IFRS 9 | The Mortgage Charter, SICR and default

In this post we look at the recently announced Mortgage Charter and think through how the various measures might be treated for IFRS 9 staging and regulatory default. The Mortgage Charter was announced at the end of June 2023 with the support of the UK Government, the FCA, and lenders representing over 85% of the mortgage market. It outlined a set of standards that lenders will adopt to help borrowers who are worried about higher mortgage rates.

Similar to payment holidays provided during Covid-19, the key measure in the Charter - the option to reduce payments to interest only level for six months - is available for up-to-date customers, with no affordability assessment, so lenders may not know if a customer taking up one of the measures is in financial difficulty. This means that the measures should not be flagged as forbearance – they are not tailored to a specific customer and a change to a loan’s terms cannot be forbearance if there is no evidence of financial difficulty.

The guidance around payment holidays and treatment for statutory and regulatory reporting issued by the PRA in June 2020 can still be considered relevant, i.e., that a customer taking up a treatment from the Charter should not automatically be transferred to Stage 2 or Stage 3/default. Nevertheless, using information they have available, lenders should consider whether any such cases have experienced a Significant Increase in Credit Risk (SICR) for Stage 2 or have hit an “unlikeliness to pay” trigger for Stage 3/default.

We anticipate lenders’ response to the Charter to be similar to that for Covid payment holidays i.e., following the PRA’s holistic framework for the consideration of SICR and Stage 3/default. During Covid many lenders used proxy data to identify higher risk cohorts of those with a payment holiday who might have experienced a SICR, but far fewer moved any payment holidays to default without observed evidence of unlikeliness to pay.

So far there has been little change in flow to delinquency and mortgage forbearance balances have been falling. However, the current rate increases many customers are facing on their mortgage refinances are getting a lot of press, and regulators and politicians (and mortgage holders) are clearly concerned. Although demand for the measures in the Charter will likely build over the autumn, lenders have some time to get ready for a ramp-up and have been through this before – the very sudden demand for payment holidays was good training, with many of the issues banks need to address for the Charter being very similar. Lenders should however remain conscious of the challenges faced not only in implementing the measures, but in recording and monitoring the performance of customers who have utilised them.

Background

In recent years the UK mortgage market has become dominated by loans where borrowers typically lock into a fixed interest rate for between two and five years. After this introductory period the loans move to a floating Standard Variable Rate (SVR), normally a couple of percentage points above base rate. However, rather than staying on the SVR, borrowers will typically refinance onto a new fix rate. UK finance statistics show that since 2013 over 90% of new mortgage originations have been on fixed rates and so are 74% of current homeowner contracts.

The rapidly rising rate environment has already led to a significant increase in refinancing costs for many UK mortgage borrowers – a situation summarised by the FSA in their recent policy statement Mortgage Charter Enabling Provisions (PS23/8):

“Over the next 12 months, 1.7 million fixed rate deals will expire, with 893,000 expiring before the end of this year. Many of these fixed rate deals were agreed at a time when interest rates were at an historic low. The median interest rate for deals expiring in Q3 2023 is 2.07% and in Q4 2023 is 1.95%. In contrast, median rates for new deals today are currently above 6% and expected to rise further. Over the next year we expect many borrowers will face a significant increase in their monthly mortgage payments. We project that the biggest increase in the number of ‘financially stretched’ borrowers will occur between July and September 2023.”

Source: Quoted household interest rates - Bank of England

The Mortgage Charter outlined a set of standards that lenders will adopt to help borrowers who are worried about higher mortgage rates. These standards were then brought out in more detail in the FCA’s PS23/8, which sets out the enabling provisions for the Charter.

The applicability of the charter and key features are as follows:

• Allows customers to switch to a new mortgage deal at the end of their existing fixed rate deal without an affordability check and allows customers to lock into a new deal up to six months ahead (and change this if a better deal becomes available).
• Allows a temporary reduction from a capital repayment basis to an interest only repayment basis without need for an affordability assessment, for a period of six months. There is an expectation that full capital payments will resume thereafter, including making up for the reduced payments over the remaining term of the mortgage. For borrowers who are unable to afford increased payments after the six-month period lenders are expected to be ready to provide tailored support
• Lenders can already allow borrowers to extend their mortgage term up to retirement without an affordability assessment. The Charter allows term extensions to be reversed to the original term without an affordability assessment if this happens within six months of the term extension having taken effect. Borrowers cannot make repeat use of this option. Where terms may be extended past retirement, an affordability assessment will still be required.
• Allows these options to be taken by customers without affecting their credit score.
• Lenders have committed not to repossess a property within 12 months of a missed payment except in exceptional circumstances.

Are the measures in the Charter forbearance?

For the purposes of prudential reporting, forbearance is well defined in FINREP: “concessions towards a debtor that is experiencing or about to experience difficulties in meeting its financial commitments (“financial difficulties”)” [para 240 reg 680/2014] with a “concession” referring to “a modification of the previous terms and conditions of a contract that the debtor is considered unable to comply with due to its financial difficulties (“troubled debt”) resulting in insufficient debt service ability and that would not have been granted had the debtor not been experiencing financial difficulties” [para 241a reg 680/2014]. Conversely “exposures shall not be treated as forborne where the debtor is not in financial difficulties” [para 252, reg 680/2014].

