Residential Mortgage Loans Research Articles

Correction: Wong, M.C.S.; Ho, H.M. A Framework for Integrating Extreme Weather Risk, Probability of Default, and Loss Given Default for Residential Mortgage Loans. Sustainability 2023, 15, 11808

The authors would like to make the following corrections about the published paper [...]

A Framework for Integrating Extreme Weather Risk, Probability of Default, and Loss Given Default for Residential Mortgage Loans

This paper considers a hypothetical case in which a bank wants to build a routine climate stress test exercise on residential mortgage loans. The bank has regularly updated the probability of default (PD) and loss given default (LGD) on each residential mortgage loan under the internal-rating-based (IRB) approach of Basel II/III. Additionally, the bank estimates the stressed PD and stressed LGD associated with a predetermined extreme weather event. Using simulation techniques, this paper shows that the loss of the bank’s residential mortgage portfolio can reach a median of around 36% of the portfolio value. This remarkable loss comes from the effects of default correlation and property damage. Banks should pay more attention to such impacts of extreme weather events.

Labor Mobility and Loan Origination

AbstractWe find that mortgage loans originated after the adoption of the inevitable disclosure doctrine (IDD; a mechanism discouraging loan officers’ labor mobility) have a lower default probability, a higher loan modification rate, and a lower foreclosure rate. These effects are unaccompanied by any reduction in loan supply and contribute to more stable housing prices. Using the adoption of the Uniform Trade Secrets Act as an alternative identification generates consistent results. Overall, our findings suggest that restricting loan officers’ labor mobility leads to better ex ante screening and ex post monitoring, improving the origination efficiency for U.S. residential mortgage loans.

Cooling the mortgage loan market: The effect of borrower-based limits on new mortgage lending

This paper studies the effects of regulatory measures concerning maximum loan-to-value (LTV), debt-to-income (DTI), and debt service-to-income ratios (DSTI) on new loans secured by residential property. It uses loan-level regulatory survey data on about 82,000 newly granted residential mortgage loans in Czechia from 2016 to 2019 to estimate the average effects of regulatory measures and their heterogeneous effects depending on borrower, loan, bank, and regional characteristics. We find that the LTV limits decreased the mortgage loan size only slightly, while the value of collateral securing the loans increased. Only the additional DTI and DSTI limits helped reduce the average loan size significantly and prompted more risk-sensitive pricing that heighten the average lending rate. The mortgage market response to the limits appears heterogeneous in loan size rather than pricing or collateralization. The loan size reduction varied with the borrower’s income, age, and location. High-income, high-repayment, and 25 to 35 years old borrowers reduced their contracted loan size more. Borrowers located in regions with higher GDP per capita and lower unemployment rates reduced the contracted loan size more. Longer-maturity loans showed a greater drop in the contracted loan size.

Systemic risk contribution of banks and non-bank financial institutions across frequencies: The Australian experience

The Australian financial sector (AFS) is highly concentrated and interconnected. Besides, Australian banks' lending portfolios are dominated by residential mortgage loans, and 70% of insurance companies' revenues arise from non-policyholder sources. The AFS also performed relatively well during the global financial crisis (GFC). Given these distinctive features, in this paper, we examine the systemic risk contribution of Australian banks, insurance companies, and other financial services providers. We use a flexible copula-based delta conditional value-at-risk (ΔCoVaR) method across different frequencies. Further, we study the systemic risk determinants in a panel setting. We find that the major Australian banks are systemically more important than all other financial institutions. Systemic risk is typically higher after the GFC than in the pre-crisis period, despite the introduction of more stringent capital requirements. In addition, the short-term ΔCoVaR is significantly higher than the medium- and long-term ΔCoVaRs. Finally, institution-specific characteristics and market-wide variables explain the cross-sectional and time-series variation in systemic risk, and their explanatory power varies across frequencies.

