You are here

December 16, 2020

Long Run Effects of Quantitative Easing


Rodney Ramcharan
Rodney Ramcharan | Professor of Finance and Business Economics, Marshall School of Business

Rodney Ramcharan joins Richard K. Green to look back at 2009 and the quantitative easing used to inject money into the US economy during the financial crisis. Ramcharan shows that the effects of government intervention in the economy can last a long time, up to six years, with refinance activity providing a key indicator for a business’s future health. Green and Ramcharan discuss how the data gathered since 2009 could inform monetary policy as the effects of the pandemic continue, as well as the varying viewpoints economists have had over the years about the debt ceiling.

Listen via podcast



- My name is Richard Green and I am Director of the USC Lusk Center for Real Estate. And it's my pleasure to welcome you back to Lusk Perspectives. We have a terrific guest with us here today, Rodney Ramcharan of the Department of Finance and Business Economics at the Marshall School of Business here at USC. Many of you already know Rodney quite well for the splendid presentations he's done while he has been associated with the Lust Center for Real Estate. He joined us from the Federal Reserve Board of Governors, where he was instrumental in putting into place for risk management profiles in the aftermath of the great financial crisis. And I think it is fair to say that this time around, we learned quite a lot from the last time around. And so at least in the financial sector, things are going far more smoothly than they did 11, 12 years ago, so we have Rodney in part to thank for that. Thank you, Rodney. And Rodney is gonna be speaking of an issue that I think is of particular interest to the real estate business, not gonna let him give the title, but it's about how do big changes in monetary policy flow through to business decisions and consumption decisions? How much do they matter immediately to the trajectory of the economy, and in particularly through the real estate channel. So with that, Rodney, thanks for coming back and doing this, take it away.

- Thank you, it's my pleasure to be back. It's a pleasure to have the chance to talk about these questions so that the focus of the talk is looking at the effective QE. And there's been a number of papers and folks have talked about this extensively and I'm guessing that you can see my screen and so forth and that's all fine. So what I'm gonna do is I'm gonna look at the effects of QE, but look at it long-term so let me give you some background.

- [Man] Rodney? [Rodney] -Yes.

- Just one second, I think we need to reshare your screen if you wish, if you, I can see this.

- Reshare the screen okay. And..

- [Man] Big green button at the bottom.

- Yep, let's try that again. And can you see that?

- [Man] Yes.

- Okay, let me now do the slideshow and hopefully this works. If not then, 'cause I think I already have the slides and can you see this? Your screen screen-sharing is pause

- Yes.

- So let me resume sharing.

- And you could see that, right?

- [Man] Yes.

- And you could still see my slideshow.

- [Man] There we go.

