S&P Global Ratings' base-case economic forecast for the U.K. assumes a transition arrangement with the EU through year-end 2020. However, we believe the risk of a disruptive Brexit in March 2019 is increasing as negotiations continue (see "Credit Conditions EMEA: Any Cure For The Trade And Brexit Blues?," published on June 28, 2018). To test the stability of our ratings on U.K. asset-backed securities (ABS) and residential mortgage-backed securities (RMBS) if a disruptive Brexit were to occur, we considered the potential impact of increased credit losses if key macroeconomic variables were to deteriorate.

Under our stressed credit assumptions, coupled with structural features such as sequential note repayment, 'AAA', 'AA', and 'A' rated ABS and RMBS tranches have a cushion to withstand potential deterioration in credit losses from political and economic events. On the other hand, our current ratings on tranches rated 'BBB' and lower could be vulnerable to downgrade in the event of a disruptive Brexit accompanied by a severe economic shock. We will therefore closely monitor developments in the lead-up to March 2019 to assess the impact of potential macroeconomic deterioration on outstanding ABS and RMBS ratings.

Disruptive Brexit Scenario Analysis

To test the stability of our ratings in the event of a disruptive Brexit, we considered an adverse macroeconomic scenario consistent with The Bank of England 2018 annual cyclical scenario stress tests (2018 ACS). This is a hypothetical scenario that we selected solely to test the ratings sensitivity of structured finance transactions. There remains a lot of uncertainty on the economic implications of a disruptive Brexit, which could be more or less severe than the scenario outlined below. In addition, our scenario analysis focused specifically on the expected credit performance of collateral backing U.K. ABS and RMBS transactions. It is important to note that a disruptive Brexit could impact other risk factors, such as counterparty or sovereign risks, but this is outside the scope of our scenario analysis.

According to the 2018 ACS, the peak-to-trough deterioration in key macroeconomic variables over a period of approximately three years are:

• Real GDP falls by 4.7%;
• Unemployment rises to 9.5%;
• Home prices fall by 33%;
• Consumer price index (CPI) inflation increases to 5.1%; and
• The bank rate rises to 4%.

This scenario is approximately between a high 'BB' stress and a low 'BBB' stress under our rating definitions. We believe it would have material negative implications for the creditworthiness of U.K. consumers.

Auto ABS

A disruptive Brexit is likely to create more-challenging conditions for the U.K. auto industry. However, the current performance of U.K. auto ABS transactions remains stable. We also believe that such a Brexit would have less impact on existing deals that are currently amortizing due to rapid deleveraging, the subsequent buildup in credit enhancement, and the time it would take for the consequences of a Brexit-induced recession to crystalize (unemployment would take over a year to peak, by which time a significant proportion of senior notes would have paid down). Although transactions with a revolving period will not benefit from a buildup in credit enhancement, triggers in these transactions generally remain relatively tight and should reduce the potential for any significant deterioration in performance before amortization commences.

Under the 2018 ACS, we believe that increased unemployment and a decline in household income will result in increased auto loan defaults, also known as hostile terminations. We performed a regression analysis between late stage delinquencies, as a proxy for hostile terminations, and unemployment for U.K. auto ABS deals during the 2008-2009 recession, which we view as being a period of similar economic stress to the 2018 ACS. The data suggests a close correlation between unemployment and hostile terminations. Under a scenario where unemployment more than doubles from its current level of 4.0% to 9.5%, our expectation is that our base-case default assumptions for auto deals would at most double. We generally consider our base-case default rate to be consistent with a global scale 'B' stress scenario. As outlined in our criteria, if the economy deteriorates, some level of volatility around the base case can occur without affecting the ratings assigned to securities rated above 'B'. In other words, there is typically some level of cushion, or tolerance, for higher default rates before a rating action would be necessary. The level of tolerance for pool performance deviation away from the base case tends to be higher as ratings move up the scale from 'B' to 'AAA' (see "Methodology And Assumptions For European Auto ABS," Oct. 15, 2015, and "Global Methodology And Assumptions For Assessing The Credit Quality Of Securitized Consumer Receivables," Oct. 9, 2014).