COVID-19 Mortgage Forbearances: Drivers and Payment Behavior

This article examines mortgage forbearance in Freddie Mac loans during the COVID-19 crisis and compares this period to two previous episodes: a baseline and a natural disasters sample. We find that the rates of mortgage forbearance in the COVID-19 period are similar to those in the natural disaster sample but are much higher than in the baseline sample. In all three periods, we find higher forbearance rates are associated with lower FICO scores, higher DTI ratios, and higher LTV ratios. In addition, higher levels of the cumulative cases of COVID-19 are associated with higher forbearance rates. We also examine the payment behavior of borrowers who entered forbearance during the COVID-19 crisis. We show that most mortgages in forbearance missed at least one payment within the first three months or were delinquent when entering forbearance. However, by the seventh month after entry into forbearance, about half of the mortgages are current, whereas about 40.7% are at least 90 days delinquent. <b>TOPICS:</b>MBS and residential mortgage loans, financial crises and financial market history, legal/regulatory/public policy <b>Key Findings</b> ▪ Mortgage forbearance rates in the COVID-19 crisis are similar to historic rates in natural disaster areas but are much higher than in the months just prior to the crisis. ▪ Higher forbearance rates are associated with lower FICO scores, higher DTI ratios, higher LTV ratios, higher cumulative COVID-19 rates, larger mortgage payments, two- to four-unit and investor properties, single borrowers, and refinances. ▪ By the seventh month in forbearance, about half of the loans have transitioned to current status, but a similar share of loans remains delinquent.

The Lowdown on Low-Down-Payment Mortgages: Is It Safe?

Low down payment mortgages were prevalent before the financial crisis, slowed down during the financial crisis, but have reemerged since the financial crisis. The Federal Housing Administration has a history of insuring 3% down payment mortgages to help lower- and middle-income households purchase a home, but Fannie Mae and Freddie Mac recently have introduced 3% (and lower) down payment programs. In addition, mortgage innovation in the form of the “wealth building mortgage” now adds the possibility of a 100% loan-to-value (LTV) ratio. In this paper, we examine the performance of low down payment (or high LTV) lending. We find that low initial down payment mortgages are significantly more likely to become seriously delinquent (90 days or more) than mortgages with traditional down payments of 20% or more. TOPICS:Fixed income and structured finance, MBS and residential mortgage loans, risk management, credit risk management, legal/regulatory/public policy, financial crises and financial market history Key Findings ▪ This study finds meaningfully accurate estimates of the likelihood of more severe delinquency payment status changes by using individual mortgage loan, borrower, and county-level risk factors as explanatory variables. ▪ Low initial down payment mortgages are significantly more likely to become seriously delinquent (90 days or more) than mortgages with traditional downpayments of 20% or more. ▪ Initial borrower and mortgage characteristics, such as LTV, borrower documentation, or FICO score, are dominant factors in the likelihood of delinquency and default.

Rebuilding the Non-Agency Market: Why Modernizing Data and Reporting Is Critical

Twelve years have passed since the US mortgage and housing bubble served as a primary catalyst that plunged world economies into a global financial crisis. The bubble was fueled by the usual culprits: fraud, excessive risk, and lax standards, particularly in the non-agency mortgage universe. Underlying all of these factors was a systemic lack of transparency and quality reporting that prevented these issues from being reconciled and resolved. Since then, legislation has been imposed to regulate the mortgage industry, protect consumers from predatory lending practice, and rebuild the economy. Similarly, issuers have made substantial underwriting changes to prevent prior lending excesses. The improvements made, however, have not addressed the core issue, and legacy systems for accessing, managing, and analyzing loan data have remained largely untouched. Data are the basis of safe and critical decision making, but it continues to be inaccessible, costly, and unvalidated. To restore investor confidence, especially as COVID-19 proves to be a destructive force, the industry requires modernization; otherwise, it may never truly recover. <b>TOPICS:</b>Asset-backed securities (ABS), CLOs, CDOs, and other structured credit, MBS and residential mortgage loans <b>Key Findings</b> ▪ While substantial improvements have been made to underwriting and borrower verification, antiquated legacy systems and inefficient data-reporting processes remain fully intact, which stifles innovation, hinders transparency, and jeopardizes accountability. ▪ The legacy systems’ lack of innovation proved incapable of quickly adapting to COVID-19-related performance reporting or increasing the availability of new mortgage data, which contributes to the sector’s slow recovery in volume and investor participation. ▪ A broader investor base is critical in propelling the non-agency mortgage market back into prominence. To attract new entrants and welcome previously scarred participants, investor confidence must be restored through data integrity and modernizing sector infrastructure.