- Fantastic it works. Okay, so here's a little bit of background, after the crisis, what the Central Banks all over the world learned is that when you hit the lower bound at zero on the short end of the yield curve, that is the interest rate that the Central Banks traditionally had control over, people then said, well, look, most folks tend to borrow the longer end, and so Central Banks really beginning with the Japanese case in the nineties, but at least in the case of the Fed in 2008 said, let's try to manage the yield curve, not only in the short-term, the traditional part of what the Central Bank does, but now that's focused on the long end. And so you have a picture here of QE1, the QE1 announcement, where the Fed, I believe November of 2008 said that we've hit zero at the federal funds rate, we'd like to bring down the long-term 10 year rate. So what the Fed began doing is began buying long dated treasuries, and long dated MBS as an attempt to sort of bring down the long rate, that's the rate at which most folks tend to borrow. And so the picture that you're looking at is the difference between the federal funds rate or excuse me, I phrased it poorly, that the difference between the 10 year treasury and the federal funds rate. And what my mouse is going to do is gonna focus on the red line, which is the QE1 announcement. And what you see from that is all through the throws of the crisis, you see that the 10 year rate really stays quite hard. And that was a real problem at the Fed because people were very concerned. And so once QE1 gets announced, you see that gap there between the 10 year rate and the federal funds rate begins to drop quite quite sharp. And so the way that Central Bankers think about this is they think that if they're able to manipulate the long end of the yield curve, there are a couple of mechanisms that they have in mind. They're able to provide liquidity to banks, and then that can induce banks to lend more, so that's one particular channel. The other channel is all of you will know you all investors is that if you lower the long-term rate for safe assets, you're gonna induce investors to rebalance their portfolios to risky assets. And that's what the Fed wants. In addition, because you've lowered the risk-free long-term discount factor, you're gonna boost asset prices as well and we've seen that now. And then the fourth big effect that tends to work at least in the United States is the mortgage market. So in this talk, I'll give you a little bit of, well, not give you the background, but show you some examples as to how QE1 might've worked in the housing market. So the one thing I want you to take away from this exercise is that QE1 did work. So the picture you're looking at is the average mortgage rate in the United States. This is a daily time series of that rate. And what you see is the red line, again, the QE1 announcement and mortgage rates plummet. They drop by about a percentage point almost on the day after. So it was a fairly effective strategy. And then what you see here, this picture is a little bit busy, but the blue line is the refi credit volume. So this is using daily data, in this case, it's aggregated up to the weekly level and the blue line is the volume of mortgage credit that's refied in that week. And the red line is the volume of mortgage credit that's originated for the first time that week. What you see is the red line again, is QE1, that's the arrow there that you see and refi jump sharply. So once the Fed is able to bring down longer term rates, lots of people in the mortgage, were able to capitalize on the lower rate by going to the bank and getting this new rate. And so that's a great thing because folks have shown that in some cases, the monthly savings on the refi were about $1,000 a month in some cases and so people could take that savings and go spend it, and that's a wonderful thing when there's a downturn. Another way of looking at the impact of QE1 and the refi growth is this is a little bit busy again, but the intuition is that if you sort of adjust for the trend in refi growth, then these dotted lines with these dots, the part of refi growth, that's detrending. The red line, the vertical red line is QE1. What you see is the week, so this is QE1 here. The week after QE1 is implemented, there's a dramatic national increase in refi activity. It's dramatic. It's about a 50 percentage point jump in what folks were able to go and do. So if you're sitting at the Central Bank in 2009, if you're an investor, this all looks wonderful. Now the challenge with this exercise is that not everybody could benefit. So the blue would be the areas where refi activity was weakest and the red would be the areas where activity was strongest. So let's focus on our home state. We see that the redish areas are all along the coast, okay. And the blue areas are all in the interior. So the impediment, the friction on the refi channel is if you're underwater, if you had gotten a mortgage in 2006, you had bought at the peak and you're underwater, and now the Fed has come and lowered rates and you go to the bank and say, I'd like to refi, the bank says, well, look, your equity is negative. It's not gonna happen. And so only people who didn't get really impacted by the housing shop to begin with were the ones who could benefit. So, that raises some interesting questions. Let me just step back just a second and give you the channel that people had in mind. So the channel again is, I come along with the Central Bank, I lower the long-term rate. People then refi, the monthly savings if you get a lower rate, in many cases, the rates fell by a percentage point with two or three, the monthly savings are enormous, $1,000, $800 a month in savings. The idea being that people will then use these savings to go spend, go buy cars, go out to restaurants, maybe go buy another home. And by going and spending, it would help businesses. This is a picture, this is very, very micro data from a data provider with about 20... So it's the universe of businesses in the United States is about 25 million businesses across the United States. And this aggregated up... So I know the exact... I know the location of each of the businesses and that aggregated up to the county level and the reds and looking at the growth in businesses across the United States from 2009 to 2015 Q4.

- Excuse me, Rodney are these establishments or are these firms?

- These are a star... So these are established.

- And could you explain to the audience, the distinction, 'cause it's kind of an important one here, I think.

