Private Credit, Public Debt and Financial Crises
Òscar Jordà discusses the genesis of financial crises and their aftermath using the last 140 years of data on 17 industrialized economies. Recoveries after financial crises tend to be slow. Deleveraging and household finances require time to take hold.
Download the transcript (pdf, 134 kb)
Download the presentation slides (pdf, 227 kb)
This video is also available on our YouTube channel.
00:00:10 Okay, so let me give you the bottom line of my talk. The bottom line is going to be this. The reason the recovery from the Great Recession is low is that we had a financial crisis that we had not seen in the last 70 years. And financial crises are quite different from any other recession. And in particular when you have a recession preceded by rather rapid growth of credit, then
00:00:42 it takes time to deliver and it takes time to recover from that deleveraging that Mark was mentioning earlier. And when you put those elements together and think about how history can educate us about whether this recession was different from the historical experience of previous financial crises, the answer really is not really. If you had taken the data from the last 140 years spanning over 17 industrialized economies, you would’ve reached
00:01:10 the conclusion that the recovery from this recession would be pretty much on target from what we’ve seen in the data. So that’s going to be my pitch. And to do that, let me take you through a journey that’s going to take you, like I said, over a 140 years, 17 countries. This is part of a broad project of research that I have started with a Research Associate here at the bank Early Elias, Moritz Schularick, which is at the University of Bonn, and Alan Taylor who is at the University of Virginia. I’d like to construct the
00:01:40 talk into four broad areas. It’s not often that you get to see data that goes back 140 years, and it’s going to be very useful for me to present you with some pictures that I think you’ll find interesting in the sense that they, I think, reveal trends that are quite different from what you’re used to. In a sense, when you get our outlook from us you get almost the up-to-the-minute piece of news. And sometimes you lose the forest for the trees. Let me show you longer spans of data, and let me discuss really some, I would say,
00:02:12 worrying trends in the sense that these are new elements that have come into play in the last 20 to 30 years that we need to start thinking about more seriously. That will serve me to motivate a little bit why my emphasis on credit and the relationship between financial crises, credit, and how we think about this recession in particular. There’s also been another element to this recession and perhaps, to almost any
00:02:40 other recession which is how does the government play into this picture? I’m going to give you a rather different perspective on the perspective that Mark has given you. And that is to say, I’m going to return to the beginning of the crisis and think about what was the role of government before then. And how does that impact the progress of the crisis as we move on from crisis to recovery. Okay, and then I’d like to conclude by saying something
00:03:10 about what are the potential lessons that we may have gleaned from this trip through history, economic history, okay? All right, so what is the long view. Well the first thing to notice is that crises, financial crises, are back. What you have in that slide that I have in the projection is a plot for every year since 1870 up to today of the number of countries that in any given year experienced a financial crisis. And the one thing
00:03:40 that strikes you in that picture is, well, we can identify global events very easily. Obviously as you move your eyes to the right hand side of that graph you will see the big spike that came about in 2008 with the recent financial crisis. And if you move your eyes back to the beginning of the axis you’ll clearly identify 1929. But what is the other aspect of this graph that is eye-catching? Well, the other aspect of that graph is that gap in there. That gap is a
00:04:10 period of about 20 to 25 years where nobody experienced a crisis. That’s a really long span of time. And if you were to play the odds on this particular roulette, it would be very hard for you to get that long a span of time with nothing happening. So what was going on in that point in time? Well, this is the point in our economic history when the Bretton Woods system was enforced. Just to give you a little background, the Bretton Woods system
00:04:40 came about because of the Great Depression. The war, in a sense, interrupted a process that had been started at the end of the Great Depression. And it was when all the architecture of the financial system was being rewritten. So Bretton Woods in the large sense encapsulated those ideas and encapsulated also the idea that you wanted to restore international trade. You want to end policies of beggar thy
00:05:10 neighbor, which are so in vogue nowadays that the Bank of Japan has started a different program of monetary stimulus. So in that context, what are some of the ingredients of the Bretton Woods period that perhaps we should think about. Well there are some that are perhaps not as interesting. Capital controls, and fixed exchange rates, exchange rate regimes, are possibly not the first things that come to mind when we think about why that gap of nothingness was there in the previous picture. But there are some
