Philip Bunn.
As households spend more of their income making payments on loans they are more likely to get into arrears. This risk rises gradually at first, but above a certain point they enter a danger zone where the probability of arrears rises sharply. Knowing where this danger zone lies is really important because, if it comes a little earlier or a little later, that can make a big difference to the number of people who fall into it, although as this post shows, it is hard to identify this danger zone precisely. Nevertheless, understanding what leads households to get into financial difficulty is crucial for assessing how such difficulties might increase following rises in interest rates or unexpected falls in income.
Why look at debt servicing ratios?
The proportion of income spent making payments on loans (the debt servicing ratio or DSR) is a good summary indicator of the sustainability of a household’s debt: it brings together information of the size of the payment and the amount of income available to make that payment. DSRs help us to understand households’ current financial health, but perhaps more importantly, estimating how DSRs might evolve in the future is a way of assessing how financial distress could increase in response to economic events that lead to higher DSRs, such as rises in interest rates or scenarios that lead incomes to fall for some households. This matters to policymakers at the Bank because they need to understand the implications of changing interest rates and how different scenarios might affect financial stability: widespread increases in financial distress have the potential to lower banks’ capital positons and threaten the resilience of banks.
To be able to evaluate how higher DSRs might increase financial difficulties, it is crucial to know if there is a point at which we should really start to worry about how much households spend on their loan payments, i.e. is there a danger zone for DSRs where the risk of facing distress rises substantially? If so, a given percentage point rise in a household’s DSR may have a very different impact on their likelihood of arrears depending on whether it pushes them into the danger zone or not. Defining that danger zone means looking at household level data because aggregate data masks very large differences between the financial situations of different households. And in practice, limited data availability means that it is only really possible to analyse this relationship for households with a mortgage.
The evidence
Data from a range of UK household surveys suggest that a DSR danger zone does exist where rates of mortgage arrears start to rise sharply once DSRs exceed a certain level, although it is hard to define that danger zone precisely. In the period since 2000, that danger zone appears to lie between spending 30% and 50% of pre-tax (or gross) income on mortgage payments, depending on the measure used (Chart 1). Although it is ultimately post-tax income that is used to make mortgage payments, gross income is used here because it is more widely available (we’ll come back to this later).
Chart 1: Mortgage DSRs and arrears
There is a big difference in the number of people who are potentially at risk depending on whether the danger zone is defined as having a mortgage DSR above 30% or above 50%. Data from the latest Bank of England/NMG Survey (discussed in more detail here), imply that there were around 700,000 UK households spending more than 30% of pre-tax income on mortgage payments in September 2015, but only 200,000 spending more than 50%. Correspondingly large differences are also likely to exist using the different definitions in scenario analysis.
The different household surveys used each have advantages and disadvantages, and that makes it hard to choose a single best estimate of the DSR danger zone. For example, the English Housing Survey (EHS, the Survey of English Housing before 2008) has the largest sample size and is the longest running of the relevant surveys. The Wealth and Assets Survey (WAS) and NMG Survey only have relatively small sample sizes in comparison, and only data for the post-crisis period, but the WAS may capture wealthier households better than other surveys, while the NMG survey is the only survey carried out online, where households may be more willing to disclose sensitive information about their finances than they would to an interviewer in their home (households in the NMG survey report higher rates of arrears that are closer to aggregate data than other surveys).
Where the DSR danger zone starts is also sensitive to the definitions used. As discussed in more detail below, it can depend on the sample period, how arrears are measured, whether unsecured loan payments are included and whether net or gross income is considered. Nevertheless, the overall pattern is that there is a rapid rise in arrears somewhere between DSRs of 30% to 50%.
Sample period/nature of the shocks
Since the turn of the century, the relationship between DSRs and arrears appears to have been relatively stable, at least within the EHS data, which is the only survey that covers the whole period. But the relationship does look very different in the early 1990s. Data from the British Household Panel Survey (BHPS – the only survey going back that far, before it was discontinued in 2008) suggest that arrears rates started rising substantially at DSRs of above 15% in the early 1990s, a much lower threshold than in the post 2000 period (Chart 2). That indicates how the relationship could be sensitive to the nature of the shock and the situation in which it occurs. In the early 1990s arrears may have risen sharply at lower DSRs because the combination of big rises in unemployment and interest rates led to large and unexpected increases in DSRs. (Hopefully) the more recent period ought to be the most relevant in terms of assessing how many households might face financial difficulty in future scenarios, but very severe shocks could affect the position of that danger zone.
Chart 2: Mortgage DSRs and arrears (BHPS data, arrears 2 months+)
Time in arrears
Arrears of any length are a concern, but it is more persistent arrears that are more likely to lead to losses for banks. Restricting the arrears definition to households with more persistent payment problems tends to raise the DSR above which arrears tend to start rising sharply. In the EHS, arrears rates for arrears of any length start increasing substantially beyond DSRs of 30%, but for arrears of 3 months or more, the pickup in arrears is most striking only beyond DSRs of 50% (Chart 1).
Including unsecured payments
The above analysis is all based on mortgage payments, but many mortgagors have unsecured debt that we ideally would want to take into account too. While, there is less data available on unsecured loan payments than mortgage payments, it is possible to include them in the DSR using the NMG survey. That does make a difference: 90% of the households in mortgage arrears also had some form of unsecured debt, implying that just looking at mortgage payments understates how much income these households devote to servicing their debts. Just focussing on mortgage payments, arrears rates rise mostly notably above DSRs of 30% in the NMG Survey, but that tipping point increases to around 40% once payments on all loans are accounted for (Chart 3).
Chart 3: DSRs and mortgage arrears (NMG Survey data 2014-15, arrears 2 months+)
Net versus gross income
Using income net of taxes rather than pre-tax income also increases the DSR threshold at which arrears rates rise sharply. Net income is a conceptually better measure to use because loan payments are made out of net rather than gross income, but net income is not available in all surveys. However, it is available in the more recent EHS data, and putting the analysis into net space does increase the point at which arrears rates rise more substantially, from around 30% using gross income, to 40% using net income.
Chart 4: Mortgage DSRs and arrears (EHS data 2008-2013, any arrears)
Conclusion
This post has shown how rates of mortgage arrears increase with the proportion of income spent on servicing debt. Above a certain DSR, arrears rates rise sharply, and it is the number of households who fall into this DSR danger zone that we should worry about when thinking about how financial distress might rise following increases in interest rates or unexpected falls in income.
Small differences in the definition of this DSR danger zone can have a big impact on the number of people who fall into it. It is, however, hard to precisely define this danger zone: it varies across the different surveys, time periods and definitions of arrears. But it typically lies between spending 30% and 50% of gross income on mortgage payments. And roughly speaking, 10 percentage points can be added to that range for either including unsecured loan payments or for using net rather than gross income. The fact that there are a range of estimates for the definition of this danger zone suggests that, rather than choose a single definition, it is better to monitor how adverse scenarios might affect the distribution of DSRs .
Philip Bunn works in the Bank’s Structural Economic Analysis Division
If you want to get in touch, please email us at bankunderground@bankofengland.co.uk. You are also welcome to leave a comment below. Comments are moderated and will not appear until they have been approved.
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.