The increase in our base-case default assumptions under the stress scenario would lead to higher stressed default assumptions for tranches rated 'B' to 'BBB', meaning they would likely need a higher overall level of credit enhancement to maintain ratings stability. Stressed default assumptions for 'A' to 'AAA' rated tranches, however, would generally remain constant, as at higher rating levels stressed default assumptions are less sensitive to deterioration in the economic environment. As a result, the disruptive Brexit scenario would likely leave ratings in the 'A' category and above unchanged and lead to downgrades at ratings in the 'BBB' category and below.

Along with an assessment of the impact on hostile terminations, we have also considered a potential decline in recovery values, an increase in voluntary terminations (VT), and a reduction in residual values, which in our view are the primary credit risks in most U.K. auto ABS transactions. Even in the absence of a disruptive Brexit, we forecast a deterioration in secondhand vehicle values in the U.K., which would result in lower recoveries for hostile terminations and increased residual value losses. We see this as being driven by a number of factors, such as an oversupply of used vehicles resulting from a high number of personal contract purchase (PCP) agreements maturing in 2018 and 2019, and also pressure on secondhand diesel values (see "Changing Consumer Behavior Signals A Shift In Risk Profiles For U.K. Auto ABS," July 2, 2018). In addition, lower used-car values could result in borrowers having negative equity in their vehicles, which may create incentives to exercise their VT rights under the U.K. Consumer Credit Act. However, for PCP agreements, we generally view VTs as pulling forward residual value losses that would otherwise be incurred, although the loss severity may be marginally higher because the loans have not fully amortized to the guaranteed minimum future value (GMFV). Therefore, as voluntary terminations increase, we expect less potential residual value loss in transactions with high PCP concentrations.

Under the 2018 ACS stress scenario, our base-case recovery and vehicle market value decline (MVD) assumptions for residual value losses would increase. However, as with hostile terminations, we believe our stressed recovery and MVD assumptions account for the forecast deterioration in secondhand vehicle values and would generally remain stable for the higher rating categories under the ACS, although the distance between the base-case and stress would decline. For ratings in the 'BBB' category or lower, it is likely our stress assumptions would increase to account for the deterioration in our base-case. We would also expect an increase in VTs. For most transactions with PCP agreements, we believe an increase in VTs would be partially mitigated by the credit enhancement currently available to cover residual value losses for PCP contracts that reach maturity. Another mitigating factor would be the expected reduction in new vehicle registrations, which should support used vehicle values and reduce further supply of used vehicles in the medium term.

We also make inferences from our experience of transaction performance-related downgrades during the 2008-2009 U.K. recessionary period. During and after the 2008-2009 recession, we did not downgrade any U.K. auto ABS tranches for performance-related reasons. This was a result of a number of factors, including their short weighted-average life and the rapid deleveraging and buildup of credit enhancement in these transactions, along with the fact that the vast majority of the tranches were rated above the 'BBB' category.

Our overall expectation is that, under the 2018 ACS, U.K. auto ABS ratings from 'AAA' to 'A' would generally remain unchanged. We would expect potential downgrades at the 'BBB' to 'B' rating levels, absent any mitigating factors. The majority of U.K. auto tranches we rate are in the 'AAA' to 'A' range, with only five tranches currently rated 'BBB' or below, so the impact on existing U.K. auto ABS would be limited. New transactions launched after a disruptive Brexit would face higher exposure to stressed economic conditions, which we would consider when setting our base-case and stressed performance assumptions.

Credit Card ABS

In credit card ABS, any adverse movement in one of the three key performance variables (charge-off rate, monthly payment rate [MPR], and gross yield) might be compensated for by positive movement in the other two variables, which may help to maintain the current ratings on the notes. For example, a rise in the charge-off rate might be mitigated by a rise in yield or MPR and a drop in yield might be mitigated by a rise in MPR or a drop in the charge-off rate. Credit card ABS structures also typically have asset performance, seller insolvency, and servicer-default-related amortization events that, if triggered, will cause the amortization period to begin.

