housing market

Error message

Deprecated function: The each() function is deprecated. This message will be suppressed on further calls in _menu_load_objects() (line 579 of /var/www/drupal-7.x/includes/menu.inc).

If Prices Fall, Mortgage Foreclosures Will Rise

Published by Anonymous (not verified) on Wed, 08/09/2021 - 7:09am in

In our previous post, we illustrated the recent extraordinarily strong growth in home prices and explored some of its key spatial patterns. Such price increases remind many of the first decade of the 2000s when home prices reversed, contributing to a broad housing market collapse that led to a wave of foreclosures, a financial crisis, and a prolonged recession. This post explores the risk that such an event could recur if home prices go into reverse now. We find that although the situation looks superficially similar to the brink of the last crisis, there are important differences that are likely to mitigate the risks emanating from the housing sector.

Same Old Story?

Our last post demonstrated that price increases have been unusually strong and are now at rates not seen since just prior to their peak in 2005 at 16 percent year-over-year. Prices then fell 20 percent between mid-2006 and early 2009. By 2012 the average home had lost about a quarter of its 2006 value. Because prices had risen and fallen so fast, new mortgage originations in the period leading up to the peak, including those with substantial down payments, were quickly put into negative equity, which is a major risk factor for foreclosure, as both academic research and the experience of 2007-11 demonstrate.

Are developments in the housing market now essentially the same? Well prices have certainly been rising very fast, and mortgage originations have also been strong. The former, however, has outpaced the latter in recent months. Additionally, the owner’s share of housing wealth is 67 percent as of the first quarter of 2021, its highest value since 1989. (This is the value of the property minus the debt owed on it, expressed as a share of the property value.) Note that during the previous housing boom (1995-2006) this measure didn’t rise, in spite of sharply rising home prices. Additional borrowing was large enough to keep the owner’s share roughly constant at about 61 percent.

Of course this is an aggregate figure, and the distribution of leverage is a more important indicator of risks in the housing market. To examine this, we use the method developed in this Economic Policy Review article to assess current risks in the housing market at the property level. We begin by providing an updated Combined Loan to Value (CLTV) ratio for a large sample of mortgaged properties in the United States, using sale or appraised values at mortgage origination and estimating price appreciation using the CoreLogic Home Price Index. A property’s CLTV is the value of all debt secured by the property, divided by the value of the property, and is thus equal to 100 percent minus the owner’s equity as a share of the property value. The next chart shows the updated distribution of CLTV across current mortgage borrowers through December 2020. Note that our data set is limited to mortgage borrowers and thus excludes all properties owned “free and clear” (with no mortgage).

The Current Distribution of Mortgage Performance Indicators

Each color in the chart below corresponds to a CLTV level: light blue and dark gray correspond to properties that are in negative equity—a CLTV over 100 percent – while light gray, gold, and dark blue reflect decreasing leverage levels. As the chart shows, negative equity is very rare now, while more than 85 percent of all properties have a CLTV<80, meaning they have at least a 20 percent equity cushion. This sounds quite comforting: households generally have a lot of equity in their properties. However, the situation was much the same in 2005. We can look deeper into another indicator that complements equity position and shows a difference between now and 2005: credit score.

A Small Share of Properties Have High (>80%) Combined Loan to Value Ratios

Source: Authors’ calculations using data from CoreLogic and Equifax Credit Risk Insight Servicing McDash.

Credit score is another strong predictor of mortgage performance for borrowers, conditional on their equity position. While the CLTV distribution in 2020 looks quite similar to that in 2005, the credit score distribution definitely does not. More than two-thirds of mortgage debt in 2020 was held by borrowers with a FICO score above 740, compared to just over 50 percent on the eve of the housing crisis in the early 2000s. Perhaps more importantly, about 10 percent of current debt is owed by borrowers with a current score below 660, compared with nearly 20 percent in early 2006.

What if Prices Fall?

In order to fully understand the riskiness of this stock of debt, we go one step further by calculating expected delinquency transitions under various adverse price scenarios. Those scenarios are described in the table below and include prices which revert to their level from two years ago (HPI-2) and four years ago (HPI-4) in each county. (In cases where the stress scenario produces an increase in prices, we set price change to zero.) These scenarios are fairly severe in light of strong recent price growth – the median county would see prices fall by more than 27 percent under HPI-4 . Even the 10 percent least affected counties would see double-digit declines.

Stressed Home Price Scenarios

Price ScenarioPrice ScenarioHPI-2HPI-4Δ P, 10th percentile-16.6%–27.6%Δ P, 50th percentile-11.9-20.7Δ P, 90th percentile-7.8-13.1Max State DeclineIdahoIdaho, Utah

Source: Authors’ calculations based on data from CoreLogic.

