This 18-minute podcast was released July 15, 2021. (This podcast was recorded June 17, 2021.)
“[W]hile you see a large increase in foreclosures in the 2008 financial crisis, you don’t see a similar increase in foreclosure during the COVID-19 recession. Actually, the total number of foreclosures declined during the COVID-19 recession,” says Juan Sanchez, a vice president and economist at the Federal Reserve Bank of St. Louis. He talks with Kristie Engemann, economic content coordinator, about trends in credit card debt, mortgage debt and foreclosures during the pandemic and how they compare with those of the financial crisis.
Kristie Engemann: Welcome to Timely Topics, a podcast series by the St. Louis Fed. I’m Kristie Engemann, your host for this episode, and with me today is Juan Sanchez, an assistant vice president and economist at the Federal Reserve Bank of St. Louis. Hi, Juan, thanks for joining me.
Juan Sanchez: Hi.
Engemann: Today we’ll be talking about your research on consumer debt during the pandemic. And we’ll mostly focus on credit card debt and housing debt, but I’d like to start with a general question for you. Could you give us a broad overview of how consumer debt patterns have evolved during the pandemic and did they surprise you at all?
Sanchez: Well, early in the pandemic we saw that people decided to buy more goods related to housing and more housing. So, that had an impact on housing debt that surprisingly we were coming from the 2008 big mortgage crisis so that market was actually pretty strong during the pandemic. Non-housing debt, on the other hand, did decline at the beginning, so we’ll talk more about credit card, but it has like a big decline at the beginning but then it started increasing quickly. So, I would say that early in the pandemic a lot of people were out of jobs, so we were expecting an impact on repayment of debt. But as we will discuss, we saw very little of that, so I think that was the more surprising component.
Engemann: So, now I’d like to ask a few questions about credit card debt specifically, which you wrote about in an On the Economy blog post from May titled Credit Card Deleveraging during the COVID-19 Pandemic. Could you talk about overall changes in credit card debt during that period?
Sanchez: Yes, just to put it in perspective, prior to the COVID-19 pandemic aggregate credit card debt had increased since 2013—from around $675 billion to $930 billion. So, during that period it was increasing steadily. So, however, in the third quarter of 2020 this debt dropped by about 13%. So, this is more or less half of the increase that happened over a seven-year period decline in a quarter, I would say. So that was like a very large deleveraging in credit card. So, in the article that you mentioned, what we do is to use a methodology that uses debt creation and debt destruction to try to understand where the change in debt comes from.
To explain it, let me just say that although at the aggregate, debt may increase or decrease, at the micro level, even if the aggregate, say, is increasing, at the micro level there is going to be households that are increasing debt and households that are decreasing debt. So, it may be, for instance, that these two households cancel out and at the aggregate, it doesn’t change. But when there is more creation than destruction, the aggregate debt is going to increase, and the opposite is true when the aggregate debt is declining.
So, what we look here is during the COVID period, what was happening with creation and destruction? So, again, creation is the group of families that are increasing their debt. Destruction is the group of families or households that are decreasing their debt. What we saw during this period is that both creation and destruction contributed to the deleveraging. So, creation, because there was much less creation than during the previous period and destruction because there was more destruction. So, this means that more people were repaying debt during this period. So, I think that last factor is important to differentiate this episode from previous episodes.
Engemann: Okay so, speaking of a previous episode, how do these trends differ from those of the Great Recession, and what explains these differences?
Sanchez: Yeah, so, during 2009 – 2011 there was also a deleveraging in credit card debt, but at first it was not that sharp, it was during several quarters. But also, the main difference is that there was not that increase in debt destruction. So, part of the large share of the decline in debt during the 2008 financial crisis had to do with the creation of debt. So, people were taking less debt so in that case, you can also think that there was less supply of debt, so the household could maybe not access debt during this period. But we didn’t see what we saw now, an increase in debt destruction.
So, a potential reason for the spike in destruction is that during the COVID-19 recession there were more stimulus programs by the government. For instance, the census has a report in which they ask households about what they used their stimulus money and around 16% of the recipients used this money to pay down debt. So, it may be that that’s increasing destruction in part, it’s coming from these government programs. And that, I think, is going to be some clear difference with previous episodes.
Engemann: With people taking advantage of stimulus payments to pay down debt and increase their capacity to take on new debt, could this help in the economy’s recovery in the coming months? Or do you think that the behavior reflects a level of caution that may continue until COVID-19 is under control?
Sanchez: So, in part I guess it will depend on the evolution of the pandemic. So, if things continue to improve as we all expect—we do expect a strong recovery on households’ demand or household expenditure. We are already seeing this. The other thing that some households used this money for was to increase savings. With more savings and less debt, we do expect this factor to contribute to increasing demand in the coming months.
Engemann: OK, thank you. And you and your colleagues at the Richmond Fed and the Kansas City Fed have also researched extensively on the topic of communities with high levels of financial distress. Will the deleveraging of credit card debt help reduce some of the levels of financial distress? Or are these households often those who are unable to participate in deleveraging because of their financial needs?