Charge-offs of credit card receivables are positively correlated with unemployment, bankruptcies, and individual voluntary agreements (IVAs). Our base-case assumptions consider the performance of the receivables through an economic cycle, so there is some scope for performance to deteriorate from current levels without any upward revision in charge-offs. Our current base-case assumptions for charge-offs range from 5.5% to 7.5%, while actual charge-offs generally range from 2% to 4%. However, the deterioration in key macroeconomic variables under the 2018 ACS is beyond what we would consider normal business cycle conditions, and would therefore likely lead to an increase in our base-case charge-off assumptions. During and following the 2008-2009 recession, charge-offs in our U.K. credit card index peaked at around 10%. Given the comparable level of stress in the 2018 ACS, we believe our base-case assumptions could increase by five percentage points. Our peak charge-off assumptions for U.K. credit card portfolios associated with a 'AAA' stress reflect our view of performance under conditions of extreme economic stress, which we consider a fixed anchor point; meaning that we would be unlikely to adjust these assumptions in the disruptive Brexit scenario. Therefore, we would expect to increase our 'B' charge-off assumptions in the 2018 ACS, while 'AAA' assumptions would remain constant.

We believe that an increase in charge-offs would lead to tighter credit conditions, improving the credit quality of any new account originations. If designated to the securitization trusts, this could support receivables performance. In addition, in our view, the increase in the bank rate to 4% under the 2018 ACS would result in a discontinuance of balance transfer offers, which have been one of the primary sources of receivables generation in recent years. We would expect the existing receivables in portfolios on promotional balance transfer offers to naturally amortize, which could result in higher payment rates if borrowers choose to repay balances in full. Alternatively, those receivables that do not repay in full at the expiry of the promo periods would begin to accrue annual interest, which would increase portfolio yield to mitigate any increase in charge-offs. Notably, during the 2008-2009 recession, payment rates and yields generally remained stable. We therefore believe that our base cases and stresses for these variables would be unlikely to change under the disruptive Brexit scenario.

Lastly, our purchase rate assumptions--or the rate at which new receivables are created as cardholders use their credit cards to make purchases or cash advances--are linked to the unsecured credit rating on the originator. If originators are downgraded due to a disruptive Brexit, we may reduce our purchase rate assumptions. This could result in a rating impact for all rated tranches absent any mitigating factors such as available levels of enhancement being in excess of the minimum required amounts to support the current ratings. For the purpose of this sensitivity analysis, a change in originator ratings are outside of the scope.

Overall, we believe a disruptive Brexit would have an impact on new credit card originations and likely on the collateral composition of securitized portfolios as balance transfer receivables run off. We expect that our base-case charge-off assumptions would increase, although there remains a cushion in our current assumptions versus actual portfolio performance. We believe our peak charge-off default assumptions at the 'AAA' level would remain unchanged. Payment rates would be expected to remain stable, while yields may increase as promotional periods expire, which could mitigate any increase in charge-offs. We therefore would not expect any rating impact at higher rating categories, but 'BBB' and lower rated tranches could be more vulnerable to downgrades from an increase in our charge-off assumptions.

RMBS

For RMBS, as with ABS, the application of 2018 ACS stress leads to potential downgrades only for low-investment-grade and speculative-grade credits. The key RMBS performance metrics of delinquencies, defaults, and loss severities would inevitably increase in such a scenario, but we do not expect these increases to be significant enough to see material ratings migration. In addition, any ratings impact would likely be felt gradually, rather than immediately.