What can we say about how mortgage performance would evolve under these scenarios? To start with, the price declines would drive many borrowers into negative equity, putting them at risk of default. Because of strong price growth since 2016, we estimate that Idaho, Utah, Nevada, and Arizona would all experience negative equity rates of more than 30 percent under HPI-4.

Yet the favorable credit score distribution would ameliorate the effects of these price declines on mortgage defaults. In the maps below, we show expected mortgage default rates under the two scenarios. More specifically, the maps show the 24-month transition rates for loans that were current (and not in forbearance) as of December 2020. To estimate these transitions, we use default rates from the 2007-10 period for each CLTV and FICO score combination, as described in Chart 12 of this paper. We estimate that 3.9 percent of mortgage balances overall would transition to delinquency by December 2022 under HPI-2 and 5.1 percent under HPI-4. These figures would be a significant increase in defaults over those observed in recent years, but they would fall far below the double-digit default rates observed during the crisis, when price declines were more severe, and the credit distribution was far less favorable.

As the maps show, however, there is significant geographic dispersion in the expected default rates. Relatively high rates of delinquency transitions appear in some expected places—Arizona, Florida, and Nevada are all above average risk for the HPI-2 and HPI-4 scenarios. But California is now at below average risk from these shocks, while Georgia, Idaho, Indiana, Mississippi, and Utah have emerged as newly vulnerable, given strong recent price growth in many of those states. Still, none of these states would be expected to match the national average default rates observed during the crisis, let alone the very high rates witnessed in Arizona, California, Florida, and Nevada.

24-month Serious Delinquency Forecasts: HPI 2 Years Ago

24-month Serious Delinquency Forecasts: HPI 4 Years Ago

Source: Authors’ calculations using data from CoreLogic and Equifax Credit Risk Insight Servicing McDash.

There is at least one caveat to this fairly benign scenario: mortgage forbearances. As noted above, our default estimates exclude loans already in forbearance. Those represent about 2.7 percent of loans in June 2020. Because widespread forbearance is a new approach to avoiding default and foreclosure, little is known about how these borrowers will fare when the programs end. Some share will likely be able to resume making payments, while others may have to sell their homes. Given strong price growth and very tight for-sale inventories of housing, these borrowers will generally have positive equity if they have to sell, enabling them to pay off their loans and avoid default. Nonetheless, the transition out of these programs is one additional factor to keep track of when monitoring housing risk.

Andrew Haughwout is a senior vice president in the Federal Reserve Bank of New York’s Research and Statistics Group,

Belicia Rodriguez is a senior research analyst in the Bank’s Communications and Outreach Group.

How to cite this post:
Andrew F. Haughwout and Belicia Rodriguez, “If Prices Fall, Mortgage Foreclosures Will Rise,” Federal Reserve Bank of New York Liberty Street Economics, September 8, 2021, https://libertystreeteconomics.newyorkfed.org/2021/09/if-prices-fall-mor...

Related Reading
Does the Rise in Housing Prices Suggest a Housing Bubble?
Tracking and Stress-Testing U.S. Household Leverage (Economic Policy Review)
Houses as ATMs No Longer 
Mapping Home Price Changes (interactive)

The views expressed in this post are those of the authors and do not necessarily reflect the position of the Federal Reserve Bank of New York or the Federal Reserve System. Any errors or omissions are the responsibility of the authors.

Does the Rise in Housing Prices Suggest a Housing Bubble?

Published by Anonymous (not verified) on Wed, 08/09/2021 - 7:08am in

House prices have risen rapidly during the pandemic, increasing even faster than the pace set before the 2007 financial crisis and subsequent recession. Is there a risk that another dangerous housing bubble is developing? This is a complicated question, and the answer has many components. This post, the first of two, provides a more detailed look at the recent rise in home prices by breaking it down geographically, with a comparison to the pre-2007 bubble. The second post looks at the potential risks to financial stability by comparing the currently outstanding stock of mortgage debt to the period before the financial crisis and projecting defaults should prices decline.

The Sharp Rise in Housing Prices during the Pandemic

The U.S. economy shut down in March 2020 due to the pandemic. Yet, by the summer housing prices started to rise sharply despite high unemployment. How similar is this to the early 2000s? We would be worried if the housing market were playing out exactly as it did in the prior boom. In the time series, we aren’t there yet: so far, we’ve had about one year of double-digit price growth, compared to the national average compound annual growth rate of more than 14 percent between 2003 and 2005.

