Reporting and Governance

The forbearance measures being rolled out by the government around debt repayments will mean that in the future banks will have to give greater thought to the treatment of data and its impact on their models. By Damien Burke, a partner at 4Most.

On March 11, 2020, the UK government and the Bank of England announced packages to directly support businesses and consumers and enable banks to continue to provide lending and forbearance to those affected by Covid–19. 

These initial announcements were a first step, and since then there have been numerous statements covering forbearance across mortgages, personal loans, current accounts, credit cards and car leases with regard to customer treatment from lenders and how the credit reference agencies (CRAs) should record it. These present a number of challenges in terms of identification, customer treatment, modelling and reporting.


In simple terms, there are three types of occasion when lenders need to identify their customer’s circumstances in relation to Covid-19: at the point they request a forbearance treatment, the point at which they are exiting the forbearance treatment, and when calculating capital and impairment. These assessments then needs to be consistent with reporting to the CRA.

Identifying customer circumstances at any stage has its own challenges. At the moment, the imperative is to try to support every customer through a challenging period. In this light, full customer interviews are impractical without limiting the number of customers you are able to support, hence some level of pragmatism or triage is required.


While identifying at entry may now be too late for some customers, given the uncertainty, and the possibility of further regulatory guidance to extend forbearance for a longer period, it is likely that further flexibility and opportunities for discussion will arise. At these times it will be important to ensure lenders are able to accurately identify customer circumstances and try to achieve the right customer outcome. For example, introducing a questionnaire to capture customer circumstances, such as whether the income shock they are experiencing is related to redundancy or reduction in overtime. In addition, capturing updated information that could indicate whether they may be more likely to be in financial difficulties in the future, for example, based on industry or job title. Analysis could be performed on  pre-delinquency strategies in the run-up to Covid-19 to establish accounts that were already marginal in terms of potential financial difficulties and stress performed on industry-level macro-economic variables to identify customers that might struggle in the future. This information could be used to indicate which forbearance measure is most appropriate. and inform the likely long-term impact. 

Given the focus on supporting as many customers as possible, it is likely that the Financial Conduct Authority (FCA) will apply a level of pragmatism in terms of identifying the presence of coronavirus-related difficulties, particularly if the lender errs on the side of caution and provides Covid-19 forbearance. As the customer approaches the end of the Covid-19 forbearance, however, decisions will need to be made as to the next course of action. It would be reasonable for the regulator to expect that there is more certainty from the lender as to the customer’s circumstances, given the time they will have been dealing with the customer, more clarity about what sectors of the economy are under long-term stress and internal developments using technology and data at the lender’s disposal. 


At the point at which the regulatory advice ends, lenders will have to provide next steps. These will largely depend on the customer’s new affordability, the products available and any new regulatory advice.

For example, for a standard residential mortgage the most likely solution is recapitalisation of the amount, either extending the term for a short period or slightly increasing the amount and remaining within the term. Either approach may increase the overall amount paid but given current low rates this is unlikely to be materially detrimental to the vast majority of consumers. Other products will have slightly more complex considerations: for example, the £500 interest free overdraft facility offered aligned to FCA guidance. If a customer has run up an overdraft balance whilst being furloughed on only 80% pay and then reverts to full pay after three months, it is unlikely they will be able to quickly pay down the balance. Will the lenders simply revert to an interest-bearing position immediately, take a staged approach to pay down with reduced interest, or create a separate interest free facility to be paid off over a longer period of time?

In terms of capital and impairment, there are shorter- and longer-term implications. While the guidance from the Bank of England seems to suggest that an increase to arrears and default is not necessarily expected at the moment given the forbearance and CRA reporting, it also follows that there should not be a decrease as a result of an event that clearly will have a detrimental impact on at least some parts of the economy. In the early stages at least it may be difficult to identify exactly which cases are affected, therefore some collective assessment will be required, likely linked to industry data and via a post-model adjustment. 

In the medium to long term however, serious thought will have to be given to data treatment and subsequent model impacts. For example, if a customer coming out of a three-month payment holiday misses all or a portion of their following payment, should they move immediately to default as they are 120 days past due? If they then pay back and resume paying in line with original contract terms (but without repaying the payment holiday amounts), should they then be held in default for 12 months to ‘cure’ as with other assets? These questions and many others will aggect the modelling approach, and how to include, adjust or exclude this data and what it represents for years to come, and may require the redevelopment of scorecards, as well as capital and impairment models.

This article is free to read, request a no obligation trial access to Global Risk Regulator.