00:05:41 other elements in the Bretton Woods System that do resonate with one of the reasons that might have prevented that outbreak of crises that we saw in that period. One of them is obviously low leverage banking, regulation coming out of the Great Depression and into Bretton Woods was much, much, much stricter than it had previously been. As a result, what you see in the next bullet point is that government securities were a part, or a much bigger part, of portfolios at banks. And just to top it off
00:06:13 basically there was – this was a period of really rapid growth in investment. So it didn’t seem like the delivery of finance at a very primitive level was a hindrance to real economic growth. I’d like to focus on those aspects of the Bretton Woods system and perhaps leave aside the first two. And perhaps if you asked any European nowadays they would tell you what a bad idea those two
00:06:38 are. Okay? Well there are some other differences that came about once the Bretton Woods system died off. And in this picture, you have something that to me is very eye catching. So this picture, displays, the ratio of bank assets to GDP, bank loans to GDP, and the monetary base to GDP. Monetary base, by which I mean, basically cash and broad money, and the deposits in checking accounts. What you see there is that if you look at the
00:07:11 ratio of money to GDP, it has been remarkably stable for a long period of time. The two things that have not been stable are the bank assets to GDP ratio, and the bank loans to GDP ratio. So around the mid-70s those two ratios started to take off very, very quickly. They’ve basically doubled in size in a very short period of time. And this is going to be one of the longtime trends that I’d like you to focus on as we move forward. It used to be in the
00:07:41 Friedman and Schwartz days that if you knew the monetary base, you had a pretty good idea of what the bank asset side was going to be. And that is true for the period that pretty much predates World War II. Since then though, that has completely broken down. As you can see in that picture that ratio of bank assets to money and bank loans to money has just continued to increase more and more over time. Okay well that is just the
00:08:13 advent of modern finance and – and so we have this period of innovation in finance. Which by the way, when we think about financial crises, it’s amazing just how often we forget that people will just find ways to get around regulation. It starts with something as simple as backdating checks, as opening Euro dollar accounts outside of the US when there’re restrictions. So people
00:08:41 will always find ways to get around regulation. But it is really unprecedented to see something so visible in the data in the last 30 to 40 years. Here’s something that nobody except you have seen, you and Early because she put this graph together, have seen before. So this is data that is part of this project I was telling you about. And what we’ve done is go through a bunch of dusty books and collected data and thought about well, okay, is all
00:09:12 lending the same, and what has happened to lending at the bank level? What you see here is a really striking evolution of what has happened to lending. This is the ratio of total loans that is going to real estate. That is the black line. And I’ve divided that black line between the portion that is going to households so to speak, and the portion that is going to commercial real estate. And what you’ll see in that graph, I think, is that the ratio that is being
00:09:41 dedicated to real estate relative to other productive activities, what I call here from nuts and bolts to bricks and mortar, is becoming larger over time. And part of that increase is – is a big jump obviously that happened before World War II, but also over time just a big jump that commercial real estate lending has experienced in the last 30 to 40 years. Okay, so this is a new development also to keep in our bag of tricks. Now, here you
00:10:10 have two interesting pictures that put these numbers in a different light. You have here a picture that belongs to data in 1929 for 17 countries, and a picture for 2007. The year before each of the two most recent global financial crises that we’ve experienced in the last 140 years, so to speak. And the one thing that strikes you is the switch in the mix of the liabilities. You look at 1928 and what
00:10:41 you see is that government liabilities were just about as important, if not more important, than the liabilities of the private sector. That completely reversed in 2007. In 2007, private liabilities were much, much, much more substantial than public liabilities. And that starts to bring me to the topic of public debt and the role that public debt plays into financial crisis, you’ll see in a minute. There are also other developments to keep in mind that are
00:11:12 perhaps somewhat unprecedented. In 1990’s we had the Asian financial crisis. One of the lessons that especially Asian countries took from that period was this, that if you have a big cushion of foreign reserves it is a lot easier to handle a run on your currency. There are anecdotes that during the financial crisis a country like South Korea, which by almost any
00:11:39 other measure was doing well economically, was faced with the situation in which Nike would send a contract for sneakers to a Korean factory and the Korean factory would be unable to secure a loan from its local bank. Even with that contract in hand worth multimillion dollars. So with that hard lesson learned, a lot of the Asian economies have decided to engage in a really outstanding accumulation of reserves. Much more than would have been predicted
