The U.K.'s seven major financial institutions have passed the Bank of England's (BoE) latest annual stress test. In the results published Dec. 16, 2019, each firm cleared the required capital and leverage hurdles, with none required to revise its financial plans or shareholder distribution policies. The exercise subjected the institutions' balance sheets to a severe but plausible scenario that resulted in material credit impairments, trading losses, and misconduct charges. The seven firms' aggregate impairment losses were £151 billion over the five-year stress period, 80% of which were in the initial two years.

S&P Global Ratings believes the results confirm the large institutions' resilience to potential downside risks, both domestic and international. This balance sheet profile underpins our stable trends on the sector's economic and industry risks and our stable outlooks on five of the seven groups that participated in the stress test (see chart 1).

Chart 1

As it has with previous versions, the BoE calibrated this year's stress test to be more demanding than the 2008 global financial crisis. In addition, it judges that the U.K. domestic stress assumptions encompass the potential outcome of a "worst case" disorderly Brexit. We consider that the near-term risk of a disruptive Brexit has diminished following the amended Withdrawal Agreement reached with the EU and last week's U.K. general election outcome. However, ratification of the Withdrawal Agreement will not mark the end of the Brexit process because the U.K. will then begin to negotiate replacement trade arrangements (see "Credit Conditions EMEA: Low Growth, Lower Rates," published on Dec. 3, 2019).

The stress test results were not a surprise to us given that the participating institutions are comfortably positioned relative to current regulatory capital and leverage requirements, including buffers. While their balance sheets have strengthened materially, U.K. banks' earnings remain pressured by sustained low interest rates, mortgage price competition, and the subdued economy. As a result, banks will likely need more time to meet their return on equity targets and some plan to announce targeted restructuring measures in the new year.

We revised the outlook on HSBC Holdings PLC to negative from stable last month because its revised strategic plan may weaken our favorable view of its business stability and diversification (see "HSBC Holdings PLC Outlook Revised To Negative On Accumulating Threats To Earnings Prospects; 'A/A-1' Ratings Affirmed," published on Nov. 12, 2019). Although we expect to retain stable outlooks on most U.K. institutions, we could take similar rating actions on any firms that struggle to adapt to current operating conditions.

Stress Test Outcome Indicates Robust Capital And Leverage Profiles

The common equity Tier 1 (CET1) low point of the stress is 9.3%--a little lower than last year's 9.7%, reflecting the similarity of the starting points and marginally harsher stress assumptions (see chart 2). The starting CET1 ratio is now stable since banks have achieved their regulatory capital and leverage targets and therefore have no immediate need to accumulate more capital. The final Basel III reforms will likely require some further capital building in the longer term, depending on their final calibration.

Chart 2

The main elements that contribute to the downturn in the aggregate CET1 ratio at year-end 2020, the low point during the stress period, are:

• Loan impairment charges of around £121 billion (lowering the CET1 ratio by 5.7 percentage points), split roughly evenly between U.K. lending and overseas lending;
• Traded risk losses, lowering the CET1 ratio by 2 percentage points;
• Risk-weighted asset inflation, lowering the CET1 ratio by 1.7 percentage points;
• Misconduct costs of £13 billion, lowering the CET1 ratio by 0.7 percentage points; and
• Other changes, lowering the CET1 ratio by 0.3 percentage points.

These effects are partly mitigated by:

• Cuts to discretionary distributions, raising the CET1 ratio by 2.2 percentage points--the bulk of this arises from cuts to ordinary dividends;
• Increased net interest income, raising the CET1 ratio by 1.3 percentage points;
• Transitional relief under International Financial Reporting Standard (IFRS) 9, raising the CET1 ratio by 1 percentage point;
• Lower expenses and taxes, raising the CET ratio by 0.9 percentage points; and
• Additional Tier 1 (AT1) conversions, raising the CET1 ratio by 0.6 percentage points.

Chart 3

IFRS 9 transitional relief arises because the BoE, like other regulators, is phasing in the application of IFRS 9 impairment rules, which require a forward-looking assessment of asset quality. This is the second BoE stress test conducted under IFRS 9, and it again contributed to the early recognition of credit losses. Of the £151 billion aggregate impairments incurred under the stress scenario, 80% arose in the first two years.

The BoE also disclosed results without IFRS 9 transitional relief, though the banks were not assessed against this metric. On this nontransitional basis, the results show that all the banks remained above their CET1 hurdle rate, and Barclays and Lloyds fell below their Tier 1 leverage ratio hurdle rate. The nontransitional results do not reflect the fact that the BoE is still considering a long-term framework to set hurdle rates following the phase-in of IFRS 9 . As such, in our view these nontransitional results are not, in themselves, any indication of relatively weaker or stronger capital positions of the banks in the stress test.

