mortgages

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).

How does house price indexation affect the valuation of equity release mortgages?

Published by Anonymous (not verified) on Fri, 05/11/2021 - 8:00pm in

Craig Turnbull

Equity Release Mortgages (ERMs) are different from traditional mortgages. Both mortgages provide an upfront cash lump sum. But traditional mortgages are tied to an immediate home purchase that is repaid over a set period, while equity release mortgages are tied to a share of a future home sale. In this blog post, I examine some of the challenges with valuing equity release mortgages. Specifically, I focus on the approaches used to estimate the current home value – a key input to the mortgage valuation – which typically involves applying a simple house price index return to the most recent house survey valuation or transaction price. I show this widespread approach may understate equity release mortgage risks and overstate portfolio values.

Valuations for equity release mortgages

In the UK, equity release mortgages almost always include a ‘No-Negative Equity Guarantee’ (NNEG). These guarantees ensure that the lender cannot recover more than the proceeds of the house sale at the repayment date. Option pricing techniques play a natural role in the valuation of the mortgage’s NNEG. These option pricing approaches are now widely used both in Prudential Regulation Authority supervisory tools such as the Effective Value Test (EVT) and in firms’ asset valuation methodologies.

Such approaches to mortgage valuation take an up-to-date price of the underlying house as a given. But the house’s most recent transaction price may be decades old. And, for mortgages that have been in-force for many years, a considerable time may have passed since the house was last subject to a surveyor’s valuation.

There is wide recognition that this approach to estimating the current house value has the potential to overstate mortgage portfolio values: the use of the index return ignores the idiosyncratic risk element in the evolution of the house price and this will, on average, understate a portfolio of NNEG values and hence overstate mortgage portfolio values.

An illustrative model to value ERMs

In this section, I set out a simplified model to value ERMs. This framework is used to illustrate the impact that different house price estimates can have on ERM portfolio values.

Let’s suppose the mortgage has a fixed and known maturity date which is 30 years from origination of the loan (ERMs will actually mature on death or entry into long-term care, and may have prepayment options, but none of this is key to this analysis).

The example mortgage has a starting loan to value ratio of 30%. The mortgage is written with an interest rate of 3.63% (ie the loan amount compounds at 3.63% so that, at maturity, it has increased from 0.30 to 0.874, assuming a starting house value of 1). A Black-Scholes option pricing approach is used to value the NNEG such that the starting valuation of the mortgage is consistent with what has been advanced to the borrower. These assumptions are set out in the technical appendix.

Next, suppose we write 1,000 mortgages on the above terms, on different houses. We stochastically project the value of this mortgage portfolio over time. Within the projection, we will value the mortgage portfolio in two different ways: one, by using the simulated ‘true’ house prices projected by the model up to that point in the projection; two, by using house price estimates that have been produced by applying the simulated house price index returns to the starting house prices.

We need to make one more assumption. How volatile is the house price index relative to individual house prices? This is key to our analysis. If house prices are perfectly correlated and there is no idiosyncratic house price risk, then the process of using index returns to update prices will not produce any valuation errors (as the house price volatility and the index volatility will be the same). But we don’t expect house prices to be perfectly correlated, and the diversification benefit that this delivers means index volatility will be lower than the volatility of individual house prices. In this example, we assume the index volatility is 12% (see Technical appendix).

The impact of different house price valuations on ERM portfolio values

We can now generate some results from the model. Chart 1 and Chart 2 show the probability distributions of the projected mortgage portfolio values in the two cases: first, where the actual modelled house prices are used in the NNEG valuations; and second where the house prices are updated using index returns.

Chart 1: Projected ERM portfolio values using accurate house price updates

Chart 2: Projected ERM portfolio values using indexed house price updates

A comparison of the two charts highlights that the indexed approach results in higher returns being generated by the portfolio over time, as a result of the NNEG values being systematically understated by the use of indexation. At maturity, the true house prices are ‘revealed’, and this sometimes results in unanticipated write-downs at maturity.

Chart 3 shows the projected behaviour of the ratio of the portfolio valuation that results from indexation to the portfolio value using the accurate house prices. We define this ratio as the portfolio valuation error.

Chart 3: Projected portfolio valuation errors

You can see from Chart 3 that the portfolio valuation error is always greater than 1: the use of indexation in the presence of idiosyncratic risk systematically biases the portfolio valuation upwards. After 10 years, the median portfolio valuation error is around 3%. As the period over which indexation is applied grows, the potential for very material valuation errors also grows.

Chart 4 shows how the simulated mortgage portfolio valuation errors after 29 years of indexation behave relative to the simulated 29-year index return.

Chart 4: Portfolio valuation errors and the level of index returns (29 years)

Chart 4 highlights that the mortgage portfolio valuation error after 29 years has a very strong dependency on the level of index returns that has been experienced over the same period. This is intuitive. If index returns have been very strong for 29 years, it is likely the actual NNEG losses of the portfolio will be very low, and the indexation approach will therefore not overstate the final losses – even bad idiosyncratic risk outcomes are unlikely to result in NNEG losses when index returns are so strong. And if the index returns have been very poor, such that the NNEGs tend to be deeply in the money, then the impact of the idiosyncratic risk on the mortgage lender will be symmetric, and the under-recognition of the idiosyncratic risks will not be consequential (the borrower will participate pound for pound in good and bad idiosyncratic risk outcomes, because even the positive idiosyncratic risk outcomes will still tend to result in NNEGs that are in the money). However, if the index returns have been such that the NNEGs are likely to mature close to the money, then idiosyncratic risk can really matter: the lender will tend not to get the upside of the good idiosyncratic risk outcomes (as the final NNEG cash flow cannot be negative) but will be exposed to the downside. Here, the effect of ignoring idiosyncratic risk in the updating of house prices can result in a material overstatement of the mortgage portfolio value.

In reality, insurance firms may implement various strategies to mitigate these potential over-valuation effects. For example, firms may undertake regular ‘drive-by’ house valuations for the larger mortgages in their portfolios. But Chart 4 (albeit based on the fairly pathological example of 29 years of indexation) suggests another approach to adjusting mortgage portfolio values – using a model such as this to derive valuation adjustment factors that are applied to mortgage values where indexed house prices have been used. The following, Chart 5 shows the valuation adjustment factors produced for our example mortgage.

Chart 5: Valuation adjustment factors for the 30-year 30% LTV mortgage

The scale of these adjustment factors will heavily depend on the assumed level of house price idiosyncratic risk. The greater the idiosyncratic risk, the greater will be the implied valuation adjustment. The purpose of this analysis is not to propose a specific parameterisation, but to highlight that analytical techniques can be used to shed light on the mortgage valuation errors that can arise from the use of indexed house prices.

Technical appendix: modelling assumptions

The mortgage is valued using an illiquid risk-free rate of 2.5%; a house price volatility of 15%; and a 30-year house deferment rate of 2.5%. (Please note these numbers are used for illustrative purposes only.)

These assumptions imply an initial pre-NNEG value for the mortgage of 0.413 and a NNEG value of 0.113, and hence a starting mortgage value of 0.300.

In the stochastic projections a house risk premium of 3.0% is assumed (and so expected house price inflation will be equal to the illiquid risk-free rate (2.5%) plus house risk premium (3.0%) minus deferment rate (2.5%) equals 3.0%).

