Stress testing the UK banking system: 2021 Solvency Stress Test results

Executive summary

The Bank’s 2021 ‘solvency stress test’ (SST) shows the major UK banks are resilient to a severe path for the economy in 2021–25 on top of the economic shock associated with the Covid-19 (Covid) pandemic that occurred in 2020. These results support the FPC’s judgement that the system is resilient to outcomes for the economy that are much more severe than the Monetary Policy Committee’s (MPC’s) central forecast.

Reflecting the circumstances of the pandemic, the SST differs from the Bank’s usual annual cyclical scenario (ACS) stress test. The aim of the SST has been to cross-check the FPC’s judgement, made in December 2020 following a ‘reverse stress test’, that banks have sufficient capital to continue to support UK households and businesses even if economic outcomes are considerably worse than expected.

The major UK banks’ resilience in the test reflects a strong end-2020 starting point, with an aggregate Common Equity Tier 1 (CET1) capital ratio of 15.9%, and Tier 1 leverage ratio of 5.7% (excluding the software asset benefit that is due to expire at the start of 2022). That robust starting position is in part due to the build-up of capital since the global financial crisis, reflecting post-crisis reforms including higher capital requirements. Further support to capital ratios was provided by actions taken during 2020 by the banks, the Bank, including the PRA, and public authorities more broadly in response to the pandemic. These actions included the banks’ cancellation of final 2019 dividends.

The major UK banks’ aggregate CET1 capital ratio falls by 5.5 percentage points in the stress to a low point of 10.5%. A significant driver of this reduction is credit impairments, and the majority of these occur on UK-based lending. This low point compares to a 7.6% ‘reference rate’, which comprises banks’ minimum requirements and systemic buffers, and is adjusted to account for the impact of International Financial Reporting Standard 9 (IFRS 9). The aggregate Tier 1 leverage ratio low point of 4.8% is also above the reference rate of 3.7%. All eight participating banks remain above their reference rates for both CET1 capital and Tier 1 leverage ratios.

As previously indicated, the aim of the SST has been to update and refine the FPC’s assessment of banks’ resilience and ability to lend in a very severe intensification of the macroeconomic shock arising from the pandemic. Consistent with the nature of the exercise, the FPC and PRC will therefore not use the SST as a direct input for setting capital buffers for UK banks. No individual bank is required to strengthen its capital position as a result of the test.

The Bank intends to revert to the ACS stress-testing framework for 2022. Due to its broader and countercyclical nature the ACS framework is well suited to informing the setting of capital buffers, including the UK countercyclical capital buffer. The FPC judges that the results of the SST, together with the central outlook and the return to the ACS framework for 2022 are consistent with the PRC’s decision to transition back to its standard approach to capital-setting and shareholder distributions.

1: Overview of the solvency stress test

The Bank’s 2021 solvency stress test (SST) has acted as a cross-check on the FPC’s judgement that the banking system is resilient to outcomes for the economy that are much more severe than the MPC’s central forecast.
As described in Box A, once the economy enters a stress, such as that driven by the Covid outbreak, the Bank of England (hereafter referred to as ‘the Bank’) uses stress tests to assess whether the buffers of capital that major UK banks and building societies (hereafter ‘banks’^{footnote [2]}) have built up are large enough to deal with how the prevailing stress could unfold.

The Bank’s ‘reverse stress test’ (RST) exercise in 2020 found that very severe economic paths for the UK and global economies would be needed in order to deplete regulatory capital buffers by 5.2 percentage points, in line with the drawdown in the 2019 annual cyclical scenario (ACS). Based on that, in December 2020 the Financial Policy Committee (FPC) judged that UK banks, in aggregate, had capital buffers that allowed them to lend in, and remain resilient to, a wide range of possible outcomes for the UK and global economies.

The SST has acted as a cross-check on the FPC’s judgement of how severe the pandemic-related stress that began in 2020 would need to be to jeopardise banks’ resilience and challenge their ability to absorb losses and continue to lend. The macroeconomic scenario underlying the SST is a severe path for the economy in 2021–25 on top of the economic shock associated with the Covid pandemic that occurred in 2020. It is broadly consistent with the ‘double-dip’ scenario generated in the RST.

The Bank published interim results in the July Financial Stability Report, which took into account early credit projections from participating banks, and for areas other than credit were based on Bank staff’s desktop analysis. This document sets out the full updated and final results of the 2021 SST, including bank specific outcomes. Eight major banks are included in the test including, for the first time, Virgin Money UK.

The 2021 SST shows that the banking system in aggregate is resilient to an extremely severe stress scenario.
The eight participating banks started the SST with an aggregate Common Equity Tier 1 (CET1) capital ratio of 15.9% of risk-weighted assets (excluding the software assets benefit that is due to expire in 2022, as explained in Section 2) (Chart 1). That is somewhat higher than previous years, in part due to actions taken during 2020 by banks and public authorities, including the banks’ cancellation of final 2019 dividends (see Section 2 for more details). At the low point of the stress in 2022 the major UK banks’ aggregate CET1 capital ratio, on a transitional International Financial Reporting Standard 9 (IFRS 9) basis, is 2.8 percentage points above the ‘reference rate’.^{footnote [3]} Banks’ aggregate Tier 1 leverage ratio also remains above the reference rate in the stress.