Everyone Passed, With AT1 Conversions For Some

While all of the banks passed the stress test even without converting AT1 instruments into common equity, the stress test still shows AT1 conversions (see chart 4). This is because under the specific contractual terms of banks' AT1 instruments currently in issue, their conversion is based on a definition of CET1 that excludes the benefit of the CET1 transitional arrangements for IFRS 9. For two banks, Barclays and Lloyds, the CET1 ratio was below 7% (Barclays, 6.5%; Lloyds, 6.1%) before AT1 conversions in the stress test on a nontransitional basis. As such, their AT1 instruments convert into CET1. We rate the AT1 instruments issued by Barclays PLC and Lloyds Banking Group PLC at 'B+' and 'BB-', respectively, which in both cases is six notches below their group stand-alone credit profile.

Chart 4

Nationwide is an outlier in terms of the material percentage point impact of the stress scenario on its CET1 ratio (see chart 5). This reflects its point-in-time capital models for residential mortgages, which it expects to replace in 2020. Nationwide also has the largest post-stress CET1 buffer over the hurdle rate, which is largely because the leverage ratio is its principal regulatory constraint.

Chart 5

The stress test's effect on the aggregate leverage ratio was primarily from higher loan impairment charges, which reduced the ratio by 1.5 percentage points. At the low point of the stress, Barclays and Santander UK were closest to their hurdle rate, exceeding it by 0.17 percentage points (Barclays) and 0.23 percentage points (Santander UK). (See chart 6.)

Chart 6

The stress test's modeled impairment charge rates on U.K. lending are disaggregated by the type of lending for each bank in the test (see table 1). To put these stressed losses into context, the test estimates five-year cumulative loss rates of just under 28% for U.K. consumer credit, for example. This compares to the actual figure, by our calculations, of about 9% over the five years to end-2018. Our base-case forecast for losses on consumer credit averages 2.7% per year over the three years from 2019 to 2021 (see "Low For Longer: Our Credit Loss Estimates For U.K. Banks," published on Dec. 5, 2019).

For consumer credit, HSBC has the lowest domestic impairment charge over the stress scenario period (see table 1), and it also performs the best in U.K. mortgage lending. Barclays and Lloyds have the highest stressed loss rates for consumer credit, the former partly reflecting its large U.S. credit card book, and Lloyds ranks among the weakest across all asset classes.

We believe that the BoE's data and rankings are extremely useful as a guide to potential relative performance by banks. Yet we caution against taking away too many precise conclusions.

Table 1

The BoE estimates that the stock of global leveraged loans has increased by 30% since 2015 to $3.4 trillion, of which about 4% is held by major U.K. banks. Given this meaningful share and weaker underwriting standards across leveraged lending, the stress test includes a focus on losses arising from leveraged loan exposures. Total leveraged loan exposures held by banks in the test amounted to £90 billion, with stressed losses of £9.7 billion over the first two years of the stress. This equates to a loss rate of 11%, or a 0.4% decrease in CET1. The loan book element of leveraged loan exposures (in other words, hold positions) amounted to £76 billion, with cumulative losses of £9 billion over the five years of the stress, a loss rate of 12% (compared with 8% during the global financial crisis; the higher assumed loss rate reflects the sharp rise in rates in the stress scenario and the looser underwriting standards today). The £9 billion of losses represent just under 13% of total corporate losses in the stress.

Changes In The Countercyclical Capital Buffer And Pillar 2A Will Increase The Quality Of Capital

The BoE announced an increase in the countercyclical capital buffer for domestic exposures, from 1% to 2%, to take effect from next year. At the same time, it will propose changes to reduce banks' Pillar 2A requirements by 50% of the increase in bank-specific buffers. The recalibration will also allow the BoE to vary capital requirements with more flexibility in response to changes in economic conditions.

The net effect of these changes would be a small increase (35 basis points) in aggregate Tier 1 capital requirements. The changes will also result in a greater proportion of capital being comprised of CET1 than currently, because the buffer requirements must be fully met by CET1, whereas Pillar 2A requirements can be met by a combination of CET1, AT1, and Tier 2 capital.

Stress Test Based On A Severe But Plausible Downside Scenario

The stress test results reflect a synchronized global recession, coupled with a traded risk stress and elevated misconduct charges. The BoE required the participating institutions to model materially adverse moves in key variables including economic growth, household and corporate incomes, and asset prices (see chart 7). The domestic stress assumptions are broadly similar to the 2018 exercise. The assumptions are slightly more severe in the U.S. and eurozone because the BoE judges that vulnerabilities have increased in those regions. This is due to higher corporate sector leverage in the U.S. and pockets of high public debt in the eurozone, coupled with interlinkages between banks and sovereigns. The stress scenario also incorporated a fall of almost 8% in Hong Kong GDP and falls in Hong Kong property prices of more than 50%.

Chart 7

The domestic stress assumptions include sharp changes in unemployment and interest rates (see charts 8 and 9). The BoE's "worst case" disorderly Brexit scenario assumes a larger fall in GDP than the domestic stress test scenario but a smaller rise in unemployment. Overall, the BoE sees the macroeconomic severity of the two scenarios to be similar. Of course, the stress test scenario additionally assumes a severe downturn in international economies and financial markets, as well as stressed misconduct costs. Therefore, the aggregate impact of the stress test on banks' capital positions is greater than the effect of the U.K.-specific disorderly Brexit scenario.