The stochastic projections assume a house price index volatility of 12%. The model assumes all houses have the same level of idiosyncratic risk and a simple single-factor structure to their correlation. These assumptions, together with the house price volatility assumption of 15% above imply an idiosyncratic risk of 9%.

Craig Turnbull works in the Life Insurance Actuaries Division at the Bank of England.

If you want to get in touch, please email us at bankunderground@bankofengland.co.uk or leave a comment below.

Comments will only appear once approved by a moderator, and are only published where a full name is supplied. Bank Underground is a blog for Bank of England staff to share views that challenge – or support – prevailing policy orthodoxies. The views expressed here are those of the authors, and are not necessarily those of the Bank of England, or its policy committees.

The Housing Boom and the Decline in Mortgage Rates

Published by Anonymous (not verified) on Thu, 09/09/2021 - 6:01am in

During the pandemic, national home values and housing activity soared as mortgage rates declined to historic lows. Under the canonical “user cost” house price model, home values are held to be very sensitive to interest rates, especially at low interest rate levels. A calibration of this model can account for the house price boom with the observed decline in interest rates. But empirically, we find that home values are nowhere near as sensitive to interest rates as the user cost model predicts. This lower sensitivity is also found in prior economic research. Thus, the historical experience suggests that lower interest rates can only account for a tiny fraction of the pandemic house price boom. Instead, we find more scope for lower interest rates to explain the rise in housing activity, both sales and construction.

Since February 2020, national home values have risen more than 15 percent across several house price indices. At the same time, existing home sales and building permits for new privately owned housing units have soared to levels last seen in 2007, and Q4/Q4 real residential investment, as measured by the Bureau of Economic Analysis, grew about 16 percent in 2020. Thirty-year fixed rate mortgage rates dropped to an historic low of 2.7 percent in December 2020. At 3 percent during the summer of 2021, mortgage rates remain depressed and 50 basis points below February 2020 levels. How much of the housing boom can be explained with the lower level of interest rates?

Elasticity of House Prices to Mortgage Rates in Theory: The User Cost Model

Standard calibrations of the most popular theoretical framework of housing valuation—the user cost model—can in fact quantitatively explain the rise in house prices with the decline in interest rates. In its simplest form, the model postulates that the raw return on housing, including both the rent yield and growth of rent, should be equal to the sum of borrowing cost and property taxes, maintenance, and insurance (taking housing supply and rents as given):

where is the rent to price yield, g is the expected capital gain rate, ρ is the effective borrowing cost (mortgage rate after tax deduction), and τ accounts for property taxes, maintenance, and insurance.

From this formula, we can calculate how much home prices rise for every 1 percentage point decline in the mortgage rate, or the semi-elasticity of house prices to changes in mortgage rates, which we refer to as the “semi-elasticity.” The chart below illustrates the predicted semi-elasticity under a set of commonly used parameters from Himmelberg et al. (2005) (marginal tax rate at 25 percent, property taxes, maintenance and insurance in total at 4 percent, growth of rent at 3.8 percent). Importantly, the semi-elasticity rises as interest rates decline, meaning that house prices become particularly sensitive to interest rate changes in a low-rate environment. For example, the semi-elasticity is about 23 when mortgage rates are at 4 percent, but it increases to about 30 when mortgage rates are at 3 percent. These statistics suggest that a decline in the mortgage rate from 3.5 percent to 3 percent would cause home prices to rise about 14 percent, which is just about as we observed. But as we show in the rest of the post, these predicted effects are much larger than our empirical estimates and those found in the economic literature.

User Cost Model Predicts High Sensitivity of House Prices to Mortgage Rates

Note: This graph plots the semi-elasticity of house prices to changes in the mortgage rate as a function of the mortgage rate, or how much home prices rise for every 1 percentage point decline in the mortgage rate.

Source: Authors’ calculations.

Elasticity of House Prices and Activity to Mortgage Rates in Practice: Empirical Evidence

We use a Jorda (2005) linear projection framework and quarterly macroeconomic data between 1975 and 2020 to study the semi-elasticity of the FHFA house price index, building permits for single-family units, existing home sales, and residential investment (rescaled as a contribution to real GDP). Estimating the semi-elasticity to interest rates using macroeconomic data is challenging because movements in mortgage rates depend on the state of the economy, which is a confounding factor. Our first econometric specification (no contemporaneous controls) accounts for the economic state by including past realizations of the unemployment rate, the 1-year Treasury rate (as a proxy for monetary policy), and CPI inflation. In the second specification (contemporaneous controls), we include these controls contemporaneously, so that the estimated semi-elasticity only reflects changes in the residual mortgage rate component, which is driven by term premia for long-term rates and the mortgage basis (or spread). All specifications also include lags of each housing variable, so that the linear projections are equivalent to impulse responses from vector autoregressions commonly used in empirical macroeconomic analysis (Plagborg-Moller and Wolf 2021).

The chart below shows the response of each variable when excluding (orange) and including (blue) the contemporaneous controls after a 1 percentage point decline in the mortgage rate. We find significant effects, with the expected signs, and with stronger and more persistent responses when not including controls. The (maximum) semi-elasticity of house prices to mortgage rates is -2, with or without contemporaneous controls. This is less than a tenth as large as what is predicted by the user cost model. Housing activity is, instead, very sensitive to mortgage rates. After a 1 percentage point decline in mortgage rates, permits rise more than 10 percent, existing homes sales increase 5-10 percent, and residential investment expressed as a contribution to real GDP increases 0.3 percentage point (0.2 percentage point with controls).

Mortage Rates Affect House Prices and Housing Activities

Notes: The above charts show impulse responses to a 100 basis point (or 1 percentage point) decline in mortgage rates from linear projections (LPs). The LPs include lags of each dependent variable, mortgage rates, and a set of controls that include the unemployment rate, the 1-year Treasury rate (as proxy for monetary policy), and CPI inflation. The “contemporaneous controls” specification includes these controls up to quarter 0. Blue and orange bands are 90 percent confidence bands for the model, including contemporaneous controls and only lagged controls.

Sources: Authors’ calculations, based on data from Federal Reserve Board, Bureau of Labor Statistics, Freddie Mac, FHFA, Census Bureau, National Association of Realtors, and Bureau of Economic Analysis.

Elasticity of House Prices to Mortgage Rates in Practice: Prior Research

To what extent are our results specific to our statistical model?  Prior research uses both macro and micro data to estimate the semi-elasticity of house prices to interest rates (in contrast, few studies look at the response of housing activity to interest rates). As shown in the table below, most empirical estimates from this literature suggest that house prices increase by less than 5 percent for every 1 percentage point decrease in (long-term) interest rates–substantially less than implied by the user cost model and consistent with our results.