00:12:11 by normal economic models. And it is not that economic theory has left us behind here, what you still see is the normal flow of private capital from developed economies into emerging economies. What you see that is new is a flow that goes the other way from official, basically the official side of the balance sheet in these developing economies basically investing in
00:12:39 our shores. So much so that they are basically undoing all the private side of this investment equation. And in fact, when you look at the numbers they’re pretty staggering. So what I have there is the period of this great reserve accumulation that has happened in the last 20 years has – basically means there’s something of the order of 10 trillion dollars officially held by emerging economies somewhere perhaps at the level of 14 trillion dollars, if you’re counting wealth sovereign funds. That is very
00:13:11 impressive and that relates to some of the statements that Chairman Bernanke made in 2005 regarding the savings glut. So let’s keep in mind when you – when you look at the next slide. This is one of the other emerging trends that is quite different from what we’ve seen previously. Here’s what this graph tells you, this graph tells you that the demographic trends are about to hit a turning point in the next five years going forward into the next 30 to 40 years. What is that big switch in the demographic
00:13:43 trend? What you see there is basically developed economies are now aging and their dependency ratios are going up. They’re going up because basically the retired population is growing larger and larger relative to the working age population. It’s a complete reversal of what has happened up to that point. Up to that point what you saw in emerging economies was really young economies with lots of children, lots of
00:14:12 dependents. That has come down over time and now they’re starting to look more like we did maybe 30 or 40 years ago. That crossing of the lines there is telling you that soon enough developed economies will not require the same level of investment to keep the same level of development. The savings that we had seen from the Boomer generation, they will start turning to these savings and more consumption. Right at the time when some of these emerging market economies have
00:14:42 reached their transitions into new demographic steady states. So what are the takeaways from looking at this long period of data? Well there’re some new emerging trends. There’s a sense in which this time is different, to paraphrase the title of Reinhart and Rogoff’s new book. So what is new that perhaps we should be paying more attention to? One is the unprecedented growth or
00:15:11 expansion of credit. So what I have there is – is a couple numbers to give you the flavor, but you’ve seen it from the graphs. So something like financial assets to GDP in 1975 around 150 percent to right before the financial crisis reaching something like 350 percent. That is a massive increase in financial liabilities. The banks’ asset mix has also changed over time as regulation after regulation has come down. So we have
00:15:40 something like 60 to 70 percent of the portfolio being made up of government securities. And now that has virtually disappeared from banks’ portfolios. There’s also been a switch to wholesale funding. So the shadow banking system that Mark was alluding to in his talk has basically exploded in the last 20 years. Unsecured lending is the game of the day. Secured loans, or deposits which are secured by the FDIC, now represent a much smaller
00:16:13 piece of the pie. And then one other aspect that is a trend change is the increasing weight of public debt on almost every country. Not just in the US, and not just because of the financial crisis, but this has been a trend that was coming at the heels of the end of Bretton Woods. Coincidentally, not because of Bretton Woods necessarily, and it has been growing up over the next 20 to 30 years and I will show you a picture of this. Some of
00:16:41 these are the broad trends that I want you to keep in mind especially the ones that relate to the importance of credit. Okay, so, how do we think about credit and financial crises? I’m going to show you a bunch of these pictures so it’s perhaps useful for me to spend a little bit of time telling you what’s going on in here. This is a picture that displays the following exercise. It looks at all the recessions that have happened in the US in the last 140
00:17:09 years, puts them into different bins: normal and financial. And then it takes the average in each case and it says, okay, let me normalize so that in year zero when the recession stars I normalize at zero and then I see what happens, in this case to real GDP per capita. Real GDP per capita falls in the first year and then starts recuperating. That’s the usual pattern that we’ve seen in recessions in the last 140 years. And by the way, the duration of
00:17:41 recessions being one year, you can go back 140 years, that hasn’t changed at all. Other things have changed, but not that. Financial recessions, financial crises are much more longer lasting. As you can see in the graph they last about two years and they take much, much longer to recover from. Now, before we start drawing too many conclusions, this is a little bit like the following experiment. You ask a bunch of people whether they have a headache or not. You then ask a bunch of people whether they
00:18:10 have taken an aspirin. And then you look at the correlation and you run the risk of concluding that aspirins cause headaches. Okay we don’t want to do that, so we have to be a little bit more sophisticated. This gives us a flavor, okay. This also gives us a flavor of what is the typical pattern of financial crisis versus normal recessions when you look at investment and when you look at inflation. So the red line again is that monstrously negative