Home Prices Are Rising Faster Now than during the Bubble

Source: CoreLogic Home Price Index, January 2003-June 2021.

Spatial Patterns of Home Price Growth

What about trends at the regional level? It turns out that the boom is taking place in different places within and across metro areas this time around. For most places, recent home price growth has been even stronger than during the previous boom: 79 percent of metropolitan areas in our data saw higher growth rates during the pandemic than during the peak years 2003-05. Of the thirty metropolitan areas containing the most populated cities in the country, 63 percent saw higher growth during the pandemic compared to 2003-05. In the chart below, we plot a 45-degree line, colored in gray, to differentiate which of the metropolitan areas with the largest population saw their fastest growth during either the pandemic or the housing bubble. Austin, Charlotte, Seattle, and Atlanta are a few metropolitan areas above the 45-degree line, meaning they have had higher growth rates during the pandemic. On the other hand, Las Vegas, Los Angeles, Miami, and New York had higher growth rates during 2003-05 and are below the 45-degree line. Some areas, however, saw similar paces of growth: Sacramento had minimal variation between its pandemic and housing bubble growth rates, putting the city close to the 45-degree line.

At the regional level, the northeast and south have positive trends in the graph, meaning that price increases are positively correlated in the two boom periods, whereas the midwest and west have slightly negative trends. The midwest points are clustered between growth rates of 10-20 percent for the pandemic and between 0-10 percent for 2003-05, whereas the other regions are more spread out. The west has the majority of its points above the trendline, while the south has most of its points near or below the trendline. The northeast points have the strongest positive relationship when compared to the other regions.

The blue regression line shows there is a positive relationship in the whole data set between house price growth during the housing bubble and the pandemic, meaning metropolitan areas that had high annual growth between 2003-05 saw higher growth rates during the pandemic, and vice versa. But note how flat the regression line is and how far away most of the dots are from the line, suggesting the relationship is weak. Many metropolitan areas that experienced fast-growing housing prices in 2003-05 have had slower growth rates during the pandemic and vice versa.  

Most Metro Home Prices Have Grown Faster during the Pandemic than during 2003-05

Source: CoreLogic Home Price Index.

Note: Each city represents the home price index of its respective metropolitan statistical area.

House Prices in Urban Areas Have Been Growing More Slowly than in Suburban and Rural Areas

The data above cover metropolitan areas and include both urban and suburban housing. A breakdown along these lines shows that house prices in urban areas have grown at a slower rate than those in suburban areas during the pandemic. To arrive at our urban classification, we first define the zip code that has the highest employment density, which we call the employment hub. We categorize zip codes as “urban” if they are within five miles of the employment hub, belong to a metropolitan statistical area, and have a population density greater than the 95th percentile. For suburban areas, we categorize zip codes as “suburban within 5/10/15/15+ miles” if they are within 5, 10, 15, or 15+ miles of the employment hub and if they are not already classified as urban (or any other suburban category).

As seen in the chart below, urban areas defined in this way have usually had the higher year-over-year house price growth compared to suburban areas, but starting around November 2018, these urban areas began to see lower rates of growth compared to suburban areas. Once the pandemic took hold in March 2020, urban areas did see a sharp increase in price growth, but suburban areas grew much faster and are above 15 percent year-over-year growth, whereas urban areas are around 10 percent. There are exceptions to even the relatively modest growth in urban areas: Manhattan (New York County) saw a price decline of 4.3 percent year over year in June, the largest county price decline nationwide.

Of course, many factors other than relative location may affect price growth. But urban classification is a significant characteristic even controlling for some of these other factors. The significant lag of home price growth in the past year isn’t attributable to zip code income or the level of home prices before the pandemic. When we control for these factors, it turns out that dense urban areas had been growing at a pace close to that of other parts of metro areas, until 2020 when they fell way behind.

Urban Home Prices Have Underperformed during the Pandemic

Source: CoreLogic Home Price Index.

There are also regional differences within urban areas. The northeast is not growing as rapidly as the midwest, west, and the south. Up until the end of 2020, all regional lines were following similar trends throughout the pandemic. At the beginning 2021 the west, south, and midwest continued to grow rapidly while the northeast began to see a slight stagnation in growth. These regional differences may have to do with the different rates of growth of cities in these areas compared to cities in other areas, and this shows how the urban classification can manifest differently depending on the region.

Urban Zip Codes Have Slower Home Price Growth in the Northeast

Source: CoreLogic Home Price Index.