Estimated Effects of Interest Rates on House Prices

PaperU.S./
International
MethodHome Price Appreciation After a 1 Percentage Point Drop in the Mortgage RateMacro Papers:Del Negro and Otrok 2007U.S.VAR0.8Goodhart and Hofmann 2008InternationalVAR1.6Jarocinski and Smets 2008U.S.VAR2Sa, Towbin, and Wieladek, 2011InternationalVAR1.2Williams 2015InternationalFixed exchange rate6.3    Micro Papers:   DeFusco and Paciorek 2017U.S.Bunching around CLL[1.5, 2]Adelino, Schoar, and Severino 2020U.S.Diff in diff around CLL[1.3, 5.3]Davis, Oliner, Peter, Pinto 2020U.S.Cut in FHA insurance premium3.4Fuster and Zafar 2021U.S.Consumer survey2.5

Notes: This table summarizes the literature on the relationship between house prices and interest rates. The “macro papers” panel summarizes five papers from the large literature using macroeconomic data. The second panel reviews four papers based on microeconomic data. The “U.S./International” column reports whether the study is based on U.S. or international data. The “Method” column reports the identification strategy. “CLL” stands for conforming loan limit. The last column reports the estimated effect on house prices after a 1 percentage point interest rate shock. For macro papers, these effects are 10 quarters after a 1 percentage point monetary policy shock. For the micro papers, the effects are for a 1 percentage point shock in the mortgage rate.

One difference between the macro and the micro literatures is in the measure of interest rate used. Macro papers tend to study monetary policy shocks or shocks to long term rates, either nominal or real. Micro studies often focus on mortgage rate shocks using arguably exogenous cutoffs at mortgage origination. For example, one such cutoff is the conforming loan limit (CLL). Most mortgages that are smaller than the CLL are guaranteed by Fannie Mae or Freddie Mac, enjoying lower interest rates than loans that are larger than the CLL, also known as jumbo loans. By looking at house prices around this exogenous cutoff and how they change when the CLL increases, Adelino, Schoar, and Severino (2020) find a semi-elasticity between -5.3 and -1.3.

Rather than using mortgage and house price data, Fuster and Zafar (2021) use the housing module of the New York Fed Survey of Consumer Expectations to elicit how much survey respondents are willing to pay for the same house in two hypothetical scenarios: when the mortgage rate is 4.5 percent or 6.5 percent. They find that even a 2-percentage point increase in the mortgage rate only lowers borrowers’ willingness to pay by 5 percent.

Conclusion

The semi-elasticity of house prices to interest rates implied by the theoretical user cost model suggests that the decline in mortgage rates during the pandemic can quantitatively account for the national house price boom. But our empirical estimates and prior studies suggest that the decline in mortgage rates can only explain low single-digit house price increases. Instead, we find that housing activity, both sales and construction, are very sensitive to interest rates.

Haoyang Liu is an economist in the Federal Reserve Bank of New York’s Research and Statistics Group.

David Lucca is a vice president in the Bank’s Research and Statistics Group.

Dean Parker is a senior research analyst in the Bank’s Research and Statistics Group.

Gabriela Rays-Wahba is a senior research analyst in the Bank’s Research and Statistics Group.

How to cite this post:
Haoyang Liu, David Lucca, Dean Parker, and Gabriela Rays-Wahba, “The Housing Boom and the Decline in Mortgage Rates,” Federal Reserve Bank of New York Liberty Street Economics, September 7, 2021, https://libertystreeteconomics.newyorkfed.org/2021/09/the-housing-boom-a....

Disclaimer
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.

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)

Disclaimer
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)

Disclaimer
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.

Forbearance Participation Declines as Programs’ End Nears

Published by Anonymous (not verified) on Wed, 01/09/2021 - 6:54am in

Tags 

Credit, mortgages

The Federal Reserve Bank of New York’s Center for Microeconomic Data today released its Quarterly Report on Household Debt and Credit for the second quarter of 2021. It showed that overall household debt increased at a quick clip over the period, with a $322 billion increase in balances, boosted primarily by a 2.8 percent increase in mortgage balances, a 2.2 percent increase in credit card balances, and a 2.4 percent increase in auto balances. Mortgage balances in particular were boosted by a record $1.22 trillion in newly originated loans. Although some borrowers are originating new loans, struggling borrowers remain in forbearance programs, where they are pausing repayment on their debts and creating an additional upward pressure on outstanding mortgage balances.

This analysis, as well as the Quarterly Report on Household Debt and Credit are based on the New York Fed’s Consumer Credit Panel (CCP), a 5 percent representative sample of consumer credit reports from Equifax, which we have used for monitoring the health of the household balance sheet and the uptake of forbearance since the arrival of the pandemic. While pandemic related mortgage forbearances were declining over time, thanks to a recovery of economic activity and stimulus payments, there were some increases in inflows to mortgage forbearance not directly related to the pandemic.

When the state of Texas and some neighboring states were hit with severe weather in winter of 2021, almost 70 percent of borrowers lost power, and many borrowers experienced damage to their properties, producing additional mortgage forbearances. Texas is a big state, enough to produce fluctuations in the forbearance rate for the United States overall. Thus, in the chart below, we break out these effects separately, we exploit the richness in our forbearance data to break out these effects separately. We observe that there has been a steady outflow of borrowers in forbearance as the pandemic has begun to wane, with just 2.7 percent of mortgages still in forbearance at the end of June 2021 (when we include storm-related forbearances, the national rate is 3.3 percent).

Forbearance participation, entry, and exit

Source: New York Fed Consumer Credit Panel / Equifax.

In the map below, we look at the forbearance participation rates of all states, beyond Texas. Texas and Oklahoma rank the highest, both having been affected by the storms: it is unlikely that this reflects forbearances associated with the CARES Act and more likely are weather related. The map reveals substantial geographic differences in the participation in pandemic mortgage forbearance programs, with states in the Southeast having the highest lingering forbearance rate, excepting the Carolinas. This dispersion likely captures a combination of the severity of the pandemic’s local economic impact and differences in the borrowers’ awareness, interest in, and access to these programs.

Percent of Mortgages in Forbearance by State, June-2021

Source: New York Fed Consumer Credit Panel / Equifax.

Finally, we look at the composition of borrowers who remain in forbearance. During the course of the past fifteen months, 9.3 million mortgages were in forbearance at some point. However, only 21 percent of those mortgages, or just under 2 million, remained in forbearance as of the end of June 2021. As we described in our series in May (see Related Reading below), the remaining pool of forborne borrowers has a notably lower average credit score compared to participants a year ago. The chart below examines the pool of mortgages in forbearance by the borrowers pre-pandemic credit score. A year ago, 32 percent of mortgages in forbearance were associated with prime borrowers with credit scores of 760 or higher. But the composition has shifted, and that share has shrunk to 25 percent, with 39 percent of the remaining mortgages in forbearance held by borrowers with credit scores under 620. This is especially remarkable when we consider the overall pool of mortgages skews much more toward the prime–only 18 percent of mortgages outstanding are associated with subprime borrowers.

We find similar, although less pronounced, results when we look at the same share broken out instead by the borrowers’ neighborhood income. Although we do not directly observe income on credit reports, we are able to group our borrowers using the average income of their zip code using data from the IRS Statistics of Income. Mortgages held by borrowers living in the lowest income neighborhoods, comprise only 16 percent of the pool of outstanding mortgages but were 21 percent of the mortgages in forbearance at the end of June. This is consistent with the finding that uptake of mortgage forbearance is correlated with the mortgage origination credit score.

Subprime Borrowers Are Much More Likely to Be In Forbearance, and More Likely to Remain

Source: New York Fed Consumer Credit Panel / Equifax.