00:18:41 line that shows you that investment really plummets in financial crises and it takes a long time to recover. It also shows you that prices tend to push the deflation boundaries rather than the inflation boundaries, okay. All right, well, this is great but we need a little bit more analysis here to get a sense of what might be going on. How can credit help us think about financial crises, is it helpful to predict financial crises, are financial crises driven by the
00:19:11 embarrassment of governments, as an early writer put it? The answer is that financial crises are incredibly hard to predict, but there is some predictive power in looking at how quickly credit grows prior to a crisis event. You can put in public debt, you can look at the external balance, none of those are going to help you think about financial crises. But the one thing that does, not majorly, but the one thing that does is the growth of credit. And if
00:19:42 you look at that problem from this perspective, I know this is a busy slide it’s meant to check your eyesight, what you’ll see is basically again the same normal path for a recession and recovery and then you see two groupings of lines. One that is closer to the normal recession, one that’s farther away. The difference between the two groupings is high versus low credit. Whether you have high or low level of deficit accumulation prior to the financial crisis
00:20:12 really doesn’t make a difference, okay. So again this is a story of credit, it’s not a story of the government. Well what about the government? These are some of the trends that I was describing earlier in regard to what has happened to public debt in the last 40 years. What I have displayed for you here is the ratio of public debt to GDP and this included state, local, and federal debt for the US, so it’s a litter higher than the numbers that you may have
00:20:40 seen. But what you’ve seen are two things. Obviously you have the big ramp up of World War II, the big pay down of that debt after World War II all the way through the middle of the 70s. And then basically the welfare state taking over in Europe and in the US and those levels of public debt to GDP rising back again well before the financial crisis hit again. So I don’t want to say that you shouldn’t care about the level of public debt to GDP
00:21:11 when thinking about recoveries from financial crises, that was not the message. And in fact, let me tell you why that’s not the message. Let me break this analysis down carefully for you. Now we’re going to start moving away from aspirins cause headaches” to now you have a scientist running a controlled experiment. Trust me, this is the closest I can get to doing that. So this is what this graph is showing you. It’s saying, okay, look at normal recessions and look at financial recessions and think about the following
00:21:41 counterfactual: everything stays the same, you control for all the variables in the economy, you put the kitchen sink in there very cleverly and then you say, okay, now let’s do the one experiment in which prior to the outbreak of the recession, or prior to the outbreak of the financial crisis, what you do is you have a higher than normal accumulation of credit. What happens? Well it doesn’t really matter whether you’re in a normal recession or you’re in a
00:22:12 financial crisis, your recession is going to be deeper, your recovery is going to be much slower, regardless. Okay so this is a graph of real GDP per capita across all 17 economies to make everything much more comparable. If you look, if you break down, and again this is an eye check, I know. If you look at lending, if you look at public debt, if you look at prices, what I’ve done is remove the dashed line for the normal recession and just focused on the
00:22:41 financial crisis. Lending goes down massively if you have a lot of credit buildup before the expansion. Public debt grows massively because of course you have to bail out a bunch of banks. So public debt is going to explode in that respect. And of course prices are going to be going down. There’s going to be deflation because there’s going to be an immense shortage of demand as a consequence, okay? Now I’m able to tell you what happens with
00:23:12 the level of public debt. Well, if you have a financial crisis and you have to bail out the private sector, mostly banks, and you’re going to basically have your level of public debt to GDP explode, the more precarious your balance sheet as a government is, the more the sovereign falls under attack by markets. That’s easy to see and you
00:23:39 can see it in Europe nowadays. Spain came into the recession with a relatively benign, actually kind of the lowest level of debt to GDP of the 17 countries that I showed you earlier. And look at Spain now. You know, the level of debt to GDP is expected to go to 100 percent. It’s not there yet, but it’s going rapidly toward that direction and look at the attacks on the sovereign debt for Spain in the last year. You know, obviously with Cyprus things were a little shaky,
00:24:13 but well before then you saw attacks on Spanish debt. What happens then if you come into the recession with a debt to GDP level that’s about 50 percent, which is about average for the sample versus an extreme which is 100 percent debt to GDP level. Again your recession becomes so much more painful. Because, you know, you start bailing out your private sector and at some point the sovereign comes into attack. The sovereign comes into attack that means ramp up the austerity. Ramp up the austerity means stop