Although prices are increasing rapidly nationwide, the data show we are not simply repeating the housing market bubble of the early 2000s during the pandemic. This boom is taking place in different metro areas and in different locations within metros. Still, home price growth in excess of 15 percent per year can’t be sustained forever, so a remaining question is how price growth will normalize and what the consequences of a decline in prices could be. We turn to this question in our next post.

Andrew Haughwout is a senior vice president in the Federal Reserve Bank of New York’s Research and Statistics Group.

Belicia Rodriguez is a senior research analyst in the Bank’s Communications and Outreach Group.

How to cite this post:
Andrew Haughwout and Belicia Rodriguez, “Does the Rise in Housing Prices Suggest a Housing Bubble?,” Federal Reserve Bank of New York Liberty Street Economics, September 8, 2021, https://libertystreeteconomics.newyorkfed.org/2021/09/does-the-rise-in-h....

Related Reading
Mapping Home Price Changes (interactive)
Keeping Borrowers Current in a Pandemic (May 2021)
Do People View Housing as a Good Investment and Why? (April 2021)

The views expressed in this post are those of the authors and do not necessarily reflect the position of the Federal Reserve Bank of New York or the Federal Reserve System. Any errors or omissions are the responsibility of the authors.

Keeping Borrowers Current in a Pandemic

Published by Anonymous (not verified) on Tue, 27/07/2021 - 5:56am in

Andrew Haughwout, Donghoon Lee, Joelle Scally, and Wilbert van der Klaauw

Keeping Borrowers Current in a Pandemic

Federal government actions in response to the pandemic have taken many forms. One set of policies is intended to reduce the risk that the pandemic will result in a housing market crash and a wave of foreclosures like the one that accompanied the Great Financial Crisis. An important and novel tool employed as part of these policies is mortgage forbearance, which provides borrowers the option to pause or reduce debt service payments during periods of hardship, without marking the loan delinquent on the borrower’s credit report. Widespread take-up of forbearance over the past year has significantly changed the housing finance system in the United States, in different ways for different borrowers. This post is the first of four focusing attention on the effects of mortgage forbearance and the outlook for the mortgage market. Here we use data from the New York Fed’s Consumer Credit Panel (CCP) to examine the effects of these changes on households during the pandemic.

Background: Who Qualifies for COVID-19 Mortgage Forbearance?

Initially, under the CARES Act, borrowers with federally backed mortgages could request up to twelve months of forbearance, made up of two 180-day periods, if they experienced financial hardship because of COVID-19. Several agencies have subsequently granted extensions. Specifically, borrowers with mortgages backed by the government-sponsored enterprises (GSEs) Fannie Mae and Freddie Mac can request up to two additional three-month extensions (for a maximum of eighteen months of total forbearance) if they were in an active forbearance plan as of February 28, 2021, while borrowers with mortgages backed by the Department of Housing and Urban Development/Federal Housing Administration (HUD/FHA), the U.S. Department of Agriculture (USDA), or the Department of Veterans Affairs (VA) can enroll in forbearance until June 30, 2021, and receive up to eighteen months of total forbearance. At the same time, the CARES Act (section 4013) eased the accounting treatment of pandemic-related modifications for loans in bank portfolios, and the federal banking agencies released guidance to that effect in early April 2020.

How Does Forbearance Work?

Widespread forbearance is a new policy, although similar programs have previously been rolled out on a smaller scale in the wake of natural disasters. Typically, the missed payments will be added to the end of the loan; for a borrower in the first year of a thirty-year mortgage, a forbearance thus amounts to a twenty-nine-year interest-free loan of the forborne amount. These forbearances are safe loans in part because they are incentive-compatible: in order to preserve their housing equity, borrowers must resume payments when they are able. (Note that renters, with no equity in their property, do not have strong incentives to pay back forborne rent payments, making the provision of relief to renters more difficult than to owners.)

Who Entered Forbearance?

As we reported back in November, large numbers of mortgage forbearances began to appear on credit reports in April 2020, and by May 2020, 7 percent of mortgage accounts were in forbearance. By June, however, exits from forbearance began to outweigh entries, and the number of mortgages in forbearance began a slow decline. The following chart shows that by March 2021, the overall forbearance rate had fallen to 4.2 percent, accompanied by reductions in both entries and exits, suggesting a relatively stable group of borrowers in forbearance for a relatively long period of time. In fact, of the 2.2 million mortgages still in forbearance in March 2021, 1.2 million entered forbearance in June 2020 or earlier. (In March, the inflows into forbearance are likely affected by additional payment relief offered in Texas as a response to the effects of the winter storm there.)

