Household indebtedness and financial stability − speech by Colette Bowe

Speech

Thank you for this invitation to open the workshop today.

The topics you are planning to discuss are of the utmost importance, and I am delighted to be able to contribute to the discussion. I have long been interested in the contribution that mutuals, in their various forms, can make to the array of market possibilities open to borrowers and savers – and indeed served on the board of one of our largest mutuals some years ago.

And, Andre, it is a pleasure to be here again at the Business School and to reconnect with colleagues.

As you will all know, the Bank of England’s Financial Policy Committee (FPC), of which I am an external member, is responsible for monitoring and acting to remove or reduce systemic risks in order to ensure the resilience of the UK’s financial system. This obviously covers a very wide range of issues, and I have chosen today to focus on some that I think will be of interest to this audience. It is of course equally obvious to all of us here today that the many challenges posed by rising inflation require intense vigilance.

I am planning to cover two topics this morning: first, to review our current thinking about the importance of household indebtedness to financial stability – and in this section to update you on the progress of some important work ongoing at the Bank about how we are deepening our understanding of how household debt burdens have changed over time and what debt might mean for household behaviour, especially in the current economic environment.

The second topic I will touch on is the recent review that the Bank has done of its so-called housing tools - really, they ought to be called mortgage market tools – and the outcome of the consultation we have done, and whose results we have implemented, on one of these tools, the affordability test.

Importance of household debt to financial stability

Most, if not all of us, in this room will recall the interplay between mortgage debt and financial stability that was characteristic of the global financial crisis (GFC) that began in 2007. The crash in the US subprime mortgage market, combined with opaque securitisation and inadequate bank capitalisation led to significant spillovers to the global financial system. In the UK, loose underwriting standards and high mortgage leverage acted as some of the weaknesses in financial stability and amplifiers of the crisis. All of this has been somewhat exhaustively documented but it is worth reminding ourselves of a couple of points today.

First, unsustainably high mortgage debt has historically affected UK financial stability via two channels. The first is through the effect on borrower resilience, where highly indebted households face challenges from real income cuts and/or mortgage rate increases and cut spending sharply - which can amplify a downturn, with potentially systemic consequences for financial stability.

Second, shocks can be transmitted to financial stability through the lender resilience channel, where lenders experience losses when highly indebted households face re-payment difficulties. These losses can result in a reduction of the supply of credit to households, businesses and the wider economy.

As we have said in our most recent Financial Stability Report (FSR), which was published on 5 July^{footnote [1]}, our view is that the current shock does not at this stage pose the same risks to lenders and the financial system as previous shocks. Lenders are well capitalised and have limited direct exposures to Russia and Ukraine, as well as being resilient to risks stemming from sectors which are particularly exposed to higher commodity prices. The FPC continues to monitor the resilience of banks to further downside risks. And to support our understanding of those risks, the 2022 annual cyclical scenario (ACS) stress test will be launched later this month. It will assess the resilience of the UK banking system to deep simultaneous recessions in the UK and global economies, real income shocks as well as large falls in assets prices and higher global interest rates.

But even in a less severe scenario, we are conscious that the rise in living costs and interest rates will put increased pressure on the finances of UK households in the coming months, with lower-income households in particular finding it difficult to adjust their spending behaviour in response to the rise in prices. This will test the degree of borrower resilience in the system as borrowers struggle with bills or [to] manage their other spending commitments.

We are facing a highly uncertain economic outlook, as my colleagues on the MPC have been reporting. So, at the risk of restating the blindingly obvious, I have to say that everything I am saying now has to be seen against that backdrop.

I would like first to give you a quick overview of some useful granular research being done by colleagues at the Bank on debt service ratios – DSRs for short - recently published in the latest Financial Stability Report. In the time available, this can only be a curtain raiser – but if any of you would like to follow up with my Bank colleagues, I would be delighted to connect you.

Household debt – scene setting

Before I delve into the research on the DSRs, let me take a step back and provide some scene setting on the household debt.

The total stock of UK household debt (excluding student loans) was just under £2 trillion in 2022 Q1, equivalent to around 124.5% of total household income. Although high compared to historical standards, this is materially below its 2008 peak of 146%.