Chart 1: Banks started the stress test with a strong capital position in aggregate, and even at the low point remain some way above the reference rate

Aggregate CET1 capital ratio of major UK banks and impact of the 2021 SST scenario ^{(a) (b) (c)}

Footnotes

• Sources: Participating banks’ Stress Testing Data Framework (STDF) data submissions, PRA regulatory returns, published accounts, Bank analysis and calculations.
• (a) The CET1 capital ratio is defined as CET1 capital expressed as a percentage of risk‑weighted assets, where both terms are defined in line with the Capital Requirements Regulation (CRR) and the UK implementation of the Capital Requirements Directive (CRD V) via the PRA Rulebook. Unless otherwise stated, any references to EU or EU-derived legislation throughout this document refer to the version of the legislation which forms part of retained EU law.
• (b) Major UK banks are Barclays, HSBC, Lloyds Banking Group, Nationwide, NatWest Group, Santander UK, Standard Chartered and, from end-2020, Virgin Money UK. Prior to 2011, data are Bank estimates of banks' CET1 ratios. Capital figures are year-end.
• (c) 2020 capital ratio and impact figures exclude the software assets benefit (see Section 2 for a full description of this).

The CET1 capital low point in the SST is higher than that observed in the 2020 RST and the 2019 ACS, despite a broadly similar drawdown, reflecting the higher start point (Chart 2).The drawdown of 5.5 percentage points in the SST is also broadly in line with the interim results published in July 2021.

Chart 2: The impact of the SST is broadly consistent with last year’s RST and the 2019 ACS, but the capital low point is higher

Impact of recent Bank stress tests ^{(a) (b) (c) (d) (e)}

Footnotes

• Sources: Participating banks' STDF data submissions, published accounts, Bank analysis and calculations.
• (a) See footnote (a) to Chart 1.
• (b) The RST and SST capital ratios incorporate the effect of amendments in the CRR ‘Quick Fix’ applicable from 27 June 2020 to allow 100% transitional relief of eligible IFRS 9 provisions until the end of 2021 (see footnote 5).
• (c) The impact of the 2019 ACS does not include the conversion of Additional Tier 1 (AT1) instruments that was a feature of that test.
• (d) Capital ratio figures for the 2021 SST shown in the chart exclude the software assets benefit.
• (e) The reference rate used in the SST is determined in the same way as the hurdle rate used in the 2019 ACS except for changes in the transitional relief implemented as part of the CRR ‘Quick Fix’. Both reference and hurdle rates comprise the sum of each bank’s minimum capital requirements and any systemic buffers that they are required to hold.

The results confirm the FPC’s previous judgement that UK banks, in aggregate, are resilient to a much more severe scenario than the MPC’s central forecast and have sufficient capital to support UK households and businesses even if economic outcomes are considerably worse than expected. While the SST was designed as an exercise specifically to test banks’ resilience against an intensification of the macroeconomic shocks arising from the pandemic seen in 2020, the FPC judges that the results are informative about banks’ overall resilience to other shocks that could arise.

Since the end-2020 balance sheet cut-off date for the 2021 SST, the aggregate CET1 capital ratio of the major UK banks has picked up further to 16.5% in 2021 Q3 (see Financial Stability Report December 2021, ‘2: In focus – Resilience of the UK banking sector’).

2: Broad narrative of the scenario and results

The macroeconomic scenario is very severe, representing an intensification of the shocks seen in 2020.
Earlier in 2021, the Bank published the scenario to be used in the SST. The scenario incorporates paths for economic and financial market variables, including GDP, property prices and unemployment (Chart 3).

The scenario used for the SST has been designed to assess banks’ end-2020 balance sheets against a severe path for the economy in 2021–25 on top of the economic shock during 2020 associated with the Covid pandemic. It is broadly consistent with the ‘double-dip’ scenario generated in the FPC’s RST of August 2020 and represents an intensification of the macroeconomic shocks seen in 2020.

Between end-2020 and the trough of the stress, UK GDP falls by 9% (Chart 3), with a sharp dip at the beginning of 2021. When combined with the economic shocks already seen in 2020 it implies a cumulative three-year loss over 2020–22 (with respect to the pre-Covid baseline) of 37% of 2019 UK GDP. UK residential and commercial property prices fall by around 33% and unemployment rises by 5.6 percentage points to peak at 11.9%. Certain sectors such as hospitality, leisure, construction and transport are more affected than others. Although GDP growth turns positive later in the scenario and unemployment starts to fall back, output remains below its pre-recession projected path throughout the scenario, and even during the recovery phase confidence remains low. Underlying these projections it is assumed that no new government support measures are introduced, and that existing schemes end in March 2021.

Chart 3: The 2021 SST is a very severe path for the economy, and more so in some respects than the 2019 ACS

Peak-to-trough changes in key variables (and peak in unemployment): 2019 ACS and 2021 SST ^{(a) (b)}

Footnotes

• Sources: Bank of England and Bank calculations.
• (a) The unemployment bars show the peak level of the Labour Force Survey UK unemployment rate.
• (b) The Bank Rate bars show the peak-to-trough change in percentage points.