The resilience of tranches currently rated in the 'AAA', 'AA' and 'A' categories reflects the fact that higher rating categories already factor in shocks of a similar nature and magnitude of the 2018 ACS. RMBS tranches rated 'BBB' and below are more vulnerable to large uncertainties or major exposure to adverse conditions such as a disruptive Brexit. For example, we determine our U.K.-specific residential property MVDs according to the current indexed valuation of a property. As house prices fall, MVDs applied at higher rating categories reduce, reflecting that part of the anticipated decline has already occurred. The tables below show the estimated ratings impact by asset subclass (see Appendix for the same data expressed as percentages instead of number of tranches).

The 2018 ACS would likely cause a day-one increase on our weighted-average foreclosure frequency (WAFF) assumptions for most U.K. RMBS transactions, reflecting the transition of the U.K. residential mortgage market a non-benign economic state from a benign one, but we would expect immediate ratings changes to be minimal. This is because house price declines and unemployment increases are expected to manifest themselves over a timeframe of months or years rather than immediately. For most transactions, especially ones with large exposures to seasoned repayment loans, higher default and loss expectations will be offset by deleveraging as borrowers pay scheduled principal or prepay either partially or by remortgaging (see "Our analytical approach," below, for more details on how we applied the stresses).

Table 1

Table 2

Table 3

Table 4

The above tables show implied ratings migration two years after a disruptive Brexit. On day one of the disruptive Brexit, we anticipate increases in default and loss estimates, but the day-one anticipated ratings migration is limited to tranches currently rated 'BB' and 'B', with migration on higher rated tranches occurring only as house price declines materialize. Immediately after a disruptive Brexit, the average rating migration of a tranche currently rated 'BB' would be a 0.5 notch downgrade. For a tranche rated 'B' the average would be a downgrade of 2.0 notches.

Two years on from a disruptive Brexit, average ratings migrations for tranches currently rated 'AA' and 'A' would be downgrades of 0.4 notches and 0.66 notches, respectively. Individual ratings that move between rating categories at the 'AA' and 'A' levels are typically tranches moving to an implied 'A+' rating from 'AA-' currently or to an implied 'BBB+' rating from 'A-' currently.

The 2018 ACS stress is comparable to the financial crisis, except interest rates rise

The U.K. RMBS sector has already endured a period of significant stress during and after the 2008-2009 global financial crisis. During this time, the U.K. saw a rise in unemployment to approximately 8.4%, according to the Office of National Statistics, and a national peak-to-trough house price decline of approximately 19%, according to the Nationwide Building Society. These compare to the 9% unemployment spike and the 33% house price fall we have assumed for this analysis. From September 2007 to March 2010, the U.K. prime RMBS index saw 90-day-plus delinquencies (excluding repossessions) rise to 2.0% from 0.6% and the U.K. nonconforming index rise to 18.7% from 7.1%. The BoE stresses anticipate the BoE base rate behaving in an opposite fashion to that in the aftermath of the crisis, when interest rates fell to 0.50% in March 2009 from 4.50% in October 2008. If rates rise to 4% from today's level of 0.75% in the event of a disruptive Brexit, it would have a varying impact on RMBS transactions depending on whether borrowers are paying fixed rates during the peak stresses. It is worth noting that U.K. mortgage regulation, introduced since the crisis, requires lenders to assess borrowers' affordability using a stressed rate.

The impact on buy-to-let is difficult to predict

The buy-to-let (BTL) sector is currently facing headwinds in the U.K, which make it especially hard to assess the fallout of a disruptive Brexit for this sector (see "New U.K. Buy-To-Let Mortgage Rules Will Change The Face Of The Lending And Securitization Markets," published Nov. 27, 2017, and "Buy-To-Letdown? Recent RMBS Loans Will Struggle With The Perfect Storm Of Regulation And Tax Hikes," published Jan. 24, 2018). After the crisis, interest-rate reductions and higher deposit requirements that pushed aspiring buyers into the rental sector supported landlords' profitability. Although it is feasible that rental demand could hold up despite a disruptive Brexit, the prospect of a higher rate environment may be challenging for BTL borrowers on variable rates. The recent trend toward longer-term fixed-rate loans for BTL landlords may also provide a level of protection to parts of the sector.