- Yeah, actually I should take that statement back. I think the way that the company gathers the data, I think these are businesses in the sense that if you have a credit score, then this would be the... So if you have a... If you have a coffee shop and you use credit, you then get a credit score, and they keep track of where you are and the amount of credit that you've used. So it's not gonna be at the establishment is gonna be at the business level. And so it's not establishment, it's the incorporated business. And so the takeaway from the picture is that on average, the U.S. lost a great deal of businesses in the last, well from 2009 to 2015 Q4, but the decline tended to map into the areas where people could'nt refi. And so it raises the... I'll show you another picture of this. So this is a picture that brings this home quite clearly I think. So the left-hand sides of scatterplot, the left-hand side is the change in the total number of businesses in the county from 2009 Q4 to 2015 Q4. And on the X axis, that is the axis at the bottom, it's the volume of mortgage credit that was refied during QE1. And what you see from the pictures of those areas, those counties where people were able to refi, those counties also tended to have the biggest or the largest business growth in the next six to seven years. And that raises the possibility that the effects of QE1 may have been highly uneven across the country. So there's a number of econometric, and this is just a webinar that's intended to raise questions, not to solve questions, I'll save you that. So there's a number we econometric metric things that you need to do to convince yourself that this is accurate. And I'll just give you a flavor of the intuition and then I'll move on, okay. So the main concern, let me go back to the previous slide, the main concern you might have with that slide is where people could refi, those are the areas where incomes are higher and so forth, and so on, so what we're picking up isn't really the effect of QE refinancing on business activity, we're simply picking up the effect of income or race or the demographic structure of the county on businesses. And so what I can do with the high-frequency data I have on the mortgage market, it is like an extract, the part of refi activity that comes immediately after QE1 in the week or two after. Now, by focusing on that narrow window, right after the Fed makes the QE1 news, I can be reasonably certain that the in pulse and refi activity is coming directly from the policy change and not from something else, okay. And that allows me to then cut the refi activity into a very fine sliver that's attributable to QE1. And I call that fine sliver a shock, okay. It's the part of refi activity that people didn't expect, but they engaged in once QE1 occurred. And this is just to give you a figure, just to give you a picture of what that shot looks like. What you could see is that there's no trend in refi activity until QE1 takes place. Then the rectangle goals measure the size of that shock. And you could see the after QE1 gets announced, the size of those shocks become quite, quite large, okay. So then all I'm gonna show you is I'm gonna look at, and I'm sparing you all the econometric details behind this. I just want you to take away from this exercise, the idea that those counties, so this is the QE1 shock that my mouse is on. These are the shocks that come after QE1, so week one, week two and week three. And I'm trying to relate the growth in business activity inside the county to those initial QE1 shocks. And what you find... And so this rectangle here is the size of the estimate and the bars there are above the zero access so that tells you that effects of the QE1 shock up to three weeks after the event has this statistically significant effect on business growth inside the county six years, Oh, I'll say it again. I have focused on a narrow sliver three weeks after QE1, and I can then predict business activity inside that county six years up. So that begins to tell me that the effects of these shocks were presumably large and persistent. Another way to look at the data is to look at the... So whenever you see businesses changes, excuse me, I said that badly, whenever you see businesses changing, it's a mix of entrance and exits, right? Some folks are entering the county for the first time and some are exiting. And so you could look at the ratio of that to get the measure of the business expansion of the extensive margin. And what do you again see, is that a couple of weeks after QE1, those counties that reified significantly more tended to have relatively more entrance compared to exits six years later. And I just have a few more breaths and then I'll stop. This is looking at the average credit score of businesses inside the county. What you again see is that looking ahead, five, six years out and this... And I'll talk about why this is pertinent for the events that we're facing now, you could predict how well those businesses are going to be doing five, six years, hence by looking at the refi activity today. This is a few more... So this is using about 20 million businesses and asking what's the probability that the business that was in existence in 2009, what's the probability that the business exited by 2015 Q4. And here, I'm looking at small businesses and this is the QE1 shock, okay. So if you have a positive QE1 shock, any business or any small business inside that county is significantly less likely to exit because it's negative. Likewise, if you look at five to nine employees, you see a similar pattern as well. And now you might think, well, not every business in every county depends on that county, right? If you have a... If you have a conglomerate that sells stuff all over the country, economic activity inside the county is less relevant for a business activity for that kind of business. Well, one businesses or one set of businesses that depend on economic activity inside the county would be like retail food establishments. So these are the definition of non tradables, McDonald's or a coffee shop does not really sell the coffee much beyond five, six months. So here you see looking at the probability that the retail food establishment exits you see again, that the effects can be explained by QE1 economic activity, the same thing in the construction space as well. Clearly, construction is non tradable, and you could predict whether your construction company survives as a function of refi activity in QE1, the week of QE1. So let me just give you the big picture and then I'll open it up, are these tell you what I think is some of the questions? I think a couple of takeaways is that the housing market, at least in United States, in the EU, it's different than in the U.S, the housing market is a pretty important channel for what the Fed does in the traditional way of cutting short term rates, but also in the new way of buying bonds and trying to bring down the longer term rates. What this evidence just shown you and the sketches, I think is a good way to think about it. What they've shown you is that these events, so these effects seem to last a pretty long time. So again, I'm looking at the effects of QE1 in 2009, and I'm explaining business activity six years later, okay. And then finding those areas that we're able to benefit more where people have positive equity businesses, five, six years later, we're better off and would enter those kinds of counties. Now, the flip side, if this is not everybody benefits. So one implication of this is, is unconventional policy by the Fed making inequality worse, right? Because what we're seeing is that some counties have people with positive equity who benefit greatly when the Fed cuts rates. Some counties have people who are not so lucky with negative equity, and when the Fed cuts rates, those folks don't benefit and the businesses there don't benefit as well. It raises the question then would a different set of policies have made the country better off, would fiscal policy have been the better way to deal with the crisis? And it also raises I think, a somewhat of a deeper question that if the Central Bank is pushing rates down and is benefiting a certain sliver of the population and takes the pressure off of Congress from acting, then the Central Bank could potentially be making inequality much worse. And then the greater inequality means that the political dysfunction gets much more difficult. And so there's this case that maybe given the side effects of what the Fed is doing, it might be better to lay off the gas and let Congress act. So I'm gonna stop there. And if you have thoughts and questions, I'm happy to tackle them.