Because the borrowers who are in-scope of the Charter are up-to-date, and because there is no affordability test for the measures in the Charter, it will be very difficult for firms to determine if any given customer is in financial difficulty when being granted one of the measures and, therefore, forborne. This means that without further evidence to demonstrate the customer is in, or is about to be in, financial difficultly, the extension of these measures should not automatically lead to a forbearance classification being adopted.

However, there is an exception to this in para 254 regarding recent arrears: “there is a rebuttable presumption that forbearance has taken place in any of the following circumstances: (a) the modified contract was totally or partially past due be more than 30 days (without being non-performing) at least once during the three months prior to its modification…” which may support flagging the position as forborne.

Of course, at the end of the six-month capital payment holiday, if a customer cannot resume full capital and interest payments or experiences payment difficulties while in receipt of one of the measures, we anticipate that “normal” procedures would apply.

In summary, much like the situation observed during the Covid pandemic, these measures should not automatically be classified as forbearance. Lenders should however remain mindful that other triggers that may indicate financial difficulty of the borrower should still be assessed.

Should the measures in the Charter be classed as default for regulatory reporting or Stage 2/3 for statutory reporting?

The “Dear CEO” letter from Sam Woods published on 4 June 2020 relating to the default and staging of payment holidays can still be considered relevant given, as highlighted above, the similarity of the situation to Covid payment holidays. The main principles that read across to the Charter are:

• Counting of days past due should be on the basis of the agreed (rather than the original) schedule
• Use of a treatment should not on its own automatically result in a loan being (a) regarded as having suffered a SICR or being credit impaired for Expected Credit Loss (ECL), or (b) triggering a default under the CRR
• Although it would not be appropriate for all firms to assume all loans subject to or eligible for payment deferrals have suffered a SICR or are credit impaired, it is unlikely to be the case that none of these loans have been so affected; some should be in stage 2 or 3
• Firms are likely to have limited information for making borrower-specific assessments but should make “holistic” assessments that look beyond past-due information based on “other indicators” to determine if a treatment should be considered a distressed restructuring

Given that the measures in the Charter are available for to up-to-date customers without the use of affordability assessments, and the level of personalisation of the measures is low, they are a poor indicator of SICR or default on their own.

The PRA suggested that firms should develop frameworks for making “holistic assessments of loans subject to payment deferrals” for indicators of SICR or credit impairment and proposed four elements that should be considered: economic conditions, historical information that firms already hold about borrowers, information gathered from customers subject to payment deferrals, and application of expert judgement. The PRA give very helpful guidance for each of these categories in the letter that we will not reproduce here.

In line with the recommendations in the framework, we recall that when deciding which payment holidays should be put into Stage 2 lenders often used a combination of data such as payment history, other internal risk data (LTVs, LTIs, application risk segment), and a range of bureau indicators (absolute level or change in score / indebtedness / delinquency on other credit products) to identify cohorts of customers perceived to be higher risk. Taking a similar approach to the measures in the Charter seems reasonable.

Over time, as data emerges on the post-treatment behaviour of customers, lenders will be able to refine rules (or even build models) to more accurately identify which customers may face financial difficulty on exiting the treatment but, to start with, these choices will involve expert judgement.

In terms of default and Stage 3, which are aligned at most banks, the situation is similar to Stage 2 recognition: because of the absence of delinquency and an affordability test there is little objective evidence to place a borrower in default. In its Covid-era guidance the PRA had no expectations that lenders would investigate the “circumstances surrounding a borrower’s request for a payment deferral” but that “firms should utilise borrower information they have to assess borrowers for indicators of unlikeliness to pay”. Both these points seem relevant to the intention behind measures in the Charter. When considering indicators of unlikeliness to pay, two elements from the EBA guidelines on the application of the definition of default are relevant and may justify default status:

1. Distressed restructuring and diminished financial obligation: helpfully PS23/8 gives three examples of how the scheduling of payments could change under the measures in the Charter. Using these, the rescheduling of cash flows yields a nil impact on the net present value of the loan, as providing a firm continues to accrue interest on capital due, the increased future payments offset the reduced amounts received across the period the measure is extended. This does not therefore trigger default in accordance with paras 51, 52 EBA/GL/2016/07.
2. Other indicators of unlikeliness to pay: if there are justified concerns about a borrower’s ability to generate stable and sufficient cash flows to service the debt. For example, by the customer contacting their lender to discuss concerns over their affordability. [para 59(b) EBA/GL/2016/07]
   

As with the consideration of forbearance, we would expect “normal” processes to apply when the Charter treatments have ended.

While many lenders made adjustments to move some payment holidays to Stage 2 over Covid, they typically viewed Stage 3/default as being more objective and few firms pushed payment holidays into default without observed evidence of financial difficulty (e.g. missed payments). We would expect a similar approach for the measures in the Charter.

Repossession delay

The last element of the Charter is that lenders have committed not to repossess a home within the year following a customer’s first missed payment. Given that repossession proceedings typically take longer than this, it should have a negligible impact on ECL estimation.

What next?

To date there has been little change in flow to delinquency and mortgage forbearance balances have been falling. However, as customers continue to roll off fixed rate deals priced during a period of historic interest rate lows, the risk of payment shock and affordability challenges remain. Demand for the measures in the charter may well build over the autumn and it will be critical that lenders signed up to the charter are able to respond to any ramp up, as they did during the Covid pandemic. Lenders will also need to remain mindful of the need to monitor performance of those customers who make use of the measures, as only time will tell whether customers’ really do remain more resilient than expected, or if the measures simply have the effect of delaying defaults.