Looking forward, those borrowers remaining in forbearance are the borrowers that will be most likely to struggle when these policies are lifted. Federally backed mortgages are eligible for up to eighteen months of forbearance, depending on when the initial forbearances started. With this limit in effect, many of the borrowers remaining in forbearance will face the limits of this leniency in the coming 3-4 months, when they must either resume payment or will be pressured to sell their homes to avoid delinquencies and consequent foreclosures. It appears that the CARES Act and other policy supports to households has been largely successful at avoiding widespread damage to household balance sheets, with little evidence of increasing delinquency in our Quarterly Report. It is unclear, however, how the unwinding of these programs in a still recovering economy will impact mortgage borrowers across different income and credit score groups and geographic locations.

Chart data

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

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

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

Wilbert 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, “Forbearance Participation Declines as Programs’ End Nears,” Federal Reserve Bank of New York Liberty Street Economics, August 3, 2021, https://libertystreeteconomics.newyorkfed.org/2021/08/forbearance-partic...

Related Reading

Keeping Borrowers Current in a Pandemic
What Happens during Mortgage Forbearance?
Small Business Owners Turn to Personal Credit
What’s Next for Forborne Borrowers?

Disclaimer

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.

Mortgage Rates Decline and (Prime) Households Take Advantage

Published by Anonymous (not verified) on Fri, 13/08/2021 - 1:20am in

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

LSE_2021_HDC_mortgage_scally_460_art

Today, the New York Fed’s
Center for Microeconomic Data
reported that household debt balances increased by $206 billion in the fourth quarter of 2020, marking a $414 billion increase since the end of 2019. But the COVID pandemic and ensuing recession have marked an end to the dynamics in household borrowing that have characterized the expansion since the Great Recession, which included robust growth in auto and student loans, while mortgage and credit card balances grew more slowly. As the pandemic took hold, these dynamics were altered. One shift in 2020 was a larger bump up in mortgage balances. Mortgage balances grew by $182 billion, the biggest quarterly uptick since 2007, boosted by historically high volumes of originations. Here, we take a close look at the composition of mortgage originations, which neared $1.2 trillion in the fourth quarter of 2020, the highest single-quarter volume seen since our series begins in 2000. The Quarterly Report on Household Debt and Credit and this analysis are based on the New York Fed’s Consumer Credit Panel, which is itself based on anonymized Equifax credit data.

In the first chart, we look at a figure direct from the Quarterly Report: total mortgage originations by credit score band. Here, the boom in originations is starkly visible—originations to the highest credit score borrowers rose sharply during 2020. (We use the Equifax Risk Score 3.0). The origination volume in the fourth quarter of 2020 just surpassed the previous high, from 2003, when a dip in mortgage interest rates prompted a boom in mortgage refinancing. Although these two bumps in mortgage originations are similar in magnitude, the composition is quite different; 71 percent of originations in the fourth quarter of 2020 went to borrowers with credit scores over 760, while in the third quarter of 2003, only 31 percent of those new mortgages went to the most creditworthy borrowers. Researchers have concluded that the 2003 refi boom had long-running consequences, contributing to over-leveraged balance sheets as home prices fell.

LSE_2021_HDC-mortgage_scally_ch1-03

Purchases or Refis?

Refinances do tend to go to higher-credit-score borrowers, compared to purchase mortgages—this is because those who are refinancing already have a mortgage that they’ve been repaying for some time and are building their credit history. Thus, this high volume of superprime mortgage originations suggests we are in the midst of another refinance boom. And homeowners are taking advantage of the mortgage interest rates at historic lows, even taking some cash out with their refinanced mortgages. In the following chart, we break out mortgage originations into purchase and refinances. The origination boom in 2020 is owing to a confluence of high refinance originations as well as high purchase originations—in fact, in nominal terms, the level of purchase originations nears that seen in 2006. As for the increasing level of purchase originations, it’s consistent with high levels of home sales paired with increasing prices. Overall, refinance originations in 2020 were at their highest level since 2003. Although about 15 percent below the 2003 level in terms of nominal aggregate debt, some 7.2 million mortgages were refinanced in 2020, which was less than half the 2003 count (15 million!).

LSE_2021_HDC-mortgage_scally_ch2-02

Who’s Buying? Looking Into Purchase Originations

Although it is difficult to compare the volume of purchase originations with the levels seen in 2005-06 due to differences in home prices, the trend was unmistakably increasing this year, and to a high level. In the following chart, we break purchases out into three groups: (1) first-time homebuyers, or those whose origination marks the first mortgage on their credit report, shown in blue; (2) repeat buyers – individuals who are making a purchase that is not their first, shown in red; and (3) second home buyers and investors—those who have a new origination with one or more first mortgages already on their credit report. The boom years preceding the Great Recession—which similarly saw a brisk pace of mortgage originations, had the three lines at roughly equal volumes for several quarters. And then as house prices began to decline and the market began to crash, it was first-time buyers who were able to come in at a higher rate—incentivized by first-time homebuyer credits in the Housing and Economic Recovery Act, as well as declining interest rates and home prices. But during the slow recovery in originations in the past eight years, the first time and repeat buyer lines have essentially tracked each other, while second home and investor purchases have trailed, only picking up slightly in the past year.

LSE_2021_HDC-mortgage_scally_ch3-01

2020 Vintage Remains Very High Quality

As we describe above, the bulk of newly originated mortgages are going to borrowers with the highest credit score. But is this because of the volume of refinance originations? In the final chart, below, we look at the median credit score of newly originated mortgages, by type. Refinanced mortgages typically go to higher score borrowers than purchase mortgages, due to their established repayment histories. However, when we break out purchasers into first-time and repeat buyers, we see that the median credit scores are quite close for refinancers and repeat buyers, while the first-time buyers have lower credit scores, and always have. The median credit score of refinancers and repeat buyers was just below 800 at the end of 2020, about 60 points higher than that of first-time buyers. With a look to the series history, new mortgages are more prime—for even first-time buyers, median credit scores have slowly drifted up since 2002-06, when they hovered in the high 600s.

LSE_2021_HDC-mortgage_scally_ch4-02

Cha-Ching! Cashing Out Home Equity When the Rates Make Sense

With refinances picking up in 2020, we look to see whether consumers are cashing out, and drawing against their home equity. In the chart below, we show the quarterly cashout refinance volume, which is calculated as the difference between the borrowers’ new mortgages and the mortgages that the refinances replaced. Keeping in mind that the data we present here are not adjusted for inflation, cashout refinance volume is still notably smaller than what was seen between 2003-06. However, we do see a notable increase in cashout refinance volumes, which spiked in the fourth quarter of 2020 and show no sign of abating. Homeowners withdrew $188 billion in home equity over the course of 2020. One important note though—although this is a big increase, it’s also associated with lower average cashout amounts. Borrowers who refinanced in 2019 withdrew, on average, an additional $49,000; while borrowers who refinanced in 2020 withdrew, on average, $27,000. The median cashout withdrawal in 2020 was only $6,700, suggesting that at least half of the refinancers borrowed only enough additional funds to cover the closing costs on the new mortgage. Still, a substantial minority of refinancers took out some cash from their property, which they can use to fund consumption or other investment opportunities, including home improvements. Meanwhile, borrowers who did not choose to take out extra cash saw an average savings of $200 on their monthly mortgage payment, improving their monthly household cash flow.

