AdobeStock_336270258.jpeg
VoxEU Column COVID-19 Financial Regulation and Banking

Resolving mortgage distress after COVID-19: Some lessons from the last crisis

Debt moratoria introduced to mitigate the initial impact of the coronavirus pandemic have begun to expire. This column derives lessons for policymakers by analysing detailed micro data on arrears resolution during the recovery from the post-2008 Irish financial crisis.  It highlights trends in borrower engagement in resolution, the financial health of households engaging, details on how loans were restructured, and the success of loan restructures. 

The onset of the coronavirus pandemic in March 2020 led to the issuance of widespread debt moratoria across the globe to provide liquidity relief to households.  These moratoria have begun to expire in Ireland and will do so across Europe in the coming months, which leads to some crucial questions about the design of further policies that should be implemented to deal with the ongoing impact of the pandemic. 

Despite the severity of the last recession, most European countries did not experience widespread mortgage defaults. In contrast, Ireland experienced an extraordinary boom in house prices, followed by a similarly extraordinary bust and a collapse in employment. By 2013, nearly 14% of the Irish owner-occupier mortgage market was in arrears of more than 90 days. In this vein, Ireland’s boom-bust experience mirrored that of many of the worst-affected states in the US, which are the subject of a rich body of research and debate, and have contrasting views on the role of credit supply and demand factors put forward by, among others, Mian and Sufi (2009) and Adelino et al. (2016). 

Owing to concerns about borrowers losing their homes, widespread mortgage modification was an important feature of the banking and policy response, with repossession much less common. The Central Bank of Ireland imposed targets on banks in 2013 to ensure the vast majority of distressed mortgage cases received a “sustainable solution”. By 2016, over one in six owner-occupier mortgage accounts had been restructured. 

In our related paper (McCann and O’Malley 2020), we exploit rich banking sector data on mortgage loans and distressed household balance sheets to analyse the way in which mortgage modifications functioned in Ireland over the last decade, hoping to derive lessons for both the Irish and global policy response to the expiration of pandemic-related payment moratoria.

Borrower engagement

We focus initially on borrower engagement, defined in the Irish case as an instance where the borrower contacts the lender because of financial difficulty and completes a Standard Financial Statement (SFS) form. This SFS is a standardised industry-wide template that captures all relevant information on income, expenditure, debt, and demographic factors at the point of distress. SFS forms provide a rare and rich source of financial information on mortgage borrowers. From our research on borrower engagement, based on retail banking data on engaging borrowers between 2011 and 2018, we highlight a number of key facts: 

1. Engagement was widespread, despite the relatively low probability of repossession. More than one-in-five mortgages in Ireland had an associated SFS completed during the 2012-2020 period.

2. Many borrowers engaged early, before missing payments, stemming the potential flow to arrears. At most points in the last decade, more than half of engaging mortgage borrowers were engaging before having defaulted.  Proportionally, more than twice as many performing borrowers were engaging before defaulting, relative to defaulting without having engaged. 

The above clearly show that, even after a collapse in house prices and facing low repossession risk, a large number of mortgage borrowers nonetheless engaged with their lender to arrive at resolutions. This relative success points towards many borrowers having high default costs, in keeping with the modern literature stating that ‘strategic’ or ‘ruthless’ default is uncommon, and requires deep levels of negative equity before occurring. 

Despite this relative success, part of the reason for the ongoing issue of long-dated mortgage arrears in Ireland can be traced to the lack of engagement with the system among a relatively small group. Even where the majority of households engage faithfully to resolve their mortgage distress, lack of an effective collateral repossession regime leads to an accumulation of deep arrears cases representing just over 5% of all owner-occupier mortgage balances, suggesting that the ‘last resort’ of a functioning foreclosure regime is an important part of the overall debt resolution architecture. 

3. Non-engagement was common among those that accumulated long-dated arrears over the decade. Close to 40% of those in long-term mortgage arrears (greater than two years past due) are measured to be ‘non-engaging’ or ‘non-cooperating’ across two different studies in recent years. Among those with arrears of ten or more years, ‘non-cooperating’ rates are over 50%. 

Restructures: Temporary or permanent?

The pandemic raises a number of important questions about the nature of future restructures. Lenders’ beliefs about the shape of the recovery will partly determine restructuring. If a vaccine leads to a sharp ‘V’, with activity recovering to previous levels across all sectors, temporary relief such as additional moratoria or periods of interest-only payments are warranted. If a ‘swoosh’ or an ‘L’ prevails, some borrowers will be unable to service debts over a long time-frame, requiring deeper levels of debt relief. Finally, if a ‘K-shape’ takes hold wherein some sectors undergo long-term structural decline, deep relief over a longer term may be more appropriate, with resort to insolvency or movement to social welfare schemes when required. A challenge facing lenders and policy makers today is the uncertainty around this path. 