This has been driven by deleveraging and by slower growth in new household lending post global financial crisis. In particular, average annual household debt growth slowed from around 9% pre-GFC to around 2.4% over the last decade. At the same time, annual income growth slowed by less, from an average of 4.6% pre-GFC to 3%.

And higher debt levels have generally been more affordable than prior to the GFC given a prolonged period of low interest rates, as well as policies to build resilience in the financial system, which I will come back to later in my remarks.

So households entered the pandemic period reasonably resilient and have been able to maintain a lot of this resilience. While lockdowns brought a sharp contraction in activity, they also led to a large rise in household savings. This coupled with extensive fiscal, monetary and lender support (through payment holidays) helped many households to bridge through.

To monitor household debt affordability for the most highly indebted households, in previous FSRs, we focused on gross DSRs – that is, the ratio of mortgage repayments to pre-tax income. Using the Wealth and Assets Survey, Bank staff research has showed that households with a mortgage DSRs at or above 40% were more likely to experience repayment difficulties during the financial crisis. In 2022 H1, following the pandemic, the share of such households increased to around 1 ¾%, and were broadly in line with pre-GFC averages.

Chart 2: Share of households with high mortgage DSRs increased but remain below pre-GFC peak

Footnotes

• Source: NMG Consulting Survey, British Household Panel Survey/Understanding Society and Bank calculations

Within the context of more recent economic challenges, let me now turn to measures we developed to better assess the impact of increased cost of living.

Debt-servicing in the light of increased cost of living

As we are all only too aware, we are living through a considerable increase in the cost of living as a consequence of Russia’s invasion of Ukraine driving the price of food and energy higher, with restrictions of gas supplies to Europe starting to fully feed into household bills. And the distributional effects of such a shock are serious: households in the lower parts of the income distribution are disproportionately affected by the increases in the costs of essential items.^{footnote [2]}

The gross DSR measures we generally report are not well suited to account for the increases in the cost of living. As such in the July FSR, we introduced some innovative work which adjusts the gross DSRs to better capture debt affordability pressures that households are likely to face. The cost of living adjusted DSRs are constructed based on a measure of disposable household income which deducts taxes and an estimate of essential spending (including spending on utility and energy bills, food and transport, to name a few). The measure of debt-servicing consists of regular debt repayments.^{footnote [3]}

The new, more granular measure better captures the effects of the following factors on debt sustainability:

• Increase in cost of essential spending - this is aimed at capturing effects such as the impact of high global energy prices on households.
• Possible increase in interest rates - this is capturing tightening in monetary policy in response to rising inflation. Interestingly, at this time we estimate that the majority of households are on fixed rate mortgage contracts, so will not necessarily see their mortgage payments change in the very short term. However, we are of course aware that the fixed rate deals available in the market are changing rapidly and so many borrowers will face an increase in payments when they come to re-mortgage and is included in our analysis.
• The possibility of higher unemployment which stems from a slowing general economic outlook – where although we are not yet seeing this effect in the labour market, we include future possible stresses in our analysis.
• Offsets from nominal wage growth and fiscal policy. Increases in nominal pay can go some way to offset the impacts of higher living costs, although there is considerable uncertainty about the size of pay increases and how they will be distributed.

Taking all those together we showed in the July FSR that the share of households with high cost of living adjusted DSRs are likely to remain at around the current levels of 1.7% over the course of 2022, and below the pre-2008 levels (at 2.8%). While the share of households struggling is expected to remain below Financial Crisis peaks this year, we may still see some increase in defaults or households needing to make sharp, and difficult, cuts to their spending.

Since the publication of the FSR, the MPC published its updated projections in the August MPR. As the August MPR sets out, the rise in energy prices has exacerbated falls in real incomes, and so led to another significant deterioration in the outlook in the UK, with GDP growth slowing and the economy contracting later this year. In turn, this means that the outlook is even more challenging for UK households than it was at the time of the FSR.

Moreover, there are a number of uncertainties attached to the outlook. In particular, a lot depends on how persistent the energy price shock will be, what happens to incomes over the same periods and how wage growth is distributed.

This new approach to estimating DSRs gives us the ability to take a more granular approach to estimating the possible evolution of household debt - and hence to testing whether the financial system itself is likely to remain resilient to these macro shocks.

The FPC will continue to monitor UK household debt vulnerabilities, as well as other vulnerabilities affecting financial stability, such as those arising from pressures in the corporate sector, and major UK bank resilience to those vulnerabilities. This is a centrepiece of the job of the FPC.