Actual rating performance would vary widely in a disruptive Brexit

Our analysis looks at estimated ratings migration for rated U.K. RMBS transactions as a whole. Actual rating performance after a disruptive Brexit will be varied, with different transactions and tranches having different sensitivities to different variables. For example, borrowers with fixed rates will not be subject to an immediate payment shock caused by increasing interest rates, whereas borrowers on standard variable rates are likely to see higher monthly payments right away. Additionally, borrowers with a high exposure to interest-only loans will be more vulnerable to shocks compared to transactions with large proportions of repayment loans as the benefits of borrower and transaction deleveraging will not be as significant. Our analysis does also not take into account servicers' strategies in response to a stress scenario, which would also have an impact on actual rating migration.

We expect prepayments to decline gradually and recover faster than after the 2008 crisis

We would expect a gradual decline in prepayment rates from the point a disruptive Brexit becomes likely. Prepayment rates in U.K. RMBS transaction are influenced by a plethora of factors, some specific to individual borrowers and others a product of the wider economic environment. Falling house prices will increase the number of borrowers in low or negative housing equity, and some of these may be unable to refinance if lenders tighten their criteria at the same time. In addition, if the mortgage-lending market becomes less competitive as lenders take flight from a sector where security values are falling, credit spreads on mortgages will rise. This would leave borrowers that have the ability to repay due to their still-low loan-to-value ratio with less of an incentive to do so.

However, improvements in U.K. mortgage regulation since the global financial crisis mean that we are less likely to see U.K. transactions that are significantly exposed to mortgage products that exhibit material risk layering and products that became either de-facto illegal (self-certification) or rationed (interest-only). Consequently, any contraction in prepayment rates is more likely to be linked solely to house prices and lending appetite rather than a wholesale shift in the regulatory environment. For this reason, we would expect prepayment rates to recover to their levels prior to the disruptive Brexit sooner than they did in the financial crisis.

Our analytical approach

We applied the 2018 ACS stresses as follows. On day one of a disruptive Brexit, we would likely change our outlook for the U.K. residential mortgage market to non-benign scenario from benign. We interpret the disruptive Brexit stress as the 'B' WAFF for an archetypal loan increasing to 2.00%. In addition, we apply an increase in projected arrears owing to our anticipation of a 4% increase in employment. Both these penalties result in a higher WAFF. One year after a disruptive Brexit, we have assumed house prices have fallen 16.5% from current levels and unemployment has further increased by 1.5%, meaning unemployment has reached approximately 9.5%. Two years after Brexit, we assume that house prices are a further 16.5% down from their peak, meaning a total drop of 33% from current levels.

To determine whether expected losses could lead to downgrades, we compared the erosion of credit enhancement implied by the increase in default and loss severity expectations at each rating level to credit enhancement available for each tranche of each transaction. We performed this comparison at three periods: day one of a disruptive Brexit; the end of year one; and the end of year two.

Related Research

• Help The Aged: The Changing Landscape For U.K. Borrowers In Retirement Creates Risks And Opportunities For Lenders, June 1, 2018
• Buy-To-Letdown? Recent RMBS Loans Will Struggle With The Perfect Storm Of Regulation And Tax Hikes, Jan. 24, 2018
• Third Of Interest-Only Borrowers In Nonconforming RMBS Deals Failed To Make Bullet Payment At Maturity In Past 18 Months, Dec. 18, 2017
• New U.K. Buy-To-Let Mortgage Rules Will Change The Face Of The Lending And Securitization Markets, Nov. 27, 2017

Related Criteria

• S&P Global Ratings Definitions, April 19, 2018
• Methodology And Assumptions: Assessing Pools Of European Residential Loans, Aug. 4, 2017

Appendix: U.K. RMBS Rating Transitions By Percentage Of Tranches

Table 5

Table 6

Table 7

Table 8