Measuring Risk and Access to Mortgage Credit with New Disclosure Data

Female and minority borrowers have historically been denied equal access to credit markets. Yet conclusive evidence of discrimination is often hindered by limited availability of complete underwriting information. The recent expansion of data collected under the Home Mortgage Disclosure Act provides an unprecedented opportunity to examine disparities in access to mortgage credit in greater detail. The average female or minority applicant is associated with higher risk characteristics than the typical White male mortgage applicant. This greater risk explains some of but not all the differences in denial rates between groups. Statistically significant disparities persist after accounting for the predicted probability of loss. For example, Black applicants have a denial rate roughly 3 percentage points higher than White applicants with similar characteristics. <b>TOPICS:</b>MBS and residential mortgage loans, information providers/credit ratings, credit risk management <b>Key Findings</b> ▪ Female, minority, and same-sex applicants as groups are associated with higher than average predicted likelihood of loss. ▪ Female, minority, and same-sex applicants are more likely to be denied given their levels of predicted risk. ▪ Comparing default rates instead of denial rates will underestimate the prevalence of mortgage discrimination.

Residential Mortgage Default Risk and Market Returns of Public Mortgage Originator Firms in Kenya

The mortgage market plays a vital role in the development of the real estate sector. The mortgage industry in Kenya has experienced tremendous growth since the year 2000. Despite this growth, Kenya’s mortgage debt to GDP ratio is still relatively low when compared to other economies like South Africa. Default risk has been revealed as one of the risks that significantly impacts on the profitability of mortgagees. However, literature is inconclusive with reference to the relationship between default risk and the market returns of mortgage firms. Consequently, this study sought to determine the extent to which residential mortgage default risk influences the market returns of publicly listed mortgage firms in Kenya. Default risk in this case was measured using the non-performing loans ratio: the ratio of non-performing residential mortgage loans to total residential mortgage loans and advances. The study adopted descriptive and quantitative forms of research design. A census was conducted on the eleven NSE listed mortgage originating firms. A panel data regression model was utilized to draw inference from the secondary data collected. Descriptive statistical findings revealed a mean of 0.0796 with a standard deviation of 0.04219 for residential mortgage default risk. Inferential statistics revealed an R square value of 0.2794 between residential mortgage default risk and market returns of publicly listed mortgage originators. In addition, there was significant effect between default risk and the market returns of public mortgage originators. Consequently, mortgagees should develop strategies of reducing nonperforming loans. For instance, mortgage firms can improve their credit rating systems.

House Price Dynamics and Bank Herding: European Empirical Evidence

This paper examines house price dynamics, bank herding behavior, and the linkages between them. The analysis presented indicates that prior to the financial crisis, non-fundamental factors played a significant role in several European countries, including the United Kingdom, Spain, Denmark, Sweden, and Ireland. We also provide evidence indicative of herding behavior in the residential mortgage loan market. Finally, Granger causality tests show that non-fundamentally justified price dynamics contributed to the herding displayed by lenders and that this behavior was a response by banks as a group to common information on residential property assets.

The Impact of Data Aggregation and Risk Attributes on Stress Testing Models of Mortgage Default

Stress testing models have been developed at various levels of data aggregation with or without risk attributes, but there is limited research on the joint impact of these modeling choices. In this paper, we investigate how data aggregation and risk attributes affect the development and performance of stress testing models by studying residential mortgage loan defaults. We develop mortgage default models at various data aggregation levels including loan-level, segment-level, and top-down. We also compare the models with and without risk attributes as control variables. We assess model performance for goodness-of-fit, prediction accuracy, and projection sensitivity for stress testing purposes. We find that the loan-level models do not always win among models with various data aggregation levels, and including risk attributes greatly improves goodness-of-fit and projection accuracy for models of all data aggregation levels. The findings suggest that it is important to consider data aggregation and risk attributes in developing stress testing models.

The impact of data aggregation and risk attributes on stress testing models of mortgage default

Stress testing models have been developed at various levels of data aggregation with or without risk attributes, but there is limited research on the joint impact of these modeling choices. In this paper, we investigate how data aggregation and risk attributes affect the development and performance of stress testing models by studying residential mortgage loan defaults. We develop mortgage default models at various data aggregation levels including loan-level, segment-level, and top-down. We also compare the models with and without risk attributes as control variables. We assess model performance for goodness-of-fit, prediction accuracy, and projection sensitivity for stress testing purposes. We find that the loan-level models do not always win among models with various data aggregation levels, and including risk attributes greatly improves goodness-of-fit and projection accuracy for models of all data aggregation levels. The findings suggest that it is important to consider data aggregation and risk attributes in developing stress testing models.