LSE_2021_HDC-mortgage_scally_ch5-02

What To Look For Going Forward

Support for households who have suffered job and income losses has been a key feature of the CARES Act, through both the direct financial transfers to households as well as the moratoria on Federally backed mortgages and student loans. Meanwhile, many households are taking advantage of the historically low mortgage interest rates that are a consequence of the supportive monetary policy. It will be interesting to see whether households will maintain these high rates of home purchases and refinances into 2021 and more generally how households will adjust their balance sheets depending, in part, on whether and how long forbearances continue on payments on federally backed mortgages and student loans.

Chart data

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

Lee_donghoonDonghoon Lee is an officer in the Federal Reserve Bank of New York’s Research and Statistics Group.

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

Vanderklaauw_wilbertWilbert 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, “Mortgage Rates Decline and (Prime) Households Take Advantage,” Federal Reserve Bank of New York Liberty Street Economics, February 17, 2021, https://libertystreeteconomics.newyorkfed.org/2021/02/mortgage-rates-dec....

Related Reading

Household Debt Balances Increase as Deleveraging Period Concludes

Interactive: Household Debt and Credit Report

CMD: Housing Market

Disclaimer

The views expressed in this post are those of the author 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 author.

Who Received Forbearance Relief?

Published by Anonymous (not verified) on Thu, 05/08/2021 - 3:34am in

Forbearance on debt repayment was a key provision of the CARES Act, legislation intended to combat the widespread economic losses stemming from the COVID-19 pandemic. This pause on required payments for federally guaranteed mortgages and student loans has provided temporary relief to those affected by the COVID-19 pandemic, and servicers of nonfederal loans often provided forbearances or other relief on request as well. Here, using a special survey section fielded with the August 2020 Survey of Consumer Expectations, we aim to understand who benefitted from these provisions. Specifically, were there differences by age, race, income, and educational background? Did individuals who suffered job or income losses benefit differentially? Did renters receive more or less nonhousing debt relief than homeowners? Answers to these questions are not only key for understanding the economic recovery and implications for inequality and equitable growth, they can provide important insight into the expected effects of more recent and potential future legislation.

Background and Data

The CARES Act and its subsequent extensions established a foreclosure moratorium and made borrowers with FHA-, VA- and GSE-backed mortgages with COVID-19 related hardship eligible for a 180-day forbearance period with possible extension to 360 days. Other lenders were also encouraged to work with their borrowers to avoid defaults. In February 2021, moratorium on foreclosures was extended till end of June 2021 and forbearance relief too was further expanded. Specifically, the forbearance enrollment window was extended until June 30, 2021, and homeowners who entered initial forbearance by June 2020 became eligible for an additional six months of forbearance. For renters, the eviction moratorium mandated by the Center for Disease Control and Prevention (CDC) was extended until the end of July 2021.

Under the CARES Act, federal student debt could be deferred until September 30, 2020. This deferral was later extended to January 31, and more recently to the end of September 2021. Moreover, interest on student debt is waived during the forbearance period.

In this post, we leverage data from the August 2020 wave of the Survey of Consumer Expectations (SCE). Since June 2013, this monthly survey has collected information on the economic expectations, choices, and behavior of household heads. The SCE covers about 1,300 nationally representative U.S. households and, in addition to monthly core questions, special modules focusing on specific topics are fielded frequently. As a part of the June and August special modules, we asked respondents about debt repayment, difficulties/successes in obtaining relief, the type of relief received, the reasons for obtaining relief, and their near and medium term expectations.

An earlier two-part Liberty Street Economics (LSE) series leveraging the New York Fed’s Consumer Credit Panel—consisting of detailed Equifax credit report data for an anonymized, nationally representative sample of individuals and households—looked at who could expect to benefit the most from CARES Act debt relief and which borrowers would continue to feel the most strain. This analysis found that mortgage relief was more likely to be concentrated in high income and majority white neighborhoods. In contrast, they found that borrowers who do not have a mortgage or student debt – those that would not benefit from the loan forbearance provisions of the CARES Act – were more concentrated in lower income and majority Hispanic or Black neighborhoods.

A more recent four-part LSE series, using data from the New York Fed’s Consumer Credit Panel, focused on different aspects of mortgage forbearance. The series found that forbearance rates were markedly higher for delinquent borrowers, forborne borrowers saw a sharp decline in delinquencies and credit card balances, and nearly two-thirds of forborne borrowers left forbearance, although lower income and lower credit score borrowers were more likely to linger in forbearance.

In this post, we take an important step forward: the individual-level SCE data enable us to examine who obtained mortgage and student loan forbearances as well as other forms of debt relief, whether the likelihood of getting relief varied with respect to an individual’s socioeconomic characteristics and how they were personally affected by the pandemic. Apart from the above four-part series, other recent analysis has also suggested that forbearance relief has succeeded in stemming a massive increase in delinquencies; here we investigate who the recipients of this forbearance relief were.

Who Received Forbearance Relief?

In the August 2020 SCE, we asked borrowers if they were able to obtain any kind of debt relief, such as a deferment or a reduction of payment or a decline in fees or interest rates. We find that 9.8 percent of renters reported receiving some rental relief (such as deferment, forgiveness, or reduction) while 5.5 percent of households with a mortgage reported receiving relief from their mortgage servicer. For auto loans, we find a higher rate (13.6 percent) of borrowers reporting receiving some assistance from their auto loan/lease servicer. Among credit card users, 11.0 percent reported receiving some assistance from their credit card company (including fee and interest rate reductions and credit card limit increases). Thus, higher shares of auto debt and credit card borrowers reported seeking and receiving some relief compared to mortgagors and renters. As federal direct student loan borrowers were all automatically placed in administrative forbearance, we will focus our attention in this post primarily on debts other than student loans.

In the June 2020 survey, we also asked respondents who requested relief or were planning to do so their reasons for making such a decision. The most prominent response (59 percent) was a “decline in household income,” followed by respondents who reported a desire (39 percent) to shore up their “rainy day funds.” Turning to those who decided not to seek payment relief, we find that by far the most prominent reason cited was that these households were still able to afford payments (95 percent).

Next, we use data from the August 2020 survey to zoom in on the distribution of each kind of relief. For example, were there differences in relief receipt by demographic characteristics (race, age, income, education)? Does receipt vary by how individuals were affected in the labor market?

First, we focus on rent relief. While mortgage forbearances were a part of the CARES Act, rent relief was often offered by landlords on a voluntary basis, although landlords benefiting from mortgage relief were required to provide rent relief to their renters. Focusing on the sample of renters and breaking down rental relief receipt by race, we find in the chart below that similar shares of white and nonwhite renters obtained relief (7.9 percent versus 7.7 percent). We find that the incidence of rental relief was higher among college-educated renters—12.4 percent of college-educated renters received some type of rental assistance while 8.4 percent of noncollege renters did so. Differentiating by income, we find that low income renters were more likely to have received payment relief or assistance than their higher income counterparts. While these patterns for education and income groups seem to conflict with one another, there are several potential reasons behind the apparent divergence. Individuals with higher levels of education tend to be better informed and may be more willing to put in the effort needed to complete the necessary documentation for rent relief. Also, more educated individuals are more likely to own their residences; those who rent may have lower incomes relative to people with similar levels of education. On the other hand, low income individuals may have had more difficulty paying rent and opted to apply for rent relief. Additionally, we differentiate between renters who faced an income loss during the pandemic with those that did not and find that those who sustained income losses were more likely to have received rental relief (12.7 percent versus 8.6 percent). These numbers imply that rent payment relief disproportionately benefited those who lost income or jobs.