- So, Rodney, thank you very much. And I think that I'm gonna do things a little unusually. As I'd like to make a couple of comments and if you would leave the last light up, I would appreciate it for a few minutes. Usually I just ask the questions and please respond to my comments. So maybe there'll be comments/questions. And then as the group knows, if you have any questions, feel free to type them into the Q&A box. And we have an intimate group today. I'm sure I will be able to get to everybody's questions. And first of all, I have to say so Rodney's , made a specialty out of doing this kind of research, and it's always very thoughtful, it's wonderfully straightforward, which is something economists are often not that good at being and very useful. He did a very highly cited paper a few years ago on a similar theme, which was the impact of adjustable rate mortgages on consumer behavior. And what he found was the adjustable rate channel is a nice experimental channel because people don't have to pick, they don't have to make a decision. They automatically see their cost of funds go down. And he and his colleagues found such things like people did in fact, have a greater tendency to buy cars when the interest rate on their adjustable rate mortgage went down. This is looking at different channels. And a couple of things occur to me, first of all, in 2012, a policy came into place called HARP 2, HARP 2.0, and it essentially changed the rules of refinancing quite traumatically. Before HARP came along, if you were upside down on your house, you basically couldn't get a refinance. And then HARP 1.01 said, well, if you were less than 15% upside down, you could get a refinance, but that really wasn't enough to be effective. And people still needed to get appraisals, 2.0, the appraisal requirement went away and you had this enormous wave of refinances as a result of HARP in 2012, 2013. And so a comment, because I think there's something to be learned from that right now, too, which is now that the issue is not people not having equity in their houses, they have a lot of equity in their houses, that the issue being, if you don't have a salary job, you can't get a refinance at this time when it's a really great opportunity to take advantage of a very low interest rate. And of course there are many, many people who at the moment are not in salary jobs, so they're locked out of this refinance market. So that's a very long prelude, I mean it's common. And also a question is how much do you think your 2015 results are being perhaps understated because the 2012 policy kind of got in the way and so places that couldn't easily refinance 2009, suddenly it became much easier. The point is if it hadn't been for that HARP 2.0, might you see a more dramatic outcome, even still in 2015 based on the 2009 phenomena.

- Yeah, I think that's a great question. I think I can answer that by saying, I don't know the answer, but it's a question I can answer in the sense that like QE1, I could sort of get the data on where the HARP refi occurred and then put those in the specifications as well and see whether, what the effects of those are and if the independent effects from QE1 as well. So, yes, so I think it's worthwhile to extend this kind of set up and take a look at HARP as well.

