The purpose of this publication is to highlight some of the key issues to be considered by lenders and banks in applying expected credit loss accounting in their IFRS financial statements. The financial reporting impacts other than expected credit loss accounting can be found in this Clearly IFRS.

Introduction

The coronavirus 2019 (COVID‑19) pandemic is affecting major economic and financial markets, and virtually all industries are facing challenges associated with the economic conditions resulting from efforts to address it. For example, many entities in the travel, hospitality, leisure, and retail industries have seen sharp declines in revenues due to regulatory and organisational mandates (e.g. “shelter in place” mandates, school closures) and voluntary changes in consumer behaviour (e.g. “social distancing”).

As the pandemic increases in both magnitude and duration, entities are experiencing conditions often associated with a general economic downturn, including, but not limited to, financial market volatility and erosion, deteriorating credit, liquidity concerns, further increases in government intervention, increasing unemployment, broad declines in consumer discretionary spending, increasing inventory levels, reductions in production because of decreased demand, layoffs and furloughs, and other restructuring activities. The continuation of these circumstances could result in an even broader economic downturn that could have a prolonged negative impact on an entity’s financial results.

This Clearly IFRS discusses certain key IFRS accounting considerations related to the accounting for expected credit losses (ECL) that may result from the COVID‑19 pandemic. The focus of this publication is for lenders and banks (collectively ‘banks’) though much of it will be applicable to measurement of ECL in industries other than financial services, such as in the accounting for ECL for contract assets arising under IFRS 15 Revenue from Contracts with Customers, trade and lease receivables as well as financial guarantee contracts provided to reimburse credit losses suffered by lenders to associates or joint ventures.

The accounting for ECL for banks is particularly challenging given ECL accounting is designed to incorporate estimations of credit events, and their consequential cash shortfalls, based on a probability‑weighted approach. In times of heightened uncertainty these estimations become more difficult. This publication highlights the major building blocks to ECL accounting, with particular focus on the measurement of ECL under stressed economic conditions and the impact that modifications and government assistance may have.

We also draw your attention to the document, Application of IFRS 9 in the light of the coronavirus uncertainty, published by the International Accounting Standards Board (IASB) on March 27, 2020.

Scope of IFRS 9’s impairment requirements

The scope requirements of ECL accounting are worth revisiting given much of the focus is on its application to loans, but the scope is far broader. The requirements also apply to debt securities (measured at amortised cost or fair value through other comprehensive income), trade receivables, contract assets under IFRS 15 Revenue from Contracts with Customers, lease receivables under IFRS 16 Leases, issued loan commitments and issued financial guarantee contracts which are not accounted under IFRS 4 Insurance Contracts and IFRS 17 Insurance Contracts.

Of particular note for banks is the accounting for issued loan commitments and issued financial guarantee contracts. In periods of economic stress, corporate borrowers are increasingly likely to draw down on borrowing facilities. The issuer of the loan commitment will need to reconsider the amount of ECL given future expected drawdowns can be accompanied with an increasing probability of default as corporate borrowers use facilities to meet their short‑term liquidity needs in response to declines in output and revenues. Similarly, financial guarantee contracts that reimburse the holder of the guarantee for failure of a borrower to pay under a specified debt instrument will exhibit greater probability of default when the issuer of the debt that is guaranteed is subject to financial difficultly. All other things being equal, an increase in probability of default will typically result in an increase in the ECL.

Application and timing of recognition

Under the general impairment model, an expected credit loss is a discounted probability‑weighted measurement of expected cash short falls either based on credit events arising in the 12 months from the reporting date (12m‑ECL) or based on credit events arising over the lifetime of the financial instrument (lifetime‑ECL)¹ . Where credit exposures have a term less than 12 months, 12m‑ECL equals lifetime‑ECL.

IFRS 9 is purposefully designed to be forward‑looking, reflecting expectations of future credit events (and resulting cash short falls) assessed at the reporting date. The determination of the ECL should reflect all reasonable and supportable information including that
which is forward‑looking (IFRS 9:5.5.4). Care is needed in determining what forward‑looking information shall be used as it must reflect the forward‑looking information that existed at the reporting date.