Renters Facing Income Loss Were Disproportionately More Likely to Receive Rent Relief

Source: New York Survey of Consumer Expectations, August 2020 Survey.

Next, we turn to the distribution of mortgage payment relief. In the chart below, which considers respondents with a mortgage, we find that nonwhite mortgagors were more likely to receive mortgage relief than white mortgagors (11.1 percent of nonwhite mortgagors received relief versus 4.7 percent of white mortgagors). Additionally, we find that mortgagors who were not college educated were more likely to receive mortgage relief (6.1 percent versus 4.9 percent) and high income mortgagors were more likely to benefit from mortgage relief (5.8 percent versus 2 percent). Differentiating between mortgagors who experienced a drop in income during the pandemic versus those that did not, we find that those that lost income were more likely to benefit from forbearance relief (8.1 percent versus 3.4 percent). This implies that mortgagors who faced higher financial hardship benefited disproportionately more from mortgage relief, similar to the picture above for rental relief. This also matches an earlier finding that borrowers who had higher repayment difficulties benefited more from mortgage forbearance relief.

High Income Mortgagors Were Considerably More Likely to Benefit from Mortgage Forbearance Relief

Source: New York Survey of Consumer Expectations, August 2020 Survey.

Next, we turn our attention to credit card debt and auto debt forbearance relief. Credit card debt and auto debt relief were offered by lenders on a voluntary basis. The chart below looks at the distribution of credit card debt relief among credit card debtors. We find that nonwhite credit card holders were more likely to receive credit card debt relief than white credit card holders—19.7 percent of nonwhite credit card holders reported receiving some credit card debt assistance compared to 9.7 percent of white credit card holders. Looking at heterogeneity by education, we find that a slightly lower percentage of college educated credit card holders received credit card debt relief than those that did not have college education. Differentiating by labor market outcomes, we find that credit card holders who faced an income loss during the pandemic were markedly more likely to receive relief (14.8 percent versus 8.9 percent). Additionally, we find that credit card holders who were renters were more likely to receive relief than homeowners (13.1 percent versus 10 percent).

Nonwhite, Low-Income, and Renter Credit Card Holders Were Substantially More Likely to Receive Credit Card Debt Relief

Source: New York Survey of Consumer Expectations, August 2020 Survey.

Finally, we turn to auto loan debt holders. The chart below shows the distribution of auto debt payment relief among auto debt holders. We find that nonwhite, low income, and less educated auto loan debtors were more likely to have received auto debt relief compared, respectively, to white, higher income, and more highly educated auto loan debtors, a finding that mimics our findings for credit card debt relief. As with credit card debt relief, we additionally find that auto debt holders who are renters were markedly more likely to receive relief, as were auto debt holders who experienced a decline in income during the pandemic.

Nonwhite, Low-Income, and Renter Auto Debt Holders Markedly More Likely to Receive Auto Debt Relief

Source: New York Survey of Consumer Expectations, August 2020 Survey.

Discussion and Conclusion

In this post, we investigated whether forbearance relief was equally distributed across the population.  We uncover marked differences in the distribution of forbearance across demographic groupings and across individuals who experienced different labor market outcomes. Consistent with our earlier expectations, we find that mortgage debt relief was disproportionately received by high income households. The picture is different in the rental, credit card, and auto debt markets, where low income renters/debtors reported a higher incidence of the corresponding relief. Differentiating by race and education, we find that nonwhite and less educated households with a certain kind of debt (mortgage, credit card, auto) were more likely to receive relief on payments for that kind of debt. This contrasts with our findings in the rental market, where white and more educated renters reported a comparable or higher probability of rental assistance.

Across all markets, households that lost income and thus faced financial hardship were more likely to receive forbearance relief than other households. The distributions of forbearance relief receipt by income, education, and race are more uneven across markets. The above analysis provides insights into how the recent extension of relief can affect different populations and can serve as guides to policy. It remains to be seen how the different groups will fare when these relief measures expire at the end of summer/early fall. We will continue to closely monitor the situations of consumers in various markets as economic and policy conditions evolve.

Chart data

Rajashri Chakrabarti is a senior economist in the Federal Reserve Bank of New York’s Research and Statistics Group.

Jessica Lu is a senior research analyst in the Bank’s Research and Statistics Group.

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

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

How to cite this post:

Rajashri Chakrabarti, Jessica Lu, Joelle Scally, and Wilbert van der Klaauw, “Who Received Forbearance Relief?,” Federal Reserve Bank of New York Liberty Street Economics, August 2, 2021, https://libertystreeteconomics.newyorkfed.org/2021/08/who-received-forbe....

Related Reading

Keeping Borrowers Current in a Pandemic (May 2021)

What Happens during Mortgage Forbearance? (May 2021)

Debt Relief and the CARES Act: Which Borrowers Benefit the Most? (August 2020)

Disclaimer

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.

Consumer Credit Demand, Supply, and Unmet Need during the Pandemic

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

Tags 

Credit, mortgages

Jessica Lu and Wilbert van der Klaauw

Consumer Credit Demand, Supply, and Unmet Need during the Pandemic

It is common during recessions to observe significant slowdowns in credit flows to consumers. It is more difficult to establish how much of these declines are the consequence of a decrease in credit demand versus a tightening in supply. In this post, we draw on survey data to examine how consumer credit demand and supply have changed since the start of the COVID-19 pandemic. The evidence reveals a clear initial decline and recent rebound in consumer credit demand. We also observe a modest but persistent tightening in credit supply during the pandemic, especially for credit cards. Mortgage refinance applications are the main exception to this general pattern, showing a steep increase in demand and some easing in availability. Despite tightened standards, we find no evidence of a meaningful increase in unmet credit need.

Demand and Supply of Household Credit during a Recession

How would one generally expect credit flows to consumers to change during a recession? To the extent that households perceive declines in income and wealth as permanent, they may lower current consumption and demand for debt. Similarly, an increase in uncertainty can depress consumption and increase precautionary or “buffer-stock” saving, reducing demand for credit. On the other hand, during recessions households may optimally respond to a temporary negative income shock by drawing on new and existing credit in order to smooth consumption.

On the supply side, access to credit may become more difficult during recessions due to a rise in lender risk aversion or a reduction in collateral values. Lenders may be concerned that a recession and higher unemployment could trigger a sharp rise in loan defaults, and this concern may cause them to tighten credit standards. With lenders requiring higher credit scores, more income documentation, and reduced credit limits on new and existing loans, households may find it harder to obtain credit when they most need it.

While we expect these factors to play similar roles in the current recession, the pandemic and the public health and policy responses resulted in an economic downturn with different characteristics and dynamics than prior recessions. The recession was unique in the speed with which it took hold, the types of workers it affected, and the nature and size of the policy response. In an economy with reduced consumption opportunities (especially services) due to lockdowns, a huge decline in employment was accompanied by a steep decline in spending, but also a jump in the personal savings rate to a record high, as personal income surged in response to policy measures.