- And so along those lines, I mean the issue of inequality, again, being something that you and I are worried about. I mean, it's all kind of, again, we see it manifesting itself now, people not being able to take advantage of the very low rates if they don't have a job. The other thing is, if you look at the benefit of buying... The financial benefit of buying a house right now, relative to renting, it's, I think no exaggeration to say, and it's at its strongest point in at least 10 years, and maybe 20. If you look at what we call the user cost of owner housing, relative to renter housing in large American cities, almost everywhere, now you're better off financially owning than renting so long as you're gonna stay in your place for at least five years. But, the other side of that is of course, investors in mortgages are taking a loss with these refinances. So they had its that were paying four and 5%, and they're gonna have to trade them in for instruments that pay two and 3%. To what extent, do you have any sense to what extent that this inequality issue? Is it sort of, is it minor in the overall scheme of things? I mean if we're looking at a couple of percent on $7 trillion, potentially a mortgage backed securities outstanding, so that's like 140 billion. That's that's real money, but compared to everything else you're looking at is it kind of small beer?

- Yeah, I don't know the answer there as well. I mean, I think, mmh. I think the concern that I have is if COVID continues, so there's been a stay on the foreclosures now, if COVID continues and people could no longer stay in their homes or can't buy homes, I think that's going to further widen the gap between the people who can do well and the people that don't. And I think the concern I have is if the Fed continues to keep long-term rates low, which may or may not be a good thing, it creates the appearance that the economy is booming. When in fact you're talking about significant swats of people that can access credit and it takes the pressure off of Congress to act on the fiscal side. And I think that's sort of the concern I have is, you know, people look at the stock market, they say things are going great. I'm not gonna put pressure on my congressmen to act because that helps a certain group of people, but there's a whole bunch of people that don't own stock they're trying to own homes that can't get access.

- Yeah.

- And if Congress doesn't step in and sort of try to find ways to get credit to them, but if they leave it up to the Central Bank, those tools are pretty blocked and they don't always reach the people that that should get it, and so that's a concern. And I think longer term, if that continues, if that widening between the people that own the assets and the people that don't, if that continues for a long time, I think that makes it much more difficult to act.

- And of course, even then, you know, again, the difference between a mean, and a median is key here, even in the stock market. It's not doing quite as well as the top line number suggests. About a month ago, I looked up the total value of the S&P 500 is about $27 trillion. Of that 6 billion is in four companies. It's in Amazon, it's in Apple, it's in Google and its Microsoft. Hey, in an environment like this, those sorts of companies are going to do very, very well. Whereas this audience will tell you, the REIT market for example, has not performed as well, except for cell tower REITs, which have performed exceptionally well. And so, yeah, I mean, there's a more generic problem but I think people focusing on averages too much when if you have a skewed distribution averages can make things look a whole lot better than they actually are. So, I, you know, I hate to do this Rodney and push you outside of your comfort zone, but maybe I will a little bit, cause you have so much experience with Central Banks, is there is discussion now about the importance of the Fed supporting state local governments.

- Yeah.

- And one of the mechanisms they could use is to buy state local bonds at favorable prices in a way they've been doing with other parts of the asset market. Do you think that might be a productive thing for the Federal Reserve to do?

- So I wanna applaud you for the question. I think it's a brilliant question. And it certainly provides a way out that now my understanding of public finance is reasonably weak. And I think some local governments face a bankruptcy constraint where they can't run a deficit for too long or something like that. If that is somehow waived away, and you know, and the Fed is able to allow local governments to borrow cheaply for a longer term, then I think that could be the fiscal stimulus that these towns might need. So just to put it differently, if Congress is unable to act, if the central government finds itself unable to agree on a fiscal stimulus, if you take the stuff I've showed you at face value, whether it says that large swats of the population, many counties are gonna be unable to... Are gonna have people who are unable to access credit. If as you pointed out, if the Fed can then step in and sort of be the buyer of public debt at the local level and allow those guys to expand, to continue with education, with infrastructure and to do housing at the local level or create programs at the local level, I think that could be a wonderful thing.

- So let me just make sure that our group doesn't have any questions before I move on. So if you were, I mean, you've sort of said this already about things we might've done differently back in 2009. How much are you worried? So I know a lot of people in our group asked me about how much debt can the U.S. economy really sustain? And of course the level of debt to GDP is at its highest level since world war II right now, is this in a lot of what we're talking about here involves, you know, if the Fed is buying stuff in order to keep interest rates low, that's basically it is increasing its balance sheet. Is there a limit clearly existentially there is, but are we a long way away from that limit? Should we start worrying about what that limit is?