There is a severe and synchronised slowdown across the world, with global GDP falling by 9.6% (Chart 3). Financial market participants are disappointed by progress in global macroeconomic outcomes, with perceptions of risk increasing. Equity prices in the UK fall 20%. Interest rates remains low, with Bank Rate turning negative in line with the market-implied path of the November 2020 Monetary Policy Report (MPR). Credit risks rise in a number of markets, with investment-grade US corporate bond spreads rising by 140 basis points and high-yield US corporate bond spreads by around 480 basis points. While there are shocks to several financial market variables, there is no separate traded risk scenario like in the ACS exercises.

The macroeconomic scenario for the 2021 SST is considerably more severe than the MPC’s latest central projection from the November 2021 MPR. In that forecast, UK and global GDP continue to recover from the effects of Covid. Cumulative UK GDP losses, relative to the pre-Covid baseline forecast, total 19% of 2019 GDP and unemployment rises only slightly to a peak of 4.5% in that projection.

The macroeconomic stress scenario is also more severe in some respects than the 2019 ACS. In that scenario UK GDP fell by 4.7%, unemployment peaked at 9.2% and Bank Rate rose to 4.0% (Chart 3). The pace of the recovery in the SST is faster, however. More broadly, the relative severity across the different aspects of the scenario is different to previous ACS exercises, reflecting the idiosyncratic nature of the shock related to the Covid outbreak. There is a greater stress on the real economy, including the UK, and movements in asset prices are smaller, consistent with a smaller rise in risk premia in the scenario.

The pandemic has underlined the importance of banks being able to provide credit to UK households and businesses. Banks must conduct the test on the basis that they satisfy the demand for credit throughout the scenario, with lending to the UK real economy expanding by 3.1% in total over the five years of the stress. This reflects an important macroprudential goal of stress testing – to help assess whether the banking system is sufficiently capitalised not just to withstand the stress but also to be able to maintain the supply of credit to the real economy in the face of severe adverse shocks.

Banks are assessed on the basis of IFRS 9 transitional arrangements, against an IFRS 9 adjusted reference rate framework.
As set out when the SST was launched in January, the results of the 2021 SST reflect internationally agreed transitional arrangements for the IFRS 9 accounting standard. Under IFRS 9, which was introduced in 2018, banks provide for expected credit losses on all loans before those losses are incurred, resulting in earlier recognition of losses relative to the previous accounting standard. Arrangements have been put in place to offer banks transitional relief as they adapt to the new standard, with that relief reducing gradually to zero by 2025. Banks participating in the stress test have been assessed on this transitional basis. This means that they have been allowed to ‘add back’ a proportion of capital losses that are associated with earlier recognition of impairments under IFRS 9. In response to the pandemic, transitional relief was extended by the European Union authorities as part of the Capital Requirements Regulation (CRR) ‘Quick Fix’.^{footnote [4] footnote [5]}

Each bank’s performance in the test is assessed against ‘reference rates’ for their risk-weighted CET1 capital and Tier 1 leverage ratios. These reference rates comprise the sum of each bank’s minimum capital requirements and any systemic buffers that they are required to hold, and so are determined in the same way as the hurdle rates used in the 2019 ACS.^{footnote [6]} That they are called reference rates instead reflects that they do not inform the setting of capital buffers, since the SST is aimed at assessing banks’ resilience to risks arising from the pandemic and their ability to lend.

Reference rates have been adjusted to take into account the impact of the IFRS 9 accounting standard, which reduces CET1 at the capital low point by bringing forward the point in a stress at which banks provision for losses. These adjustments follow the approach the Bank adopted in 2018 and 2019, and recognise the additional resilience provided by the earlier provisions taken under IFRS 9. Although participating banks are judged on a transitional basis, for transparency the Bank also calculates and publishes both capital low points and reference rates on an assumed non-transitional basis (see Box B).

Banks began the SST with robust capital positions, in part due to actions taken in 2020.
At end-2020, the start point for the 2021 SST, participating banks had an aggregate CET1 capital ratio of 16.2% of risk-weighted assets and Tier 1 leverage ratio of 5.8% of total exposures (including the software assets benefit). These capital positions have strengthened considerably since the global financial crisis (Chart 1), reflecting post-crisis reforms including higher capital requirements introduced by the FPC alongside other regulatory authorities. And despite banks making substantial provisions for credit losses due to the pandemic, capital ratios rose over the course of 2020. This in part reflects actions taken by banks and public authorities in response to the pandemic. In particular, the cancellation of final 2019 dividends in line with the Bank’s request supported capital ratios by 0.4 percentage points.^{footnote [7]} In addition, the extension of IFRS 9 transitional relief, as described above, provided further support.