The design and implementation of the mortgage market rules

I now want to turn to the topic of the FPC’s measures, and the role that they have played – and continue to play – in supporting the resilience of households and the financial system. In the wake of the financial crisis of 2008, regulators took various steps to guard against a further risk-related increase in household indebtedness. Following their Mortgage Market Review, in 2014 the FCA introduced tougher requirements for lenders to ensure that prospective borrowers could afford their mortgages. This included requiring verification of income and taking account of other debt and spending commitments. And, importantly, where the interest rate is variable or is fixed for less than 5 years, a requirement to test whether borrowers could still afford the loan if, at any point in the first five years, interest rates were to rise.

At the same time, the FPC supplemented the FCA rules with a recommendation specifying that lenders should assess whether borrowers could still afford their mortgages if rates were to rise by 3 percentage points (the affordability test). This was accompanied by a requirement for lenders to limit loans of more than 4.5 times income to at most 15% of their new lending (the flow limit). It is these requirements have led to the resilient position for household indebtedness as we enter this period of heightened uncertainty.

In 2014, when the test was introduced, most major lenders were using self-imposed interest rate stress tests of around 7%, compared with SVRs in the region of 4-4.5% at the time. Which implied a "stress" of 2.5-3%, broadly in line with the FPC's recommendation.

What the FPC was aiming to do was to reinforce prevailing underwriting standards and discourage an increase in borrowers with extreme levels of indebtedness.

Following a consultation earlier this year, however, the FPC has recently withdrawn its affordability test, although the underlying FCA framework remains in place. There are a number of reasons for the decision:

• Simplification of the regulatory framework. Macro-prudential mortgage tools were not widely used prior to the GFC, and the UK was among several countries to introduce such tools following the crisis. Mortgage policy tools were new both in the UK and internationally back in 2014. We have monitored and reviewed the tools since and obtained more evidence as to how they worked in practice. This included digging into loan-level data to understand the relative impact of the different measures to date, as well as modelling the impact of the measures in an alternative scenario in which house prices and household debt rapidly increased, putting more pressure on household finances.
• The evidence gathered suggested that the 300bps FPC Recommendation added little over and above the LTI flow limit and the FCA’s affordability testing rules. The FCA rules stipulate, in particular, that lenders take into account market expectations for interest rates and as a backstop that they impose at least a 100bps stress test. So the removal of the FPC recommendation ensures a simpler and more proportionate regulatory framework.
• We learned that the link to SVRs introduced uncertainty about how the affordability test would behave, particularly as Bank Rate moves. We wanted to avoid the risk of rules being overly stringent, even though some concerns were raised during the consultation about the timing of withdrawing the test.
• We also sought feedback on this decision via a public consultation. In general, the consultation responses supported the FPC judgment that the flow limit without the affordability test ought to provide appropriate level of resilience in the financial system, in a more predictable and proportionate way, especially with the FCA framework remaining in place.

However, a minority of respondents to the consultation opposed withdrawing the affordability test given concerns about the deterioration in macroeconomic outlook and the risk that borrowers could take on mortgages that are not affordable. The FPC is alert to these concerns, but judges that prudent underwriting standards over recent years have improved the resilience of households with mortgage debts, while the LTI flow limit and the FCA’s rules will continue to maintain these standards and ensure that households do not take on unaffordable levels of mortgage debt.

In light of higher interest rates and a weaker outlook, my colleagues and I on the FPC will continue to closely monitor the pressures on household finances, including through using our enhanced measure of DSRs.

Conclusion

I have spoken this morning about only one part of the FPC’s work – covering household debt and the mortgage market. Of course, we have a range of issues we constantly monitor and, if I may close with a commercial break, I hope that you will all continue to read our Financial Stability Reports summarising all the issues we aim to cover. The next FSR is due on 13 December but the FPC will also provide an update on our latest assessment of risks following the Q3 meeting in October.

Thank you.

I am grateful to Andrew Bailey, Sarah Breeden, Gabija Zemaityte, Sarah Venables, Jeremy Franklin, Elspeth Hughes, Raakhi Odedra, Neal Kilbane, Renee Horrell, Ali Moussavi, Henry Tanner, and Matthew Waldron for their assistance in helping me prepare these remarks.