US House Price Projections from the Economic Impact of the Coronavirus

Could the coronavirus-induced economic shocks hurt US house prices as much as the global financial crisis (GFC) did? This article identifies four drivers that could produce much milder house price depreciation (HPD) this time: early and broad mortgage- and housing-relief policy responses, the apparent lack of a preceding house price bubble, historically moderate mortgage- and consumer-debt levels, and lower mortgage rates. The authors apply a previously published time series–based house price model, combined with these exogenous drivers. Fitted to historical data between 1987 and 2020, the model estimates a 1.66 transmission rate from GDP underperformance to HPA underperformance, a ratio of 0.3 between HPA and a mortgage rate relative percentage rally. If mortgage rates rally from 3.7% in 2019 to 2.7% and the GDP takes a hit of 11% below-trend growth, then the model estimates roughly 10% below-trend growth for national HPA, which is much milder than the experience during the GFC. <b>TOPICS:</b>Financial crises and financial market history, MBS and residential mortgage loans <b>Key Findings</b> • National house price depreciation due to the coronavirus may be much milder than the previous crisis, due to effective mortgage and housing relief policy, lack of a preceding house price bubble and mortgage debt bubble, and lower mortgage rates. • Our house price model combines a time-series property with these exogenous drivers and estimates the sensitivity to each driver. • Our model projects 10% below-trend growth for national house price appreciation (HPA) if the US GDP takes a hit of 11% below-trend growth, supported with a 1% mortgage rate rally.

Drivers of Mortgage Spreads across the Business and Housing Cycle

This article provides a review of the underlying drivers of mortgage spreads at various points of the housing and economic cycle, with an in-depth look at the spreads between mortgage-backed securities (MBS) and Treasury rates and between mortgage and MBS rates. The authors provide evidence of the role of prepayment and credit-related risks, as well as lender capacity issues and large-scale asset purchases by the Federal Reserve in explaining these spreads. They also investigate recently observed differences in rates across purchase and refis as additional evidence of the role of capacity constraints in the persistent widened spreads during the ongoing COVID-19 outbreak. TOPICS:MBS and residential mortgage loans, real estate, legal/regulatory/public policy Key Findings • Mortgage spreads, especially the primary–secondary spread (mortgage rate versus current coupon yield on MBS), remain historically wide following the COVID outbreak in the US. • Increases in liquidity in the mortgage-backed securities (MBS) market coinciding with the Federal Reserve’s increased MBS purchases have historically been correlated with declines in the secondary spread (current coupon yield on MBS versus Treasuries), consistent with the post-COVID surge in Fed MBS purchases and subsequent decline in the secondary spread. • Increases in the primary–secondary spread have historically been correlated with increased capacity constraints faced by lenders, for instance, measured by increases in production per mortgage-related employee. The persistent widened primary–secondary spread in the aftermath of COVID-19 is consistent with this historical pattern.

But This Time IS Different—COVID Recession

Earlier this year, the US economy dropped from historical highs to a severe downturn almost overnight due to the government’s response to the coronavirus outbreak. The pandemic affected all sectors of the economy, including the MBS market. Historically, there has been a direct link between an economic downturn and the fundamentals of the financial markets leading up to the downturn. As we saw in the Great Recession, the exuberance in the mortgage market, poor mortgage loan underwriting quality, and the exponential leverage resulting from the derivative markets contributed to financial disaster. The recent economic downturn resulted not from overleverage and the kind of ‘wild west’ behavior observed in the early to mid-2000s, but from the deliberate shut down of the US economy in an effort to mitigate the spread of the coronavirus. Unlike the Great Recession and the Great Depression, the events leading to the current economic downturn were exogenous to the market. As a result, the economy and the MBS market rebounded quickly. TOPICS:MBS and residential mortgage loans, financial crises and financial market history Key Findings • The forced shutdown of the economy resulted in an economic downturn paralleling the Great Depression. • Delinquency rates on residential mortgages and forbearance requests rose quickly, with a knock-on effect on MBS pricing and prepayment speeds. • Due to the strength of the economy and the MBS market pre-COVID, the time to recovery has been shorter than in previous downturns.