Since December 2020, software assets have been included in CET1 and broader regulatory capital, as part of the CRR ‘Quick Fix’. The PRA has stated that from 1 January 2022 the treatment of software assets will be updated, requiring all intangible software assets again to be fully deducted from regulatory capital resources. This change reflects that there is no credible evidence that software assets could absorb losses effectively in stress, and is to ensure that banks’ CET1 capital helps firms to remain a going concern and sustain lending to the real economy.^{footnote [8]} The software assets benefit is therefore excluded in the 2021 SST from 2022 onwards (the system-wide low point occurs in 2022), and, in order to increase comparability of system-wide start and low points in the SST, the change in the major UK banks’ aggregate capital ratio is expressed relative to a start point that also excludes the software assets benefit. Excluding the benefit of software assets, the end-2020 aggregate CET1 capital and leverage ratio start points are 15.9% and 5.7%.^{footnote [9]}

The stress reduces banks’ capital positions substantially, but the system and all eight participating banks remain above their reference rates for both CET1 capital and Tier 1 leverage ratios.
The final results of the SST show that, on a transitional IFRS 9 basis, banks’ aggregate CET1 capital ratio falls to a low of 10.5% by the second year of the stress – a decrease of 5.5 percentage points from the start, excluding the benefit of software assets (Charts 1 and 2). A key driver of this change is substantial credit impairments. The aggregate Tier 1 leverage ratio falls by 1.0 percentage points to a low of 4.8%. If individual banks’ reference rates were aggregated, at the capital low points the UK banking sector would exceed its indicative CET1 capital and Tier 1 leverage reference rates by 2.8 percentage points and 1.1 percentage points respectively (Chart 4 and Table A). The UK banking sector remains above the indicative aggregate reference rates in all years of the stress. There is even greater headroom above minimum requirements.

No individual bank falls below its reference rate for either the risk-weighted CET1 capital or Tier 1 leverage ratio (Chart 4 and Table A). The scale of the drawdown does differ by bank, however, with many of those banks that have a relatively high share of exposures in the UK generally experiencing larger reductions in capital ratios. This primarily reflects the severity of the UK scenario, and is explained further in the next section. There is no conversion of banks’ AT1 instruments in the stress as no bank’s CET1 ratio falls below 7% on a non-transitional IFRS 9 basis. The major UK banks’ aggregate CET1 capital ratio recovers somewhat over the second half of the stress scenario to reach 13.5%. That is, nevertheless, weaker than the starting position.

Chart 4: All banks remain above their CET1 capital ratio reference rates in the SST

Projected CET1 capital ratios in the stress scenario ^{(a) (b) (c) (d)}

Footnotes

• Sources: Participating banks’ STDF data submissions, Bank analysis and calculations.
• (a) See footnote (a) to Chart 1. Aggregate CET1 capital ratios are calculated by dividing aggregate CET1 capital by aggregate RWAs at the aggregate low point of the stress in 2022.
• (b) The minimum CET1 capital ratios shown in the chart do not necessarily occur in the same year of the stress scenario for all banks. For individual banks, low-point years are based on their post-strategic management actions and CRD V restrictions.
• (c) There is no conversion of banks’ AT1 instruments in the stress as no bank’s CET1 ratio falls below 7% on a non-transitional IFRS 9 basis.
• (d) The start points for all firms and the aggregate are shown including the software assets benefit (in contrast to Charts 1 and 2). For banks with a low-point year in 2021, the low-point figures include the software assets benefit, whereas for banks with a low-point year in 2022, the low-point figures exclude the software assets benefit, in line with the regulatory treatment as set out in the PRA’s Policy Statement PS17/21.

3: Key drivers of the results

The results of the SST incorporate a number of key judgements. These judgements inform adjustments Bank staff have made to the participating banks’ projections. The next few sections describe the drivers of the results in more detail.

Key drivers of the weakening in capital positions include substantial impairments and higher risk-weighted assets (RWAs).
Banks’ aggregate CET1 capital ratio falls substantially in the first two years of the stress (Chart 1). A number of factors reduce banks’ capital positions, with other factors cushioning the impact of the stress. Table B decomposes the change in the aggregate CET1 capital ratio and leverage ratio between the start point in 2020 and the low point in 2022 into each of its constituent components. One of the key drivers of the capital drawdown for the banking system is credit impairments, alongside an increase in RWAs (Table B). Net interest income and net fee and commission income are key factors cushioning the impact of the stress.^{footnote [10]}

This decomposition differs from the way the Bank has presented results of previous ACS exercises, which showed the difference in capital impact for each line item, from the start point to the stressed low point year, in the stress scenario relative to the baseline scenario. The baseline scenario was based on a macroeconomic scenario in line with the MPC’s central forecast. This year banks were not asked to submit baseline projections, to help facilitate changes to the usual timetable and in recognition of ongoing operational challenges within participating banks as a result of the pandemic. When presented based on the difference between the start and low point of the stress instead, a number of components in Table B, in particular net interest income and expenses, appear larger than on the usual baseline-to-stress basis. That is because they present banks’ underlying revenues and costs in the two years between the start and low points.

Banks incur credit impairments of more than £70 billion over 2021 and 2022.
A key driver of the material fall in the aggregate capital ratio is credit impairments (Table B). The Bank published interim results of the 2021 SST in the July Financial Stability Report, based on early credit impairment projections from participating banks. Resubmissions since the interim results have not materially changed these results.

Credit impairments stand at more than £70 billion between end-2020 and the low point of the stress in 2022. When the impairments banks actually incurred in 2020 are included, this rises to over £90 billion of impairments between the outbreak of Covid and the 2022 capital low point.

Over the five years of the stress, and excluding 2020, banks incur impairment charges totalling around £90 billion, which corresponds to an aggregate impairment rate of 2.8% (Table C). Around 80% of these are projected to occur by the capital low point.