Keeping Borrowers Current in a Pandemic

These dynamics—a sharp rise in April and May, followed by a slow decline through the summer and fall—are common across most types of mortgages, but FHA borrowers were considerably more likely to take up mortgage forbearance initially, and have remained in the program longer. As of March 2021, more than 11 percent of FHA borrowers remain in forbearance, as shown below.

Keeping Borrowers Current in a Pandemic

What accounts for the higher forbearance rates for FHA borrowers? FHA borrowers are much more likely to be first-time home buyers and to live in lower-income areas. About 41 percent of FHA borrowers live in neighborhoods with average annual household income below $50,000, compared to 22 percent for GSE borrowers.

With this context, it’s perhaps not surprising to find that forbearance rates rose most, and were most persistent, in lower average income zip codes. As shown in the next chart, forbearance rates in the poorest quartile of zip codes approached 10 percent in May and June 2020 and remain above 5.5 percent at the end of March 2021.

Keeping Borrowers Current in a Pandemic

The likelihood of forbearance falls steadily as borrower credit score (measured at the date of mortgage origination) rises, and it is far higher for loans that were delinquent in March 2020; see the next chart. Indeed, forbearance rates remain near 40 percent for borrowers who were delinquent on their mortgages pre-pandemic. The higher rates of mortgage forbearance in poorer areas and among FHA borrowers is consistent with the uneven impact that COVID-19 and the accompanying recession have had on different segments of the population. Mortgage forbearance has been an important policy tool to mitigate the impact of these challenges faced by less-advantaged households.

Keeping Borrowers Current in a Pandemic

Since housing costs are typically one of the largest household expenses, it isn’t surprising that mortgage forbearance offers very substantial cash flow relief to the households that take it up. The table below provides details on the payment relief that different forbearance participants received. (As we show in a companion post, being enrolled in forbearance isn’t quite the same as receiving cash flow relief.) We estimate this relief using the average payment that was due prior to enrolling in forbearance for those who were enrolled in forbearance as of March 2021. (These figures have been very stable since March 2020, so we don’t show the changes over time here.)

Keeping Borrowers Current in a Pandemic

As the table shows, the average monthly cash flow relief associated with a mortgage forbearance is somewhat different across different mortgage types and grows sharply as neighborhood income rises. Indeed, aggregate cumulative payments skipped by borrowers from the poorest 25 percent of neighborhoods are about 38 percent of those skipped in top-quartile neighborhoods.

All told, in absolute dollar terms, mortgage forbearance has brought the most benefit to the highest-income areas. This is due to a combination of high homeownership and relatively expensive mortgage payments in these areas, which more than offsets the considerably higher incidence of forbearance in lower-income areas. Still, the high rates of forbearance take-up on FHA loans and in poorer zip codes makes clear that these programs have been an important lifeline to less-advantaged households.


We find that mortgage forbearance has been an important policy tool to mitigate the impact of the pandemic and has become a fairly common phenomenon since it became widely available last year. After an initial rapid rise to over 7 percent, the share of mortgages in forbearance has slowly declined and stood at just over 4 percent in late March 2021. Forbearance has been more common for FHA borrowers and mortgagors from poorer neighborhoods, as well as those who were already delinquent in March 2020. In a separate post, we look at how being in forbearance affects borrowers, and continue to look at the distribution of those effects.

Andrew F. HaughwoutAndrew F. Haughwout is a senior vice president in the Federal Reserve Bank of New York’s Research and Statistics Group.

Donghoon LeeDonghoon Lee is an officer in the Bank’s Research and Statistics Group.

Joelle ScallyJoelle Scally is a senior data strategist in the Bank’s Research and Statistics Group.

Wilbert van der KlaauwWilbert van der Klaauw is a senior vice president in the Bank’s Research and Statistics Group.

How to cite this post:

Andrew F. Haughwout, Donghoon Lee, Joelle Scally, and Wilbert van der Klaauw, “Keeping Borrowers Current in a Pandemic,” Federal Reserve Bank of New York Liberty Street Economics, May 19, 2021, https://libertystreeteconomics.newyorkfed.org/2021/05/keeping-borrowers-....

Additional Posts in This Series

What Happens during Mortgage Forbearance?

Small Business Owners Turn to Personal Credit

What’s Next for Forborne Borrowers?

Related Reading
Economic Inequality: A Research Series

Press Briefing

Keeping Borrowers Current in a Pandemic


The views expressed in this post are those of the authors and do not necessarily reflect the position of the Federal Reserve Bank of New York or the Federal Reserve System. Any errors or omissions are the responsibility of the authors.