Sanchez: Well, yeah, so with my co-authors Kartik and Jose, we have worked on financial distress at the ZIP code level. That’s a good level of aggregation because it’s a relatively small region. So, what we do there is to define financial distress in different ways and to show that the patterns are similar for the different measures. So, these definitions are 1) households are in financial distress if they are 30 days late making credit card payments, or 2) households are in financial distress if they have used most of their available credit in their credit cards. We compute that at the ZIP code level, and then we can merge that data set with our data set to look at different factors.
So, for instance, what we did during the COVID-19 pandemic is to merge this with data on income and try to see if these regions have a larger decline in income. So, what we did see, and we have a working paper where we look at this, regions with higher financial distress have suffered larger income losses during the pandemic.
However, this financial distress that we were talking about is financial distress prior to the COVID-19 pandemic. We haven’t seen changes after the beginning of the pandemic in financial distress. And we think that this is largely in part due to government programs, not just to the programs we were mentioning before. But also, to the programs that facilitated households to maybe not having to repay debt for some months, some forbearance programs. We actually have seen some decline in delinquency rates, this is the way we’ll refer to the percentage of households being late making payments. So, we have actually seen a slight decline in delinquency rates. These two factors have contributed. So, they did suffer larger income losses, but we haven’t seen increasing financial distress in these regions with more financial distress.
Engemann: You touched on this a little bit in your previous response, but now I’d like to turn to some questions on mortgage debt and foreclosures. You have a recent Regional Economist article in which you discuss the drop in foreclosures during the COVID-19 pandemic even though on-time mortgage payments declined. Could you talk about those two findings and also about how delinquency rates and foreclosure rates during the pandemic compare to those of the financial crisis?
Sanchez: Yeah. So, we looked at on-time payments using a loan level data set. So, this means a data set that the servicers of the mortgages provide. So, we don’t have household information, but each observation is a mortgage. So, we were able to see the share of those mortgages that are on-time payments. And this is coming from what I was saying before: When we just look at delinquency rate from the side of the household, say from data sets on credit reports of individuals, we don’t see an increase in delinquency rate. But when we look at it from the sides of the mortgages, we see that there is an increase in the share of mortgages that are not being repaid.
So, just to give you some numbers, from 2015 to the beginning of 2020 the portion of loans in our sample being paid on-time, so what we would call current, they move around 93-92% of the mortgages. So, in February of 2020 it was 93% of the loans that were being repaid. So, in in May of 2020 the number dropped to 88%, OK? And then, it moved around 88-89% for the rest of the year. This is a significant change, so there were mortgages that were not being paid.
So, what happened then is that normally when mortgages enter delinquency—they are not paid for 30 days, they are not paid for 60 days, they are not paid for 90 days, some of them get back to repayment. What happened before is that these loans, eventually, go to foreclosure. OK? However, what happened this time, if we take the loans that were current in 2020 and then enter the 90 days delinquent in June, so like 90 days of non-payment. So, when we look at them, we see that many of these loans remain 90 days late during the rest of the year. And many of them, actually 40% of them, were able to come back to the current. But we see almost none of these loans go into foreclosure. So, that’s going to be, again, something very particular about the pandemic compared with previous episodes.
Engemann: So, you touched on this a little bit already, but you also looked at what happened to loans that exited the 90 plus days late status in 2009. And you found that the two periods, 2009 versus 2020, were quite different. Could you tell us a little bit more about those differences?
Sanchez: Around the time of the financial crisis beginning of ’07 to the end of ’09, loans in our sample with on-time payment dropped from 93% to 84%. So, this is a larger decline in on-time payments than we observe in the COVID recession, but it took more years.
But in that sense, both episodes are similar. So, I would say the biggest difference is what happened with the transitions for the loans that were already 90 days delinquent. So, what we did is we took the same month, February of 2009, instead of February of 2020, and then we look at the loans that were 90 days late by June of 2009, again, as we did in June of 2020 for the COVID pandemic, and we look at what happened to the loans. The big difference here, between these two episodes is that in 2009 only 10% of these loans became current. Remember that I mentioned this number was 40% for the COVID pandemic. While 35% of these loans transitioned to foreclosure.
So, if you remember, I just said that almost no loans transitioned to foreclosure of these loans during the COVID pandemic. So, here it’s 35%, so that’s the large difference. So, again, to say it clearly, if you look at loans that were 90 days late in June, if you do it for ’09, by the end of the year 35% went to foreclosure. If you do it for 2020, for the COVID recession, almost none of these loans went to foreclosures. So, that’s the big difference between these two episodes, but in both of them you see a decline in on-time payments, a bit larger in the 2007 to 2009 period, but the main difference is that while you see a large increase in foreclosures in the 2008 financial crisis, you don’t see a similar increase in foreclosure during the COVID-19 recession. Actually, the total number of foreclosures declined during the COVID-19 recession.