The majority of these impairments are on UK-based lending.
Around 60% of total impairments over the five years of the stress are incurred on banks’ exposures to UK borrowers (Table C). This is higher than the share of UK lending in banks’ credit exposures, of just over 50%, which in part reflects the severity of the UK scenario. The share of UK impairments is also higher than in the 2019 ACS, which was based on a scenario embodying deep simultaneous recessions in the UK and global economies.^{footnote [11]}

There is a broadly even split between impairments on lending to individuals (mortgages and consumer credit) and lending to corporates (Chart 5). In the UK, the severity of the initial unemployment shock increases defaults on consumer credit exposures in particular, in line with the impact of unemployment as observed in the past. The strong rebound in residential property prices helps to limit losses on mortgages.

Chart 5: Around 60% of impairments over the stress scenario are incurred on UK exposures, and the split of total impairments between retail and corporate is fairly even

Aggregate cumulative impairment charges (and rates) over the five years of the stress (2021–25) ^{(a) (b)}

Footnotes

• Sources: Participating banks’ STDF data submissions, Bank analysis and calculations.
• (a) See footnote (a) to Table C.
• (b) Does not include other wholesale lending.

Impairment losses on UK corporates in the stress are material, at 7.8% of exposures, in part reflecting that debt in the corporate sector increased during 2020. Nevertheless, as explained in October Financial Stability in Focus, these impairments are mitigated by a number of factors. Actions by UK authorities – including the Government and the Bank of England – and support from the financial sector play a particularly important role in limiting losses in the SST.^{footnote [12]} The low level of interest rates in the scenario also limits companies’ debt-servicing costs, and the major UK banks have limited exposures to sectors with particularly low aggregate interest coverage ratios (ICRs).^{footnote [13]} As explained in Box C, while the results of the SST reveal vulnerabilities in some corporate sectors, these are not judged to be a material threat to banks’ resilience.

Impairments totalling £36 billion over the five years are incurred on banks’ non-UK lending to the real economy in the SST (Chart 5). The global low interest rate environment helps to limit corporate impairments by reducing companies’ debt-servicing costs. The impairment rate of 4.0% on non-UK corporate exposures is substantially below that for UK corporates, which first and foremost reflects that UK banks’ exposures to non-UK corporates are more concentrated among larger companies. These are judged to be less at risk of default in part because they generally have more diverse sources of income and greater access to sources of funding such as capital markets.

Given the relative severity of the scenario for the UK economy, many of those banks that have a relatively high share of exposures in the UK experience higher impairment rates, which is a key driver of their generally higher capital drawdowns (Table A). In addition, aggregate impairment rates are relatively high for those banks that are more active in unsecured and corporate lending, since mortgage losses are relatively low in this scenario.

Banks are also judged to be resilient to potential losses on leveraged lending.
The FPC has been monitoring closely the leveraged lending market and the loosening in underwriting standards in recent years (see Financial Stability Report December 2021, ‘Overview of risks to the UK financial system’). Leveraged loans are loans to companies that are highly indebted or owned by a private equity sponsor.^{footnote [14]} As such, they are a risky form of lending which may translate into larger losses in a stress. At the two-year scenario low point in the SST total projected losses on these exposures are estimated to be £3.6 billion. The core UK banking system is therefore judged to be resilient to direct losses associated with leveraged lending.

While substantial, total impairments are limited by the recovery embodied in the scenario.
When the impairments banks actually incurred in 2020 are included, overall credit impairments in the SST between 2020 and 2022 are £90 billion. While substantial, these are lower than the £120 billion implied by the 2020 RST, despite the fact that the FPC and PRC judge both macroeconomic scenarios to be of broadly equal severity.

A key factor explaining this difference is the more granular approach taken by the Bank in assessing the impact of the scenario in this exercise, in conjunction with the detailed credit submissions from participating banks. In particular, the strong rebound in residential property prices later in the scenario is estimated to help limit losses on mortgages in the SST. Indeed, under IFRS 9 losses are recognised before they are incurred. This means that the more the economic recovery reduces impairments later in the scenario, the fewer losses there are to be brought forward.

The impact of the transitional arrangements for IFRS 9 in the stress is relatively limited, compared to the 2019 ACS for instance. That reflects two factors. First, the steep macroeconomic downturn early in the scenario leads to defaults early in the stress, and this effect is accentuated by the increase in loans already classified as at a heightened risk of default during 2020. Those newly defaulted provisions in 2021 and 2022 are not eligible for transitional relief, which means that the amount of losses that banks are allowed to ‘add back’ is commensurately smaller than would otherwise have been the case. Second, as described above overall impairments are limited somewhat by the strong recovery in the scenario, reducing the amount of offsetting IFRS 9 transitional relief. The impact of transitional relief does, nevertheless, vary by bank (see ‘Annex 1: 2021 solvency stress test: bank specific results’ in the annex for more detail).