Using the SCE to Measure Demand for Credit

Our analysis draws on data from the Survey of Consumer Expectations (SCE) Credit Access Survey. The survey has queried respondents every four months since October 2013 about their experiences and expectations of applying for and obtaining credit. Our data differ from existing data sources in several important ways. First, unlike existing sources such as the Senior Loan Officer Opinion Survey (SLOOS), our survey captures realized and expected credit demand as reported by consumers instead of lenders. Second, we measure credit applications and rejections separately for different types of consumer loans, including mortgages, auto loans, and credit cards. Third, besides realized and expected credit demand, we identify “discouraged credit seekers”—respondents who report not applying for credit over the past twelve months despite needing it, because they believed they wouldn’t be approved.

In this post, we introduce a new, more forward-looking gauge of unmet need based on a new measure of latent credit demand. Specifically, starting in October 2016 we began asking respondents for the probability that they will apply for credit over the next twelve months if their application is guaranteed to be approved. Our new measure of unmet need is then defined as the difference between their answer to this question and their reported unconditional probability of applying over the next twelve months.

Application and Rejection Rates

The overall share of respondents who reported applying for any type of credit over the past year shows a v-shaped pattern after the onset of the pandemic, as shown in the chart below. After the share of those who reported applying fell from 45.6 percent in February 2020 to 39.2 percent in June and to a series low of 34.6 percent in October, we see a robust rebound in February 2021 to an application rate of 44.8 percent, just below its pre-pandemic value. We see a similar pattern for credit card application rates, which fell from 26.3 percent in February 2020 to 15.7 percent in October, though with a smaller rebound in February 2021 to 19.4 percent, still well below pre-pandemic levels. Requests for credit card limit increases (not shown) evolved comparably. Auto loan and mortgage application rates (not shown) were relatively more stable and in February 2021 were just above their year-ago levels.

The big outlier in these reported trends is a dramatic increase among mortgagors in mortgage refinance applications during the pandemic, with many mortgage loan borrowers taking advantage of low interest rates. In February 2021, 24.6 percent of mortgagors reported applying for a refinance over the past twelve months, compared to only 10.8 percent a year earlier.


Consumer Credit Demand, Supply, and Unmet Need during the Pandemic

Among those who applied for any type of credit, the share reporting a rejection has increased monotonically since the onset of the pandemic, from 14.2 percent in February 2020 to 18.5 percent in February 2021. As shown in the next chart, the increase was relatively modest and within the range of changes observed in our series’ history. The rejection rate for credit card applicants shows a somewhat more noticeable increase, rising from a series low of 9.7 percent in February 2020 to a series high of 26.3 percent a year later—a 170 percent increase. Rejections of requests for credit limit increases (not shown) follow the same trajectory—reaching a new high of 40.3 percent in February 2021. The reported rejection rates for auto loan and mortgage applicants (not shown) follow less of a clear pattern over the past year.

Rejections reported by mortgage refinance applicants initially declined, but then rebounded somewhat in February 2021 to 12.2 percent, remaining below the 15.8 percent reading a year earlier. The steep increase in demand and some easing in available credit is consistent with the general strength of the housing market and historically high levels of home equity. These findings from our survey are largely consistent with indicators of credit demand and credit tightening reported by loan officers in the SLOOS.


Consumer Credit Demand, Supply, and Unmet Need during the Pandemic

When distinguishing respondents by age and credit score, we find that the patterns in application and rejection rates described above are largely broad-based across age and credit score groups, with two exceptions. First, the February 2021 rebound in application rates is generally much larger for respondents with self-reported credit scores between 680 and 760, aside from mortgage refinances, where the largest increase is among those with scores of 760 or higher. Second, the increase in the reported overall rejection rate during the pandemic was by far the largest for those with low credit scores (680 or below).

Two New Measures of Unmet Credit Demand

Turning now to our two alternative measures of latent credit demand (or unmet need), we first consider our backward-looking measure (shown in the left panel below). The share of discouraged borrowers, which we define as respondents who reported not applying for a loan (despite having a need for it) because they expected their application to be rejected, initially remained stable at 6.9 percent from February to June of 2020, rose slightly to 7.2 percent in October, and then fell to 6.1 percent in February 2021. Interestingly, this overall pattern masks a decline in borrower discouragement among those aged 40 and under and 60 and over, with the rate for the middle-age group increasing somewhat from 6.1 percent in February 2020 to 9.1 percent a year later. Similarly, we see an increase in discouraged borrowing among those with credit scores of 680 and below (from 21.6 percent to 28.1 percent), while seeing a slight decrease for those with higher scores.


Consumer Credit Demand, Supply, and Unmet Need during the Pandemic

Finally, we turn to our second alternative measure of latent credit demand, our forward-looking measure, available since October 2016. We first find that expected credit applications, both unconditional and conditional upon guaranteed acceptance, declined following the onset of the pandemic and more recently have experienced a considerable rebound. The initial decline in expected demand, even under guaranteed acceptance, substantiates our earlier finding of a general decrease in credit demand during the first eight months of the pandemic. Considering the difference between the two expected application rates—our indicator of unmet credit need—we find, as expected, that the gap is persistently positive, indicating a lower unconditional expected application rate due to a non-negative probability that the application will be rejected.

The panel on right above shows a simple unweighted average of the gap for mortgage, credit card, auto loan, credit limits, and mortgage refinance applications. After the onset of the pandemic, there is a slight decline in unmet credit need across all loan types, with the important exception of mortgage refinance applications. In contrast to our backward-looking measure of unmet credit (which does not distinguish between loan types), here for most loan types we find the decline in unmet need to be the largest for those with low credit scores (under 680). For mortgage refinancing (not shown), we see instead a slight rise among mortgagors in unmet need from 6.1 percent in February 2020 to 7.3 percent in June and 9.1 percent in October. However, by February 2021 the gap had narrowed to 5.5 percent. As was the case for the backward-looking measure, the increase in unmet need for refinancing is the largest for those with lower credit scores (below 680).

Conclusion

In this post, we used new data from the SCE Credit Access Survey to study the evolution of consumer credit demand and supply during the COVID-19 recession. We find that credit demand dropped sharply during most of 2020, especially for credit cards, with a modest rebound observed by February 2021. We also observe a modest increase in credit rejection rates during 2020, especially for credit cards. Despite this credit tightening and increased unemployment, we see no meaningful increase in discouraged borrowing and unmet credit need. Mortgage refinancing is the main exception to this general pattern, showing a steep increase in demand and some increase in unmet need, especially for those with lower credit scores.

These overall results likely evince the impact of large fiscal interventions, including stimulus checks and expanded unemployment insurance benefits, which have enabled many households to pay down debt, especially credit card debt, and increase their saving. Even among those with low credit scores, while unmet credit needs remain formidable, instead of an increase in expected unmet need we actually see a decline for the group for which unmet need is typically high. The only clear evidence of an increase in unmet credit needs is found for mortgage refinancing, where lower credit score mortgagors have been less able to take advantage of the low-rate environment.

Jessica Lu is a senior research analyst in the Federal Reserve Bank of New York’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:

Jessica Lu and Wilbert van der Klaauw, “Consumer Credit Demand, Supply, and Unmet Need during the Pandemic,” Federal Reserve Bank of New York Liberty Street Economics, May 20, 2021, https://libertystreeteconomics.newyorkfed.org/2021/05/consumer-credit-de....