The Pandemic’s Effect on the Housing Market and Mortgage Delinquency

Weekly data show how housing activity contracted as the pandemic expanded. Both the home purchase and home rental markets experienced a 40% to 50% decline by mid-April relative to year-ago activity, but subsequently rebounded. Record low mortgage rates spurred home buying, driven by first-time home buyers, and a refinance boom. Home prices rose through summer 2020 as the for-sale inventory remained in short supply, but price growth is forecast to slow and fall in several states by 2021. Projections show that elevated unemployment coupled with declining home prices would cause serious delinquency and distressed sales to rise. Compared with the pre-pandemic period, the Baseline projection shows a four-fold increase in serious delinquency and distressed sales within two years, with more than 2 million home mortgages in serious delinquency by year-end 2021. TOPICS:Real estate, MBS and residential mortgage loans, financial crises and financial market history Key Findings • Both the home purchase and home rental markets experienced a 40% to 50% decline by mid-April relative to year-ago activity, but subsequently rebounded. • Projections show that elevated unemployment coupled with declining home prices would cause serious delinquency and distressed sales to rise. • Compared with the pre-pandemic period, the Baseline projection shows a four-fold increase in serious delinquency and mortgage loss events within two years, with more than 2 million home mortgages in serious delinquency by year-end 2021.

COVID-19 Highlights the Need for Servicer Access to a Government-Backed Liquidity Facility

The fundamental differences between the structure of the market for government mortgages, pooled into Ginnie Mae securities, and the market for government sponsored enterprises (GSE) mortgages are at the heart of the constraints that impact the servicing of government loans. This is exacerbated by the higher-risk profile of Ginnie Mae borrowers and heavy concentration of non-banks in the Ginnie space. The pandemic and the subsequent passage of the CARES Act demonstrated the need for a federal liquidity facility for non-banks. This is a more pressing issue in the Ginnie Mae market, with potential systemic implications for the mortgage market. This article explains these issues in detail, concludes that a federal liquidity for non-banks is necessary, and discusses two options, one for the current crisis and another for the future: a COVID-19 liquidity facility at the Federal Reserve, and in the long run, federal home loan Bank (FHLB) membership for non-banks to allow for access to FHLB advances. <b>TOPICS:</b>Financial crises and financial market history, MBS and residential mortgage loans, real estate <b>Key Findings</b> • Due to differences in market structures, servicing Ginnie Mae mortgages puts more strain on servicers than servicing GSE mortgages. • The Ginnie Mae segment is exposed to both a higher-risk borrower profile and a much larger concentration of non-banks, which have no source of government liquidity. • There is a need for a government-backed liquidity facility for non-banks.

Forbearance During the Pandemic: What Do We Know Thus Far?

The pandemic has caused severe distress for households across the country. In response, policymakers and lenders have provided for the widespread use of forbearance, providing homeowners payment relief during this crisis. This article highlights the particular features of these forbearance plans, the rate of take-up over the initial period, and then discusses the potential outcomes with respect to mortgage delinquency and foreclosure rates at the end of the forbearance period. TOPICS:MBS and residential mortgage loans, risk management, legal/regulatory/public policy Key Findings • The onset of the pandemic caused severe distress for millions of families who lost jobs or had reduced income. • At the peak, more than 4.3 million mortgages, 8.5 percent of all mortgages outstanding were placed in forbearance, with an even higher share of FHA and VA loans granted such relief. • Mortgage delinquencies have spiked. However, an improving job market and a strong housing market could be enough keep all of these delinquencies from becoming foreclosures if borrowers are able to exit from forbearance and resume making their payments.

Market structure and credit procyclicality: Lessons from loan markets in the European Union banking sectors

The aim of this paper is to investigate whether the market structure has an impact on procyclicality in the European Union bank loan markets. The cyclical responses of three types of bank loans (residential mortgage loans, consumer loans, and corporate loans) are quantified separately using the interacted panel vector autoregression model at the country level and the single-equation panel regression model at the bank level. Using a sample of 26 European Union countries, we find that the procyclical responses of residential mortgage loans and consumer loans are significantly stronger and prolonged when the banking sector is more concentrated or dominated by foreign banks. However, we find that there are nonlinear relationships between the market structure and credit procyclicality based on bank-level data. We also find some heterogeneities between advanced and transitioning European Union banking sectors. Finally, our findings confirm the leading role of residential mortgages in intensifying credit fluctuations.