, we have the Occupy Wall Street view. I'm not sure how international that phrase is. We're familiar with it. Yeah.

So, you know, targeting the top 1%. You know, Elizabeth Warren's candidacy for president today is very much a reflection of what happened in the Great Recession and the view that bankers essentially got off easy and that people with homeowners suffered, right? I mean, I think even the Obama administration, which is obviously a Democrat center, center left administration, there are a lot of people in the Democratic Party who still very much hold it against President Obama. And I think that also is affecting Joe Biden, that really, when the chips were there, you know, when we had to make decisions, you favored the banks over homeowners. So you can see very naturally why this leads to

political polarization and a lot of infighting. It seems very important that we get a handle on these household debt cycles, not only because of the political consequences and the polarization, but the foregone growth means that we might end up occupying future states that are less prosperous than what we could have enjoyed. Yeah.

Because we're shooting ourselves in the foot every decade or two. Yeah. It's like two steps forward, one step backwards. Yeah. What solutions do you have going forward? Have you thought about this? Yeah. So there are two sets of solutions

that we discuss in the book. One is the technical term is macro prudential regulation. But the easier way of saying that is just the primary financial regulator could step in and just say, we're not going to allow people to leverage up so much you know you can just put a limit on for example i don't know in australia what's the most common ratio people have a restriction on when they buy a house is it a loan to value ratio i mean lvi yeah lvi so so so um so you can have loan to value ratios you can have loan to income ratios and you just step in as a regulator and you say look we're not going to let you and you say, look, we're not going to let, you know, we're going to lot, we're not going to let people have higher than an 80% loan to value ratio when they buy a house. So South Korea does this,

the Bank of England has been doing this, Bank of Israel, this has become much more common. And, you know, that is clearly one way in which you can try to cut off those optimists. You know, if you go back to the John Acopolis framework, the optimist can't buy the house unless they can leverage it a huge amount because they have such unrealistic expectations of where house prices are going to go. It's almost surely they don't have enough money to actually buy the house. So the idea behind that approach is if you kind of cut off the top of the distribution of the optimism, then the whole thing unravels and the prices won't even go up in the first place.

Now, there's some evidence to support that. Certainly, if you talk to the people in the Bank of England, they instituted a debt to income ratio on the banking sector. They basically said, if you're a bank, you can't have more than 20% of your mortgages within loan to income ratio above some threshold. And they claim that that did slow house prices for a couple years in London. Now they started to go up again, and then there was Brexit. So you know, there's other things that happen. That's one set of solutions. I think I'm sympathetic to those solutions to some degree. I worry that at the end of the day, the fundamental reason for these credit expansions is some shock like the global saving glut or the OPEC shock like we had before, that it's a bit like the whack-a-mole

game. I don't know if you have that game here in Australia, you know, you're just trying to like bang the one thing, it's going to show up somewhere else, it's going to show up in some other asset class. So for that reason, the alternative solution, it's not really alternative, they're complementary certainly, is to try to attack the instrument of debt itself. That the instrument that we call debt is kind of a strange instrument, actually. It's ancient. Obviously, people have been using it forever. It shows up in the Code of Hammurabi, you know, in 3000 BC. It's an old instrument. But this idea that I'm going to help you buy a house, and you have to pay me back this interest payment. And if you don't pay me back, I get the house. That's kind of a weird contract, actually. I mean, why not do a contract

which says, I'm going to go into this house with you. If house prices fall, and you have to sell the house at a loss, I'll split the difference in the loss with you. So more like equity than debt. Exactly. And so you try to have a bit more equity like structures and our view is that that might be the more resilient approach because first if you believe john acopolis then you get rid of this kind of optimist pessimistic thing right now now the pessimist the three percent guys in the previous analogy they're gonna think twice before they they lend to the top ten percent because they know if prices adjust and they're're going to take a loss. And so you might be able to prevent the bubble in the first place. And even if you don't prevent the bubble, it's

crash, we argue, would be less severe. Because now you're basically having the creditor immediately take a loss. The creditor has a lower MPC or a lower marginal propensity to consume out of the shock. And so you get a little bit of a cushion on the downside. Let's finally come to Australia. There are many people who say we shouldn't be so concerned about our very high household debt levels because there's a different distribution than what you guys had in the US during the peak of the housing bubble. There was a huge subprime component in America, as you well know. In Australia, it's a little different. About two-thirds of the household debt is concentrated in the top two income quintiles. The bottom two income quintiles only account for about 15% of the

household debt. Does that immunize us from a dangerous balance sheet recession? I think quantitatively, the argument is correct, that Australia is less prone to seeing a really severe downturn for that reason. If you believe people in the top 40%, although in my opinion, we speak of the income distribution, I think sometimes we should weight each unit by their income, in the sense that like the top 1% alone has a huge amount of the resources. So I don't necessarily think of people in the 60th to the 80th percentile of the income distribution as being really high income in some sense. I think people in that part of the distribution probably do have pretty high marginal propensities to consume, especially if they have very leveraged positions in the housing market. But I do think that you get really worried when people below the 60th percentile have huge debt positions. So I think quantitatively, that's important. Your exact question was, does that

make Australia immune? And the answer is clearly no. And that is that people, even we show actually in our own work, that it's not until you get to the top 10 percent of the income distribution where the marginal propensities to consume out of housing start to start to look close to zero so even people in the 75th percentile 70th percentile the income distribution they will react if prices fall and in fact that's already happened i mean there was a correction here i know in 2017. You did see some weakness in consumer spending. I think you guys have been very fortunate that you've had some other, you know, you have China, you're a commodity producing country. I think those things help you to have avoided a more severe downturn than what you've experienced.

I do think that a lot of the patterns in Australia up to about 2018, if you were to just show me those patterns and you told me nothing else about Australia, I would have been pretty worried. And likewise, you have had a correction, and some people are worried it could get worse. And so I do think that Australia is going down the road that has traditionally led to slowdowns in economic growth,

not necessarily downturns or severe recessions. There's, of course, idiosyncratic factors to Australia. I think the fact that you're a commodity-producing country, you realize China, what happens in China is so important. Those things are obviously variables that need to be considered. But as I often say, I oftentimes, even if I don't know much about a country, if you just show me how much has household debt gone up, how much of house prices gone up, what's happening with the current

account position, which right now you see a big reversal, that's a pretty big sign that there's a consumption correction happening. I saw in your newspapers, actually, to get a bit off on a tangent, everyone kind of celebrating the fact that there's a current account surplus. And I thought, that's not a reason to celebrate. What that tells you is that consumers have cut back their consumption of foreign goods. That's what it's telling you. So it's just another way of saying the economy is pretty weak. So in any case, those are kind of some thoughts on Australia. Well, Amir, it's been an honor speaking with you. I dare say most of our audience wouldn't have heard about you until now. And I always get a thrill when I have the opportunity

to introduce them to someone who they don't know about yet, but I think they should get to know. Oh, I appreciate that. Thank you so much for your time. Yeah. Thank you. Thank you so much for listening. I hope you enjoyed that conversation as much as I did. For links and show notes for everything we discussed hope you enjoyed that conversation as much as I did. For links and show notes, for everything we discussed, you can find those on my website.

It's www.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 find me on Twitter. My handle is at josephnwalker on the bird place. And until next time, thank you for your time. I appreciate you. Ciao.