Forward‑looking information comes in a variety of forms, including, but not limited to:

• Macro‑economic forecasts, including GDP, industry‑sector growth rates, unemployment (national and regional), inflation, interest rates, property price indexation.
• Borrowers’ probability of non‑payment in response to macro‑economic factors that specifically relate to the borrower, noting that borrowers may prioritise payments of some debt obligations over others and so the credit risk of amounts owed by the same borrower can vary.
• Borrower behaviour in respect of timing of prepayment or extension options or use of undrawn facilities that impact the lender’s exposure.
• Valuation of collateral and timing of foreclosure.

This forwarding‑looking information influences three key inputs in ECL modelling:

• Probability of default (PD)
• Exposure at default (EAD)
• Loss given default (LGD)

Given ECL should reflect the expected cash shortfalls arising from non‑payment of the borrower, forward‑looking information should be designed to be as specific to the borrower’s economic circumstances as is practicable. The economic impacts of COVID‑19 are not evenly distributed. For example corporate borrowers in different industries will be impacted in different ways, as will retail borrowers depending on geography, employment status and other credit commitments.

Consideration should be given to the requirements in IAS 10 Events After the Reporting Period. IAS 10 distinguishes between adjusting and non‑adjusting events, with adjusting events being those that provide further evidence of conditions that existed at the end of the reporting period and therefore need to be reflected in the measurement of balances in the reporting period. Non‑adjusting events are those that are indicative of a condition that arose after the end of the reporting period (IAS 10:3).

At December 31, 2019, the forward‑looking information on the economic impact of COVID‑19 was significantly more limited than is evident at the date of this publication and in general a low probability was attributed to significant adverse scenarios which have since arisen. Issuers with a reporting period end date in 2020 will need to update their forward‑looking information to reflect expectations at the reporting date. Entities will need to distinguish between events that arose after the period end that reflect new events, as opposed to those that were reasonably expected at the reporting period end and so would have been reasonably assessed as being included in the forward‑looking assessment made at the reporting period end. This assessment might include, for example, assessing the status and extent of coronavirus infections in geographies relevant to the credit risk of the entity’s credit exposures at the reporting period end and considering the path and extent of the increase in infection rates in other areas that were affected earlier.

An entity may consider it reasonable at the reporting period end to forecast particular macro‑economic inputs used in ECL modelling. If those macro‑economic inputs end up not occurring or changing after the reporting date this should not be used as evidence to adjust the entity’s expectation at the period end. Doing so would represent inappropriate use of hindsight and would not reflect the conditions that existed at the reporting period end. Distinguishing between adjusting and non‑adjusting events requires significant judgement, particularly in the current economic environment where the economic severity of the COVID‑19 pandemic became apparent very shortly after the reporting period end.

The severity of the economic impact of COVID‑19 after the period end will require consideration even if those economic impacts are non‑adjusting events. In the case where non‑adjusting events after the reporting period are material an entity is required to disclose the nature of the event and an estimate of its financial effect, or a statement that such an estimate cannot be made (IAS 10:21). For banks with material exposures subject to ECL accounting, entities will need to consider the extent of disclosure and the practicality of determining the financial effect.

Where it is not practical for an entity to recalculate the ECL in order to determine the financial effect, the entity should consider disclosure of other qualitative and quantitative information, such as the identification of balances that are expected to be subject to the greatest degree of change in ECL because the impact of COVID‑19 on ECL might not be evenly distributed across the entity’s exposures.

Definitions, policy choices, and judgements made in applying accounting policies

In applying IFRS 9 entities will define a number of key terms and make a number of important application decisions when measuring ECL. One such item is the definition of default. This is critical in applying ECL accounting given that “default” is the “D” in PD, EAD and LGD. How default is defined will directly influence the staging of exposures into stage 1 (and therefore measured at 12m‑ECL) or stage 2 and 3 (measured at lifetime‑ECL) given staging is based on the probability of default happening (the PD). Further, moving exposures from stage 1 arises when there has been a significant increase in credit risk (SICR), this is assessed by comparing the lifetime PD of the exposure at initial recognition with the PD at the reporting date. IFRS 9:B5.5.37 requires that the definition is applied consistently to all financial instruments unless information becomes available that demonstrates that another default definition is more appropriate for a particular financial instrument.