Engemann: Oh, that’s an interesting fact. So, what explains the differences between 2009 and 2020? Were there lessons learned from the financial crisis that were put into place this time?
Sanchez: Yes, so I think that was a big difference. There was a federal assistance to homeowners, the CARES Act actually introduced a provision to suspend foreclosures and offer very easy access to forbearance options for homeowners with federal backed mortgages. And also, encouraged some private lenders to apply similar programs.
I remember during 2008, I was working at the Richmond Fed and we were trying to understand—there were some programs that were put in place to try to connect mortgage lenders with homeowners to try to modify the loans so that the loans would not enter into forbearance. And there was a literature that tried to explain why there were not more modifications. So, there was actually very little debt modification during that period. And so, I think now the programs were just much more easily available. So, if you have a loan, when you would go online to check your balance, you would be offered basically a modification at that point and the procedure was very easy. I think that was the lesson. There was something in 2008 that made very difficult the modifications, and this made it easier. I mean, it was not the only difference, and we’ll talk more about that.
Engemann: In your article, you also mentioned that a stronger housing market this time around is helping troubled borrowers. Could you explain how that’s the case?
Sanchez: Yeah so, exactly, so when I was saying that there was difference in how debt modifications were offered and how they operate, it’s also a large difference in what happened with house prices during this period. During the 2008 episodes many homeowners were underwater. What that means, they have negative equity in their houses. Again, what this means is that the value of their houses were lower than the amount owed on the mortgage.
During this episode, again, because it was an episode in which we have to stay at home, and people wanted to have the best possible houses. And because of this reason, it was not a strong recession for the housing market. So, that’s how house prices were actually increasing during this period. That facilitated modifications when you are in a talk with the lenders and the value of the house is higher than the value of the mortgage, so it’s easier for modifications. Even if you want to sell your house or refinance, you have equity in your house, you can easily do it.
So, typical home values in the U.S. increased approximately 5% from June of 2020 to December of 2020. In contrast, typical home values decreased about 2% from June of 2009 to December of 2009. So, that’s a big difference. There were also difference, if you remember, leading to the financial crisis there was an increase in riskier mortgages with lower down payments, with adjustable rates. So, there was also a lesson in term of—at the time of this crisis, the pool of mortgage was better than in 2007.
Engemann: Do you have a sense of how many borrowers have entered forbearance during the pandemic?
Sanchez: With the data sets that we have, to be honest, we don’t know exactly where these homeowners are. So, there are indirect ways of obtaining these numbers, and I’ve seen estimates in a report by Equifax that says that 5.8% of the mortgage loans were on some possible accommodations, so that may be forbearance. Again, this is indirectly, and they also say that at the end of March of 2021, this 5.8% was coming down from 9.5% in May of 2020. We basically have, as I had mentioned before, data sets from the sides of the family of the credit report and we have data sets from the mortgage services. So, comparing repayments in this, one can get some estimates. So, it seems that at the peak, this number was 9.5% and it’s now down to about 5.8%. But again, this is obtained indirectly, so there’s some uncertainty about this number.
Engemann: Do you have a sense of what might happen once the forbearance and suspension of foreclosures ends?
Sanchez: I would say looking forward there is some uncertainty about how this is going to be resolved but there are different options. If things continue getting better with the pandemic, I don’t see a problem with this. But these programs are finishing now, so what is going to happen, so the households are going to have the option to make, for instance, make the payments at the end of the mortgage. So, if you have like 10 more years in the mortgage now, maybe if you didn’t pay for a year, you’re going to have 11 more years in the mortgage. So, that’s a possible way.
They could also make the payments whenever they have the opportunity to make some extra payments on this. There are some other possible modifications that they can negotiate with the lender. So, there are different options, but this is something that we are trying to follow. We continue to see in the numbers that more mortgages that were not being repaid are getting repaid. That’s kind of where we are with the forbearance programs and the situations in the mortgage market.
Engemann: All right. Well, I’d like to wrap up this podcast with another general question for you. As life continues to return to pre-pandemic ways in the U.S., what are your expectations for overall consumer debt going forward?
Sanchez: Well, in terms of credit cards and consumer credit in general, things have been improving. Although, if there is a lesson from the financial crisis, it’s that the improvement was relatively slow and took many years. I think the bigger uncertainty was what I mentioned in my answer to your previous question, as these forbearance programs end, what is going to happen? And this is something quite new, I don’t think we have experienced so many loans entering this program. So, this is something that we are following closely and, as I mentioned, we do see improvement in the last couple of months. So, we don’t expect it to be a problem, but I would say that the greatest uncertainty is what will happen with these mortgages.
Engemann: Thank you so much for joining us today, Juan, and for sharing insights from your research with us.
Sanchez: Thank you for the invitation.
Engemann: For more of Juan’s work, please visit the St. Louis Fed’s website at stlouisfed.org. You can find all of our Timely Topics podcasts on the St. Louis Fed’s website, Apple Podcasts, Spotify or wherever you listen to your podcasts.