- Yeah, the last I looked at this, I think a U.S. debt to GDP is 120%. It's increased by 60%, which points in the last 12 years or something, Japanese debt to GDP is converging on 300%. The big advantage that we have is we're borrowing in our own currency. And so, you know, you're right, in principle there's a maximum, we don't know what that maximum is. If you look at the Japanese case they're nearly triple where we are and their interest rates still are zero negative.

- Yeah. So now I mean the other side, so the statistic I always said is that may be true if you look at our debt to GDP ratio, but our debt service to GDP ratio is right now very low by historical standards.

- Yeah, I mean, and that's because rates have been kept low.

- Yeah.

- That's right, that's why. I think it's the kind of thing... The way to answer this is to give you a contingent answer that you could certainly borrow and keep doing it until the first sign of prices going. If there is inflation, then I don't know what happens at that point, right? Yeah.

- So I'm trying to remember like your a younger guy than I am is when you were in graduate school, were you learning kid let him Prescott stuff about credible commitment and all that kind of stuff or was that already passed say by the time you were.

- I was learning it, and I'm gonna make a comment, but I'm gonna hold myself back because I think you have a particular direction that you're going, so I'm gonna let you go there.

- No, no, no, no, no, no. I wanna hear what you have to say. You're already started it.

- Well, I mean, you know, all those papers in the eighties and nineties were talking about fiscal discipline and that there's a maximum, and the reality is that we don't know what that maximum is. And I think if you're an economist, at least if you're an academic, you have to be humble. A lot of the theories, and I'm gonna say this, a lot of the theories that I grew up with, that they taught me in Graduate School that we currently teach. They don't explain the world that well.

- No they do not.

- It's not that they're wrong, but they're very incomplete.

- Well, which is almost the same as . There's a plot, I like to do. I plot real interest rates against the deficit to GDP ratio and they go in the wrong... The correlation goes in the wrong direction. Now I'm not ascribing any cause mind you, I'm not saying that higher deficits bring about lower interest rates, but it sure does make one wonder about a lot of the stuff that we were taught. For me it was about 35 years ago something like that. And you know, as Keynes who I rely on more and more to think about macro economic issues, even though he has been dead for 74 years now, is when the facts change, I change my mind. And the stuff that I think knew to be "knew to be true", has been pretty disappointing in terms of its usefulness. Which means we were still sort of groping around in the dark, I think.

- Yeah, yeah, no, no, I think that's right. I mean, I was at the IMF for 10 years and you know, it stands for it's mostly fiscal and I traveled all over the world telling countries that they should cut spending. And that's true if you're borrowing in dollars.

- Right. The currency can depreciate all other issue that's right.

- Yeah. But we used to also tell the U.S. that you need to bring the deficit down. I remember this in 2000 and 2003 and four, we would tell the Bush administration because they had cut taxes at that point, you need to bring the deficit down and need to be cautious. And that's just not true. And this, not it's true, the arguments that we used to make, they don't seem to be borne out with the data. The same thing applies in the EU, that you know, the EU operates in a 3% deficit maximum across the member states and they've been crossing that and you know, it doesn't seem to be necessarily a problem for the Euro yet.

- So, well, Rodney, thank you for taking some time out with us today, really, as always appreciate your insights, fascinating stuff. Again, I think we can learn now I would, as we've talked about, like to see a tarp, excuse me, HARP through-

- HARP, yes.

- So that people can refinance their mortgages regardless of their income situation. One thing I think, Oh, I wanted to ask you about the other real benefit of this is we know and so we have learned a lot from the last crisis is reducing people's mortgage payments is a really effective way of reducing halt and that just puts a lot less pressure on the entire system. And so beyond the sort of fiscal stimulus of paying a lower rate, just the fact that you know, that so many fewer people are gonna fail surely has a positive impact on the broader economy.

- Yeah, yeah, so it's not an equity story, it's a debt service. Yeah, exactly right, exactly right, yeah.

- So Rodney Ramcharan, again, thank you so much for joining us. Great seeing you and-

- Thank you Richard.

- Keep yourself safe.

- You as well, you as well. It was a pleasure.

- Thank you.