RWAs increase due to higher risk weights on credit exposures.
RWAs are a risk-adjusted measure of banks’ assets, used to calculate regulatory CET1 capital requirements, and they increase when banks’ underlying risk increases. Aggregate RWAs rise by 24% in the first two years of the stress (Chart 6), contributing substantially to the reduction in banks’ aggregate CET1 capital ratio (Table B). Higher RWAs are accounted for mainly by higher risk weights on banks’ credit exposures, largely reflecting rating grade migration on corporate lending, particularly outside the UK. Although impairments on UK mortgages are judged to be relatively low in the SST as described above, higher risk weights on mortgage lending still contribute materially to higher RWAs due to the impact of higher unemployment. The increase in RWAs overall is smaller than in the 2019 ACS. In the later years of the scenario RWAs fall back somewhat, consistent with the reduction in unemployment and improvement in economic activity.

Chart 6: Aggregate risk weights are projected to rise materially in the scenario

Participating banks’ historical and projected aggregate risk weights ^{(a) (b)}

Footnotes

• Sources: Banks’ published accounts and related public disclosures, participating banks’ STDF data submissions, PRA regulatory returns, Bank analysis and calculations.
• (a) For the period from 2007–11, the aggregate risk weight is defined as RWAs divided by Total Balance Sheet Assets. For the period from 2012–15, the aggregate risk weight is defined as RWAs divided by the Basel III leverage exposure measure, and from 2016 it is defined as RWAs divided by the UK leverage exposure measure (excluding qualifying claims on central banks). RWAs are defined using the prevailing regulatory standard at each date. Virgin Money UK is included from 2018 onwards.
• (b) The figures for 2020 and 2021 include the software assets benefit.

Banks’ overall profits fall sharply in the early part of the stress.
Banks generate substantially weaker profits in the stress than in recent years, making a loss of close to £60 billion in the first year (Chart 7). While that largely reflects credit impairments, banks’ underlying income is also weak. For instance, the income banks earn through fees and commissions for services provided is projected to fall in the first year, reflecting weaker economic activity.

From the second year of the stress, banks’ annual profit before tax returns to close to its pre-pandemic average of approximately £30 billion per year, contributing to the modest pickup in capital ratios. The recovery in profit reflects that the rebound in economic activity in the scenario leads to lower impairments and higher fee and commission income, balanced against the continued drag on net interest income from the low interest rate environment. In the five years of the stress scenario overall, banks are projected to earn around £70 billion in profit, only returning to positive cumulative profit in the fourth year (Chart 7).

Although banks’ expenses contribute to the reduction in capital positions over the stress (Table B), this reflects their underlying costs, and they are lower at the low point of the stress than in 2020. That fall reflects lower bonuses – in line with banks’ bonus policies – and planned reductions in restructuring costs and other staff costs.

Banks’ net interest income is suppressed by low interest rates, although a reduction in expensive wholesale funding provides some support.
The largest component of banks’ income is their net interest income (NII), which they earn by receiving higher interest on assets, like loans, than they pay out on liabilities, like deposits. Overall, although it contributes positively to the change in capital ratios (Table B), NII falls by 3% in the first year of the stress and thereafter remains broadly flat. Non-sterling loan margins – a measure of the spread between average non-sterling loan and deposit rates – fall sharply in the first year, whereas the decline for sterling loan margins is more gradual (Chart 8). These paths for margins are materially weaker than the 2019 ACS, in which interest rates rose.

Chart 8: Both sterling and non-sterling loan margins are lower in the SST than the 2019 ACS

Sterling and non-sterling loan margins in the 2021 SST ^{(a) (b)}

Footnotes

• Sources: Participating banks’ STDF data submissions, Bank analysis and calculations.
• (a) Sterling loan margin calculated as net interest income received on sterling loans minus that paid on deposits divided by sterling loans.
• (b) Non-sterling loan margin calculated as net interest income received on non-sterling loans minus that paid on deposits divided by non-sterling loans.

Falls in policy rates in the scenario in both the UK and abroad from their already low levels weigh on margins early in the stress. This is in part because banks would be unable to pass on these lower rates in full to depositors given that these rates are already only just above zero. The path for margins is further depressed by an assumption that UK mortgage borrowers with standard variable-rate mortgages continue to switch to cheaper fixed-rate products, in part due to the continued price incentive. Mortgage margins are judged over time to reduce somewhat, although remain at high levels historically. Banks’ NII in the stress is also supported by falling funding costs as banks move away from relatively expensive wholesale funding towards cheaper retail deposits, which increase in the scenario consistent with the expansion in lending.

The traded risk impact is relatively benign.
The Bank did not publish a separate traded risk scenario for the SST. While there are falls in asset prices across equity and corporate bond markets, as described in Section 2, they are less severe than in previous Bank stress tests. That reflects a smaller rise in risk premia in the scenario given an expectation from market participants that markets remain functional, with the stress in the real economy relatively more severe reflecting the idiosyncratic nature of the shock related to the Covid outbreak. As a result, the impact of financial market developments in the stress is relatively modest, with trading operations contributing slightly positively to the change in banks’ aggregate capital ratio (Table B).

Banks earn client revenue from their investment banking operations of £39 billion over the first two years of the stress to the low point, somewhat lower than recent years. That is only partly offset by £11 billion of losses from a range of items: losses on their trading books, valuation adjustments on the securities they hold, defaults of major counterparties, and reductions in the fair value of items held on banking books such as bonds within banks’ liquid asset buffers. Separately to the effect on banks’ profits, banks’ RWAs increase by £47 billion, consisting of higher RWAs for counterparty credit risk, market risk and credit valuation adjustment risk. The impact of the stress through all of these channels is smaller than the 2019 ACS.