Although IFRS 9 does not include a definition of default, banks will have already determined their definitions and applied them since they first applied IFRS 9. IFRS 9:B5.5.37 states that a definition of default should be consistent with the definition used for internal credit risk management purposes. Entities should apply those definitions consistently period to period, though it is acknowledged that definitions may need to be reviewed to ensure that they reflect current economic conditions of what a default is. For example, it is common for banks to consider unlikeliness to pay (UTP) criteria as indicators of default. The UTP criteria may need to be updated to reflect the current UTP indicators that are evident from borrowers’ non‑payment behaviours in the current economic environment. Similarly, definitions of default are often aligned, to the extent acceptable, to prudential regulatory definitions of default. Some prudential regulators have issued updated guidance on prudential default definitions in response to COVID‑19 and banks will need to consider whether these are appropriate for purposes of ECL accounting.

Banks will need to consider the requirements of IFRS 9:B5.5.37 where it states there is a rebuttable presumption that default does not occur later than when a financial asset is 90 days past due unless an entity has reasonable and supportable information to demonstrate that a more lagging default criterion is more appropriate. Care should be taken in rebutting this presumption, particularly in the current economic environment where an increasing amount of balances may be subject to longer ‘days past due’ which may be evidence of an increasing probability of default. Conversely, where payment holidays are introduced, during which borrowers are permitted to defer certain payments, such amounts may no longer be past due.

IFRS 9:5.5.11 also includes a rebuttable presumption that if a payment is more than 30 days past due then a SICR has occurred. IFRS 9:B5.5.19 is clear that 30 days is presumed to be the latest point at which a SICR has occurred and in practice entities may use 30 days past due as a back‑stop among other quantitative and qualitative indicators of a SICR.

In times of economic stress where there is an increasing number of late or deferred payments, banks will need to consider whether such payments will result in an expected cash shortfall. In some cases there will be no cash shortfalls, for example where the payment is deferred and the deferred payments accrue interest at the contractual rate where amounts are expected to be fully recovered. In other cases this will not hold true either because the bank is not permitted to charge the borrower interest on deferred payments, or is able to charge interest but at a lower amount than the original contracted interest rate.

IFRS 9 permits entities to conclude that financial instruments with low credit² risk at initial recognition and at the balance sheet date have not experienced a SICR and remain in stage 1. Given the speed at which the economic environment is changing, banks will need to assess whether it is reasonable to consider specific exposures as low credit risk. It is common for banks to use internal and external credit risk ratings for this purpose so banks should consider whether such ratings are up‑to‑date at the reporting date as rating classifications may lag a deterioration in credit risk and so may not reflect the credit conditions at the reporting date.

Model risk

The models used by many banks were not developed to accommodate the extreme economic conditions and the levels of government support measures being introduced and so entities may find that additional model development is required. The inclusion of more extreme economic inputs may also identify issues with the sensitivity and calibration of existing models (for example with respect to the transfer of items into stage 2).

Banks sometimes use post model adjustments (PMAs) to reflect risks and other uncertainties that are not included in the underlying ECL measurement models. These PMAs acknowledge that the model may have limitations that do not reflect the expected economic conditions at the reporting date. The speed of the economic impact of COVID‑19 may mean banks include PMAs to cater for the inadequacies of their ECL models given their models may not have been designed to cater for the economic circumstances that currently exist. Given PMAs overlay the output of models, such PMAs should be well controlled, authorised, documented and potentially disclosed.

Staging

An entity’s staging analysis (and subsequent measurement of ECL for loans in each stage) requires an assessment of forward‑looking information, principally in respect of lifetime PDs. Identifying forward‑looking information that appropriately reflects the potential impact of COVID‑19 will be one of the most significant challenges banks face in their estimation of ECLs in current conditions.

Due to uncertainty over the severity and duration of the impact some banks might assume that the adverse impact will be short‑lived as a result of government intervention and so will give greater weight to longer term forecasts based on past experience. Banks should consider official economic forecasts issued by central banks and other authorities as well as their internal economist units in assessing the severity and duration of macro‑economic deterioration.

Despite the challenges in identifying relevant forward‑looking information entities are still required to make estimates based on reasonable and supportable information that is available without undue cost or effort at the reporting date (IFRS 9:5.5.17). This may involve greater reliance than before on external sources of PD information (e.g. ratings agency credit ratings or CDS spreads) given historical information may lack relevance. However, over‑reliance on such sources can be problematic as market prices for debt and derivatives might reflect factors other than the borrower’s risk of default (such as market liquidity or the market’s assessment of the value of collateral), and credit ratings can be a lagging indicator of credit risk.