00:24:42 the presses. Economic activity freezes, okay. And you can see that in real GDP per capita, you can see massive deflation, you can see massive freezing of lending, and of course you can see a massive increase in public debt. And then that has to be undone. Okay, so yes, public debt to GDP doesn’t predict financial crises, but it matters a lot for the purpose of thinking about what happens. Now, let me retake what Mark said about the US economy. And
00:25:11 let me retake the conclusion which I announced at the beginning of the talk. If I look at the performance of the US economy, where does the US economy rank? If I had taken the following experiment, given my analysis of 140 years of data across 17 industrialized economies, take what I knew in 2007, run the model forward, and then compare the prediction of my model with the
00:25:42 actual data. And let me do that for the US and let me do that for the UK, why? Because in the UK they had a slightly different program. They had also slightly different constraints, they had a slightly higher public debt to GDP level and we know that that makes things harder, okay? On the left hand side you have the US performance of real GDP per capita. On the right hand side you have the UK’s. What is the performance of the US? As Mark
00:26:11 anticipated the US has actually done better than I would have expected from my model. I think we did clean balance sheets much more quickly, I think we didn’t go the austeritypath quite as quickly. We waited a little longer, we had a little bit more stimulus at the beginning of the recession, and that helped us move the path quite a bit. Okay, so now we’re doing, we would be expected to be about 5 percentage points where we started in 2007 and we’re almost back to – to where we were in 2007. And compare
00:26:42 that to the UK. The UK, my model actually does a very good job predicting what would have happened for the first two years. I mean, it’s uncanny how good a prediction it is. And like Mark, I will confess to getting a prediction every now and then incorrectly, despite my abilities with the data. But look at what happened . The predicted path which said the UK ought to be now back to where it was in 2007, but it’s not,
00:27:10 it’s quite a bit lower than that. It’s almost like the mirror image of the US, okay? And I don’t want to blame austerity in the UK for this result. Like I said, they had different sets of constraints all of which I’ve tried to put in my model, but, you know things are different. So it may be austerity, may be something else. Okay, so I got one minute left to tell you what should be the takeaway. And since you have your policymaker hats on, so to speak, in your function as directors of this bank, perhaps these are some of
00:27:42 things that you should keep in mind for the longer run. And some of the things that I think are worth mentioning are this: credit plays a pivotal role in thinking about the business cycle and thinking about the severity of recessions and thinking about the likelihood of crises. So we ought to be doing a much better job at monitoring credit. And we ought to keep our eyes on the ball much more thoroughly perhaps than we have, especially given the
00:28:12 very rapid innovations in finance that happened in the last 25 to 30 years. It just makes the job of policymaking that much more complicated but that much more important at the same time. And what about excess public debt? We had a ramp up of public debt in the last 30 to 40 years that is basically a trend changer from what had happened World War II. Well there, too, I think we have to keep our eye on the ball and perhaps not be
00:28:42 too quick to do this, but perhaps at some point we need to get a handle on public levels of debt relative to GDP. But there are two other stories that I don’t really know how to qualify. And those are what has happened in emerging economies and what has happened to demographic transitions. And those suggest to me that the savings glut that we experienced that was mentioned by Chairman Bernanke in 2005 may be coming to
00:29:11 an end at some point in the not too distant future. And those are really long and strong forces that we need to think about. And with that, I don’t know if it’s a high note or a low note, but I’ll leave it at that.
[End of Recording]
About the Speaker
Òscar Jordà is a Research Advisor at the Federal Reserve Bank of San Francisco, which he joined in 2011. His research at the bank covers time series analysis and applied macroeconomics broadly defined. Examples of his scholarly articles and ongoing research projects include research on the role of private credit and public debt in financial crises, methods to date business cycles, evaluating the effectiveness of monetary policy in downturns, and better ways to report forecast uncertainty.
Mr. Jordà received a Ph.D. in economics from the University of California, San Diego, and holds a B.S. in economics from the Universidad Complutense de Madrid. Before joining the bank, Mr. Jordà was Professor of Economics at the University of California, Davis. See his research page.
Elias, Early and Òscar Jordà. 2013. "Crises Before and After the Creation of the Fed." FRBSF Economic Letter 2013-13 (May 6).
Jordà, Òscar. 2013. "Will the Unemployment Rate Stall in 2013?" Economics in Person Series FRBSF.
Berge, Travis J., Early Elias, and Òscar Jordà. 2013. "Future Recession Risks: An Update." FRBSF Economic Letter 2011-35 (November 14).
Jordà, Òscar, Moritz Schularick, and Alan M. Taylor. 2012. "When Credit Bites Back: Leverage, Business Cycles and Crises." FRB San Francisco Working Paper 2011-27.