Disclaimer

The views expressed in this post are those of the author 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 author.

What Happens during Mortgage Forbearance?

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

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

What Happens during Mortgage Forbearance?

As we discussed in our previous post, millions of mortgage borrowers have entered forbearance since the beginning of the pandemic, and more than 2 million remain in a program as of March 2021. In this post, we use our Consumer Credit Panel (CCP) data to examine borrower behavior while in forbearance. The credit bureau data are ideal for this purpose because they allow us to follow borrowers over time, and to connect developments on the mortgage with those on other credit products. We find that forbearance results in reduced mortgage delinquencies and is associated with increased paydown of other debts, suggesting that these programs have significantly improved the financial positions of the borrowers who received them.

Forbearance and Mortgage Delinquency

Since March 2020, we have observed more than 6.1 million mortgagors enter forbearance. As noted in our previous post, these forbearance participants were much more likely to be delinquent prior to the pandemic than the general population of mortgagors. One of the benefits of forbearance for these previously delinquent borrowers is that commencement of forbearance is often coincident with a “cure”: a change in mortgage status to “current.” That is, for many borrowers, mortgage delinquencies are wiped away as the borrower enters forbearance, at least temporarily. (These status changes come without evidence of payment, supporting the conclusion that the cure is the result of an administrative change rather than a true cure. Mortgage servicer reports to investors, as opposed to credit bureaus, show these loans as delinquent. Importantly, however, these investor reports do not affect borrower credit histories.)

The first chart below shows the credit bureau reporting of mortgage status for those that entered forbearance by May 2020. Around 8 percent of the mortgages were already delinquent before entering forbearance. A great majority of those accounts that were previously delinquent are reported as “current” while in forbearance, some of them by making payments and some without one. A minority—about 30 percent of the previously delinquent accounts—retain this delinquent status throughout the period. These varying treatments upon entry into forbearance seem to depend on servicer practices. As such, current foreclosure data and delinquency statistics drawn from credit bureau data do not accurately give a clear indication of housing market stresses.


What Happens during Mortgage Forbearance?

At the same time, neither does the rate of forbearance itself. Why? Because a large share of mortgagors in forbearance actually continue to make their monthly mortgage payments. Indeed, the share of borrowers who continue to make payments while in forbearance is surprisingly high: in each month since June 2020, between 30 and 40 percent of the borrowers in forbearance have made their monthly payment.

This behavior suggests that some borrowers have taken advantage of the forbearance program and skipped payments while others have applied for forbearance as an “insurance policy” against which they are not making claims, and they are reducing their balances each month as originally anticipated in the mortgage contract.

But for the 60-70 percent of forbearance borrowers who are not making payments, mortgage balances aren’t falling. In 2019, mortgagors paid off approximately 4 percent of mortgage balances by making their regular payments. By contrast, borrowers in forbearance have seen their balances increase by 1-2 percent over the course of the last year as the automatic amortization that comes from making the mortgage payment has been largely absent and the interest component of the skipped payment is added back to the balance as well. As of March 2021, among the 5 million borrowers who have taken forbearance for at least one month since the pandemic and haven’t prepaid, about 26 percent have a higher mortgage balance than a year earlier.

Mortgage Forbearance and Performance on Other Household Debts

We can also use the CCP to examine the relationship between mortgage forbearance and performance on a borrower’s non-housing debts. Doing so, though, requires a slightly longer timeframe. In the chart below, we show that non-mortgage delinquency (which reflects delinquency on auto, credit card, and miscellaneous consumer debt) was persistently higher among those who had at least one month of forbearance since March 2020; indeed, prior to the pandemic this was a group of borrowers whose delinquency rates had not only been high, they had also been on the rise. (We keep student debt out of consideration here since the vast majority of student debt has been in automatic forbearance since the early weeks of the pandemic.) Immediately after March 2020, delinquency on non-housing debts leveled off briefly, but then began increasing again and stood at 5.8 percent in March 2021, a full percentage point higher than it had been one year before. In contrast, delinquency rates for those not in mortgage forbearance were roughly flat during the year ending in March 2021, at about 2 percent.

Thus we have a glass half empty/half full situation: these are clearly distressed borrowers, and mortgage forbearance provided assistance that may well have allowed them to keep their homes. Nonetheless, these borrowers were already struggling with debt repayment prior to the pandemic, and forbearance has not allowed them to close the delinquency gap with other mortgagors; instead that gap has persisted in spite of forbearance.


What Happens during Mortgage Forbearance?

A second dimension of performance, and one that is perhaps especially interesting during the pandemic environment of reduced consumption opportunities, is debt balance paydown. We’ve noted in the past that aggregate credit card balances fell a lot in 2020, and ended the year more than $100 billion below their December 2019 level. This is the largest annual decline in credit card balances for at least two decades, and it continued into the first part of 2021. The accumulation of savings by U.S. households during the pandemic was surely a key factor in this paydown of costly credit card balances. Did mortgage forbearance play a role for those households that received it?

In the next chart, we provide some evidence for that proposition. The chart shows the relative credit card balances for mortgagors who had a forbearance after March 2020 (red) and those who never did (blue). Card balances declined for both groups, but somewhat more steadily for borrowers with forbearances: by March 2021, they had reduced their credit card balances to 23 percent below their March 2020 level. This compares with a 15 percent decline for mortgagors without a forbearance. The dollar amount of credit card paydown is even higher for those with forbearance, since their initial average amount of credit card debt as of March 2020 was significantly higher at $9,000 compared to $6,000 for those without forbearance. As a result, a typical household in mortgage forbearance has reduced its credit card debt by $2,100 over the last year, compared to $900 for a mortgagor not in forbearance.


What Happens during Mortgage Forbearance?

The ability to reduce credit card obligations over the past year has not been equal across different types of mortgage borrowers in forbearance. The next chart shows that the balance decline for neighborhoods outside of the top income quartile has now reached 20 percent below the March 2020 level. In the highest income neighborhoods, which benefited from the largest share of mortgage relief as shown in the previous blog post, credit card balances have fallen more: 30 percent as of March.


What Happens during Mortgage Forbearance?

Conclusion

Our brief review of what happens to borrowers when they’re in forbearance produces some interesting conclusions. First, many previously delinquent borrowers are marked “current” as they enter forbearance, even if they don’t make a payment. As a consequence, credit bureau measures of mortgage delinquency must be viewed cautiously in a period of widespread forbearance. Second, a substantial share, around 30-40 percent, of borrowers who get forbearance nonetheless continue to make payments. This will have implications for our expectations for how delinquency measures will change when forbearance ends. Finally, mortgagors in forbearance have been able to pay down their credit cards faster than those not in forbearance, especially in higher income areas. In our next post, we will shift our focus to a group of mortgage borrowers who stand out from the crowd for a different reason: they own a small business.

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 Haughwout, “What Happens during Mortgage Forbearance?” Federal Reserve Bank of New York Liberty Street Economics, May 19, 2021, https://libertystreeteconomics.newyorkfed.org/2021/05/what-happens-durin....

Additional Posts in This Series

Keeping Borrowers Current in a Pandemic

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

Disclaimer

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.

Conclusion

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

Disclaimer

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.

Pages