Where entities struggle to identify and estimate the effects of COVID‑19 at an individual instrument level, they may perform the staging analysis on a collective basis, particularly where credit exposures are concentrated in geographic locations or industry sectors that are severely adversely impacted by the pandemic (similar to the illustrative example in IFRS 9 where a collective assessment is undertaken for a portfolio of loans that is dependent upon the continued production of coal from a specified mine that employs a substantial portion of the borrowers in the portfolio). In some cases banks may even determine that exposures to whole industry sectors and specific geographies exhibit a SICR.

Modelling lifetime PDs is one important aspect of a staging analysis but identifying a SICR requires a holistic assessment of a number of quantitative and qualitative indicators (IFRS 9:B5.5.17). The impact of COVID‑19 may have a significant impact on the relevance of many of these factors. For example, days‑past‑due metrics (including the backstops referred to above) would reflect the impact of payment moratoria where borrowers take advantage of a payment holiday and so amounts may no longer be past due. Credit systems may have automatically considered that a payment holiday is evidence of SICR. Such systems may need to be amended to reflect that in the current environment not all payment holidays are necessarily indicative of a SICR.

Among customers that apply for a payment holiday, separating those that are in financial difficulty from those that are not, will be an operational challenge. Consideration will need to be given to the payment status and history of the borrower at the date the borrower applies for the payment holiday. In some jurisdictions, the payment holiday requires a borrower to demonstrate that without the payment holiday they could not be able to meet the contractual payments under their borrowing. An assessment will be needed whether the payment holiday is providing a short term liquidity benefit or addressing a deterioration in the borrower’s ability to meet its obligation when due which, if it is a significant increase in lifetime PD, is a SICR. At the reporting date an entity may not be able to identify which specific borrowers this applies to, so collective ssessment of the credit risk characteristics of those borrowers may be necessary.

Changes to credit risk management practices in response to increased default activity may mean that previous qualitative indicators of SICR need to be reconsidered.

An entity’s staging analysis will also be significantly affected by its expectations of the duration and severity of the downturn caused by COVID‑19 and the resulting recovery profile. The historical speed by which exposures migrate back to stage 1 from stage 2, often known as cure rates, may be less applicable if the economic downturn and recovery is “V shaped”, i.e. a rapid recovery to pre‑downturn economic growth levels, if historical cure rates were based on previous financial crises where the recovery was slower. Where a financial asset is renegotiated or otherwise modified in response to a credit deterioration, IFRS 9:B5.5.27 notes that typically borrowers would need to demonstrate a “consistently good payment behaviour over a period of time” following the modification before the credit risk is considered to have decreased.

Credit enhancements and collateral often influence the amount that is recovered in the case of default, i.e. the LGD. Their impact on the PD is generally far more limited and so has little to no impact on staging. Financial guarantees may allow the lender to recover amounts from the guarantor but do not impact the borrower’s PD. This compares to some governmental measures described below that directly benefit the borrower by improving their ability to meet their contractual obligations that fall due, impacting the borrower’s PD.

Measurement of ECL

Number of scenarios and their probability weighting
The pandemic has led to economic disruption that will need to be reflected in modelling economic scenarios. Due to the pervasive nature of the economic disruption, in addition to updating general economic inputs such as GDP and rates of unemployment, it will be necessary to consider the impact of COVID‑19 on specific industry and geographic sectors. Combining macroeconomic factors such as GDP, interest rates, government support measures and unemployment with sector‑specific factors such as the recent significant reduction in the oil price on the borrower’s ability to meet its financial obligations will be challenging. Historical data that does not reflect the current economic environment is unlikely to give reliable forecasts in such uncertain times.

Banks may consider a “V”‑shaped economic recovery scenario that reflects a rapid recovery of distressed borrowers. However, given the uncertainties of whether such a recovery will occur, the inclusion of other more distressed downside economic scenarios should be considered, such as a “U”‑shaped economic recovery where the pace of the recovery is slower.

Multiple economic scenarios should incorporate different speeds by which borrowers may resume payments, i.e. the cure rate. The inclusion of multiple economic scenarios is particularly important at times of heightened uncertainty given the non‑linearity of how cash shortfalls arise in an economic downturn. The probabilities assigned to these multiple economic scenarios will often be a significant judgement warranting disclosure.

LGD and EAD
As with PD models (noted in Staging section above), there may be a lack of reliable forward‑looking information that can be used in the estimation of LGD and EAD. These estimates are further complicated by uncertainty over the impact of COVID‑19 on factors such as property values (relevant in determining LGD); borrower repayment behaviour (relevant to many aspects of ECL estimation notably EAD); and other factors such as mortality rates which may affect the measurement of certain products including equity release mortgages.