Chart 8

Chart 9

The traded risk scenario assumed widespread price shocks, a deterioration in counterparties' creditworthiness (including seven defaults), and stressed revenue and cost projections for investment banking businesses. Leveraged loans were a particular focus given weaker lending standards in that market. The BoE assumed that leveraged-loan price indices would decline by larger amounts than in the global financial crisis, including a 40%-41% fall over a one-year horizon. Similarly, the spreads on speculative-grade corporate bonds were assumed to spike up under the stress scenario (see chart 10).

Chart 10

Lower Misconduct Costs Reflect The End Of The PPI Saga

The stress scenario included a misconduct test that projects additional costs of £17 billion over the five years to end-2023 (with the bulk of this, £13 billion, realized by end-2020).

The aggregate figure of £17 billion is £8 billion lower than in last year's results. This reflects the fact that a larger proportion of misconduct costs had already been provisioned for in banks' balance sheets at 2018 year-end. The BoE has also stated that its stressed projections take into account the significant uptick in misconduct provisions related to payment protection insurance (PPI) claims that banks have incurred so far in 2019, in relation to the August 2019 final deadline for making PPI claims.

While the impact of the misconduct stress has not been disclosed on a bank-by-bank basis, the BoE has stated that the test targets a "high level of confidence that banks will settle at or below the stressed misconduct projections."

Hurdle Rates Adjusted For IFRS 9

The hurdle CET1 ratio in the stress test was the sum of the 4.5% minimum Pillar 1 requirement, the institution-specific Pillar 2A buffer, the global or domestic systemic importance buffer, and an IFRS 9 adjustment. The hurdle leverage ratio similarly incorporates the 3.25% minimum requirement, applicable buffers, and an IFRS 9 adjustment. The purpose of the IFRS 9 adjustment is to avoid requiring additional capital purely due to a change in loss recognition for accounting purposes.

Like last year, the BoE asked participating firms to assume perfect foresight in predicting future economic and market data for the purpose of IFRS 9 provision calculations. This is clearly unrealistic, but it would be difficult to conduct the stress test using a different approach.

Qualitative Assessment Added To Public Stress Test Results

This year's stress test results include for the first time a qualitative assessment of the participating institutions' stress-testing capabilities. That said, the disclosure appears to be very limited, and has not been provided at a bank-specific level. This means that, for now at least, the BoE's views remain private about banks' risk and capital management processes--an important factor in determining banks' undisclosed Prudential Regulation Authority buffers (the U.K. equivalent of the European Central Bank's Pillar 2G buffer).

We note that qualitative assessments are an established element of the U.S. Federal Reserve's stress-testing process, in which there have been several cases of banks passing the quantitative aspect but failing the qualitative review.

Future Direction Of The BoE Stress Tests

The scope of banks in next year's test will be expanded. Virgin Money UK PLC will take part for the first time because it meets the qualifying threshold following the 2018 merger of the CYBG and Virgin Money businesses. Also next year, the stress test results for ring-fenced bank subgroups will be published alongside those of their consolidated parent groups. The BoE has indicated that the focus of the 2020 stress test will adopt a different approach than those of 2018 and 2019.

All else equal, we note that the increase in the countercyclical capital buffer and the proposal for lower Pillar 2A requirements would make the stress tests easier to pass. This is because the CET1 hurdle rate in future stress tests would be lower, and the proportion of CET1 in banks' total capital would be higher.

In addition to cyclical stresses of U.K. banks' capitalization, the BoE uses stress tests to identify potential industrywide vulnerabilities to other types of risk. We see this as an important element of the regulatory toolkit in view of the increased prominence of nonfinancial risks. The BoE's biennial exploratory scenario (BES) assesses how the banking sector's collective response to an adverse event could affect the broader financial system and macroeconomy. The first BES in 2017 focused on the threat to banks' profitability and business models from new entrants and prolonged low interest rates. The 2019 BES looks at a severe liquidity stress, with the conclusions due for publication next year, and the 2021 exercise will examine the physical and transition risks of climate change. The BoE has additionally run a cyberstress test pilot this year looking at the consequences of a payment system interruption.

These moves indicate that BoE's stress testing approach is evolving in response to the changing face of the U.K. banking sector and the risks it faces. In particular we observe a growing focus on nonfinancial risks as well as the more traditional focus on credit risks by asset class, for which stress test data work is now well established.

Table 2

Related Research

• Banking Industry Country Risk Assessment: United Kingdom, Dec. 5, 2019
• Low For Longer: Our Credit Loss Estimates For U.K. Banks, Dec. 5, 2019
• Credit Conditions EMEA: Low Growth, Lower Rates, Dec. 3, 2019
• HSBC Holdings PLC Outlook Revised To Negative On Accumulating Threats To Earnings Prospects; 'A/A-1' Ratings Affirmed, Nov. 12, 2019
• Countdown To Brexit: No-Deal Risks Revisited, Oct. 4, 2019
• Capital Strength Helps U.K. Banks To Weather Additional PPI Charges, Sept. 10, 2019
• Limbo State Lingers For U.K. Banks, Aug. 28, 2019
• Everyone Passed: Stress Tests Highlight Growing Resilience Of U.K. Banks, Nov. 29, 2018