Looking at the Irish situation over the last decade, where negative equity was a prevalent problem and many borrowers suffered deep and long-lasting shocks to repayment capacity, we learn that: 

4. Policy action can have large effects. Largely in response to regulatory pressure, between 2010 and 2019, short-term arrangements moved from accounting for close to 70% of outstanding restructures to less than 15%. 

5. Temporary, short-term modifications are appropriate in the right circumstances, but require follow-on deeper restructure in many cases. Among performing mortgages with a short-term modification in 2011-12, 75% were performing at end-2017, with close to half of this group receiving a long-term modification along the way. 

Figure 1 Short-term arrangements as a share of all restructures, 2010-2020

Notes: Short-term (‘temporary’ in graph) arrangements include interest-only arrangements, temporary moratoria, and payment reductions. 
Source: McCann and O’Malley (2020) 

Financial distress

The financial and socioeconomic profile of borrowers seeking repayment relief can have implications for societal attitudes towards the appropriateness of that relief. In the Irish case, media coverage of high-income, high-debt cases in insolvency courts has at times created an inaccurate impression of the typical experience of those seeking mortgage debt resolution. Furthermore, in cases where borrower incomes have been permanently and deeply damaged by the shock being experienced, the policy solutions required are likely to extend beyond the banking system towards the social safety net. From a policymaker perspective, access to information such as the SFS is imperative in informing these discussions. In the Irish case we learn that in the last decade:

6. When borrowers engaged, many were experiencing grave financial difficulty.

a. They had suffered an average income fall of one third since mortgage origination;

b. They typically had loan-to-income ratios above 5; 

c. They were on average on the cusp of negative equity. 

7. These financial difficulties had a meaningful effect on living standards. While in the general population the median household was consuming a third of their monthly income in 2018, in the SFS population, due to lower income levels, almost three-quarters of income is devoted to consumption, implying a much lower capacity to save or service debt in this group.

8. This translated into an incapacity to service debt for many. Eighteen percent of households engaged with monthly deficits (income less expenditure less mortgage repayments) of over €1,000. 

9. Living standards had fallen dramatically. Close to 30% of households engaged with consumption levels already below ‘reasonable living’ recommendations issued by the Insolvency Service of Ireland.

Post-modification outcomes

Our data also allow us to study the way in which modifications were issued, and the factors associated with their success. The research tells us that lenders in many cases issued modifications that left borrowers in precarious situations where there was little resilience to additional shocks. Across the market, the default rate in recent years has been about five times higher for modified loans compared to performing loans with no such history.

10. Even after receiving modification, vulnerabilities remained. Those with modified mortgages have higher payment-to-income ratios than the general population.

11. Repayment relief. The average long-term mortgage modification led to a reduction in monthly repayments of 16.6% compared to an average 36% in the USA HAMP.

12. The choice of modification type is associated with wide variation in repayment relief. The average split mortgage resulted in a 50% cut in monthly repayments, while the average term extension led to a 26% cut. Arrears capitalisations resulted in the smallest falls in monthly repayments of 10.5% on average. 

13. Repayment relief works. Re-default rates are lower where the firm has made deeper cuts in monthly repayments as part of the modification (Figure 2).

14. Typical signals are relevant factors to target post-modification.  Re-default rates are lower where the modified loan has a lower loan-to-value ratio and a lower payment-to-income ratio.

Figure 2  Relationship between repayment reductions and re-default

Notes: Each graph is the empirical probability of a missed payment in the two years after a modification is issued. Vertical bars represent an approximate 95% confidence interval. 

Conclusion

Through an analysis of detailed micro data in the last great economic crisis, we derive lessons for policymakers dealing with expiring payment moratoria after the initial stage of the pandemic. Early engagement on behalf of both borrowers and lenders is crucial for successful resolution of loan arrears. Both short-term and long-term solutions play a role in restoring the stability of household finances, but long-term solutions are essential to avoid drift into difficult-to-resolve, long-term arrears cases. Many households engaging in arrears resolution have severely damaged finances and without adequate policies, many may drift into long-term arrears. Finally, we find that deeper cuts to payment burdens lead to a higher probability of loan payment.  

References

Adelino, M, A Schoar and  F Severino (2016), “Loan Originations and Defaults in the Mortgage Crisis: The Role of the Middle Class”, The Review of Financial Studies 29(7): 1635–1670.

McCann, F and T O’Malley (2020), “Resolving mortgage distress after COVID-19: some lessons from the last crisis”, Central Bank of Ireland Financial Stability Note 2020, No. 7.

Mian, A and A Sufi (2009), “The consequences of mortgage credit expansion: Evidence from the U.S. mortgage default crisis“, The Quarterly Journal of Economics 124(4): 1449–1496.

1,260 Reads