Aggregate misconduct costs are lower than in recent years.
The SST incorporates stressed projections for potential misconduct fines and other costs beyond those paid or provided for by the end of 2020. Banks are asked to provide stressed projections for misconduct costs that relate to known misconduct issues and have a low likelihood of being exceeded.

Banks have lower exposures to material historical issues, such as payment protection insurance (PPI), than has been the case in recent years. Misconduct charges total £7 billion over the first two years, around half the level in the 2019 ACS.

Strategic management actions are relatively limited overall.
Banks are invited to submit planned strategic management actions that they would take during the scenario in order to boost their capital ratios. These can include actions such as sales of parts of their business but must be consistent with the market-wide stress, be executable in stress and form part of, or be consistent with, the bank’s recovery plan.

At the aggregate level, the scale of these is relatively small in the SST, for instance compared to the 2019 ACS. This largely reflects banks’ robust starting capital ratios, which in part is due to actions banks had already taken to support capital ratios, including cancelling final 2019 dividends in March 2020. Furthermore, by the second year of the scenario the recovery in output is already beginning, and so banks are judged likely to want to continue to hold on to their existing business lines, rather than downsize or sell them, in order to benefit from this recovery.

Banks take action to cut distributions in the stress.
Banks make different types of distributions, such as paying out dividends to shareholders, paying out variable remuneration to staff, or buying back shares. The ability to not pay out or reduce distributions in a stress can support banks’ resilience. When a stress materialises banks can reduce distributions through a variety of mechanisms: established policies that link distributions to earnings, voluntary cuts (including strategic management actions) or automatically as a result of capital regulations. These regulations state that if a bank’s capital falls into its combined capital buffer, it is subject to a limit on the proportion of its profit it is allowed to distribute.^{footnote [15]} The total amount it is allowed to distribute is the maximum distributable amount (MDA), which is a share of a banks’ earnings.

In March 2020 the major UK banks cancelled planned outstanding dividends from the earnings they accrued in financial year 2019, following a request from the PRA to do so as a precautionary step given the unique role that banks needed to play in supporting the wider economy through a period of economic disruption.^{footnote [16]} In December 2020 the PRA judged that banks had the capacity again to make prudent payouts.^{footnote [17]} In the SST banks are assumed to pay out these dividends from their 2020 earnings. That banks pay these 2020 dividends in the SST reflects that banks would take final decisions on dividends in early 2021 before the most severe part of the scenario. Planned share buybacks in 2021 are assumed to be cancelled, however.

Thereafter, banks are assumed to take action to cut distributions. Most banks do not pay dividends from the first two years of the stress. In part this reflects that many banks make losses, especially in 2021, and those banks that do return to profit in 2022 prioritise rebuilding capital. Banks resume dividend payments in the later years of the scenario. Finally, a small number of banks fall into their combined capital buffer – although fewer than in the 2019 ACS – with the resultant MDA restrictions reducing the amount that they can distribute.

The impact of the stress is generally more severe for more UK-focused banks than globally active banks.
The impact and drivers for individual banks in the SST varies somewhat (Chart 4 and Table A). Those banks that have a relatively high share of exposures in the UK generally experience a relatively large capital drawdown. That primarily reflects the severity of the UK macroeconomic scenario and its effect on impairments. While there is a severe slowdown in the global economy in the SST scenario, the effects on those UK banks that have significant exposures overseas depends on the particular markets to which they are exposed. In addition, other non-credit factors contribute to larger falls in capital for UK-focused banks. The annex contains more details on individual banks.

The drawdown in the SST is broadly the same as the 2019 ACS.
Overall, the reduction in the major UK banks’ CET1 capital ratio in the SST – at 5.5 percentage points – is broadly the same as the Bank’s 2019 ACS (5.2 percentage points).^{footnote [18]} Nevertheless, the mix of drivers is somewhat different (Table D). Net interest income provides materially less support to the capital ratio, due to the lower path for interest rates in the SST scenario than the 2019 ACS, in which interest rates rose sharply (Charts 3 and 8). Net fee and commission income is somewhat weaker in the SST, reflecting the sharper drop in economic activity than the 2019 ACS.

Offsetting this are a number of factors that limit the scale of the drawdown in the 2021 SST relative to the 2019 ACS. In particular, impairments are lower (Table C). Despite more severe falls in economic activity in the SST than the 2019 ACS, the effect on impairments is more than offset by the beneficial impact of lower interest rates and the more rapid recovery in output in the SST. Although lower impairments in the SST together with other factors (as described above) imply less favourable offsetting IFRS 9 transitional relief, impairments net of this relief are still somewhat smaller (Table D). Other factors that provide more support to capital ratios in the SST relative to the 2019 ACS are a smaller traded risk impact, a lower increase in RWAs, and smaller misconduct costs.

The capital impact in the SST is also in line with the RST.
The SST has been designed to act as a cross-check on the FPC’s judgements based on the RST carried out in 2020. The aggregate capital drawdown is similar to that exercise (Chart 2), validating those judgements. The drivers are to some extent different, however, reflecting in part that participating banks’ projections have revealed dynamics not fully captured by the aggregate desktop analysis underlying the RST. The drag on capital ratios from credit impairments and higher credit RWAs is smaller, reflecting the sharp recovery in the SST scenario and the Bank’s more granular assessment of the impact of the scenario. Offsetting that are more negative impacts from a range of other factors. For instance, larger shocks to financial market variables are included in the SST than the RST, reducing slightly the gains on banks’ trading operations (even if the impact on traded risk is smaller than the 2019 ACS). And other income banks make in the test is lower than the RST, due to the lower starting level of income at end-2020.

4: Use of the findings of the solvency stress test

The FPC continues to judge that the UK banking system remains resilient to outcomes for the economy that are much more severe than the MPC’s central forecast.
The primary aim of the SST was as a cross-check on the FPC’s judgement of how severe the pandemic-related macroeconomic stress that began in 2020 would need to be to jeopardise banks’ resilience and challenge their ability to absorb losses and continue to lend. The macroeconomic scenario underlying this SST was therefore calibrated in a different way to – and in some respects is more severe than – previous Bank ACS exercises.

The results of the stress test confirm the FPC’s previous judgement about the resilience of the system, in aggregate, to a very severe macroeconomic stress, much more severe than the MPC’s current economic forecast. The FPC continues to judge that UK banks are able to continue to support UK households and businesses even if economic outcomes are considerably worse than expected.

The results show that no individual bank falls below its reference rates (Table A and Chart 4). No individual bank is required to strengthen its capital position as a result of the test. The annex contains more details on individual banks, including Virgin Money UK who participated in the stress test for the first time.

The stress test will not be used as a direct input for setting capital buffers.
As indicated when the SST was launched, the focus of the exercise was to enable the FPC to update judgements about how the banking system could continue to support the economy through the pandemic-related stress. Consistent with the nature of the exercise, the FPC and PRC will therefore not use the SST as a direct input for setting capital buffers for UK banks, including the UK countercyclical capital buffer rate and PRA buffers.

The FPC judges that the results of the test, together with the central outlook and return to the ACS framework for 2022 (see Section 5), are consistent with the PRC’s decision to transition back to its standard approach to capital-setting and shareholder distributions.

The Bank’s qualitative review found continued improvements in data and analysis.
A key objective of the Bank’s concurrent stress-test framework is to support a continued improvement in banks’ own risk management and capital planning capabilities. The experience of 2020 and the Covid outbreak have highlighted the importance of banks having robust internal stress-test processes in place. For that reason, as in previous years, the Bank has undertaken a qualitative review of banks’ stress-testing capabilities as part of the 2021 SST. Nevertheless, the scope of this review was reduced, in order to manage the burden on participating banks.

The Bank found continued improvements in the overall quality of data provided and analysis across a number of areas, especially for those banks where material deficiencies had been identified previously. While participating banks have made encouraging progress overall, the Bank observed some weaknesses in banks’ abilities to assess the impact of the stress on net interest income. In addition, some improvement is required in their articulation of mitigating management actions (even if these actions were limited in scale on an aggregate basis in this exercise, as described above).

Bank staff also found that banks are making good progress overall in addressing the feedback from the Bank’s 2019 assessment of compliance with the Basel Committee on Banking Supervision stress-testing principles.

5: Stress testing in 2022

For 2022, the Bank intends to revert to the ACS stress-testing framework.
Having used the SST to test the resilience of the UK banking system against a much more severe evolution of the pandemic and consequent economic shock, the Bank intends to revert to the ACS stress-testing framework for the coming year. Due to its broader and countercyclical nature, as explained in Box A, the ACS framework is well suited to informing the setting of capital buffers for the system and its core banks.

Further details of the 2022 ACS will be published in 2022 Q1. When designing the ACS exercise, the Bank will reflect on changes that have occurred since the Bank published its approach to stress testing in 2015, and any lessons learnt from the Covid pandemic, including on the implications of IFRS 9. Furthermore, the test will assess the ring-fenced subgroups of existing stress-test participant banks on a standalone basis for the first time.

The Committees

The Financial Policy Committee:

Andrew Bailey, Governor
Jon Cunliffe, Deputy Governor responsible for financial stability
Ben Broadbent, Deputy Governor responsible for monetary policy
Dave Ramsden, Deputy Governor responsible for markets and banking
Sam Woods, Deputy Governor responsible for prudential regulation
Nikhil Rathi, Chief Executive of the Financial Conduct Authority
Sarah Breeden, Executive Director for Financial Stability Strategy and Risk
Colette Bowe
Jon Hall
Anil Kashyap
Elisabeth Stheeman
Carolyn Wilkins
Charles Roxburgh attends as the Treasury member in a non-voting capacity.

The Prudential Regulation Committee:

Andrew Bailey, Governor
Jon Cunliffe, Deputy Governor responsible for financial stability
Ben Broadbent, Deputy Governor responsible for monetary policy
Dave Ramsden, Deputy Governor responsible for markets and banking
Sam Woods, Deputy Governor responsible for prudential regulation
Nikhil Rathi, Chief Executive of the Financial Conduct Authority
Julia Black
Tanya Castell
Antony Jenkins
Jill May
John Taylor

Annex

• The results of the 2021 solvency stress test of the UK banking system: annex