Borrower behavioural factors will be particularly important in relation to products that combine drawn and undrawn amounts such as revolving credit facilities or credit card facilities where a deterioration of the macroeconomic environment is generally accompanied by an increase in the volumes and duration of drawdowns. Acute levels of drawdowns by those in financial difficulty may impede a bank’s ability to take credit risk management actions such as the reduction or removal of credit limits which will affect the exposure period used in measuring ECL for exposures that contain both a drawn and undrawn component (IFRS 9:B5.5.40).

Modifications, forbearance and credit enhancements

Given the potential impact of COVID‑19 on borrowers’ ability to service their debts, a number of government and central bank led initiatives have been, or are due to be, implemented to support borrowers and banks. Other support and relief has been or is expected to be provided by banks and related parties of borrowers. These measures vary across jurisdictions and include:

• Payment holidays, with or without interest charged for delayed payments
• Loan guarantees from related parties, governments and central banks
• Lending to corporate borrowers or injections of capital by governments or related parties
• Government payments to individuals to replace lost incomes
• Government reimbursements to banks for losses suffered on certain loans
• Reduced taxes on corporates or delayed tax due dates
• New bank lending facilities such as overdrafts, short term loans, mortgages
• New government lending facilities
• Fee waivers on overdrafts, late payments and credit card cash withdrawals
• Renegotiated loan terms with extended maturities and reduced short term payments

Some of these measures result in a change to the contractual terms of loan assets which need to be assessed to determine how they affect the subsequent accounting for the loan (see below). There may be cases where a lender has expressed an intention to modify the contractual terms of an instrument but has not yet entered into a contractually binding agreement to do so. Such an intention does not affect the accounting for the instrument because it is not contractually binding and therefore does not change the existing contractual terms. However, such intentions may affect the staging of exposure and the measurement of ECL given that ECL is based on expected cash flows (e.g. borrowers may delay their cash payments based on expected revised cash flows before they are contractually in place).

Measures that support the borrower to pay the amounts due under the original terms of their loan will impact the measurement of ECL but will not affect the gross carrying amount of the loan. Measures that provide a reimbursement, either part or in full, to the lender to cover losses suffered due to non‑payment by the borrower that are not integral to the terms of the original loan will not affect the accounting nor the ECL measurement for the loan and are accounted for separately from the loan as they are not deemed part of the loan contract.

Where the contractual terms of a loan asset that is measured at amortised cost or FVTOCI are modified, that modification needs to be assessed under IFRS 9 to determine whether it results in derecognition of the asset. For the avoidance of doubt, any change in expected cash flows that is contemplated under a loan asset’s existing contractual terms is not considered a modification. For example, if the loan includes an option to delay payment and that option is exercised, this is accounted for under the amortised cost measurement requirements and does not represent a modification of the contractual terms – it represents an application of the contractual terms.

Assessing modifications for derecognition
There is limited guidance in IFRS 9 with respect to determining whether a modification of a financial asset results in derecognition. Given the IFRS 9 derecognition criteria, a modification of a financial asset can only lead to derecognition when the contractual cash flows expire. Derecognition may apply to specifically identified cash flows or a fully proportionate (pro rata) share of cash flows or a fully proportionate (pro rata) share of specifically identified cash flows from a financial asset. If the contractual rights to such cash flows are forgiven and hence expire, other contractual terms may be unaltered, only this part of the underlying financial asset is derecognised.

In the absence of more specific guidance to determine when the cash flows of modified loan asset expire, an appropriate way to assess whether a modification or renegotiation of a financial asset gives rise to derecognition is to consider the guidance on substantial modifications of financial liabilities.

For modifications of financial liabilities IFRS 9 states that “a substantial modification of the terms of an existing financial liability or a part of it (whether or not attributable to the financial difficulty of the debtor) shall be accounted for as an extinguishment of the original financial liability and the recognition of a new financial liability”.

Consequently, a financial asset should be derecognised if a modification or renegotiation gives rise to substantially different terms. This assessment may consider both qualitative and quantitative factors, and in some cases qualitative factors only where qualitatively it is clear the terms are substantially different. The assessment of what are substantially different terms may be subject to judgement. Entities should provide disclosure about their accounting policy for determining whether a modification of a financial asset is substantial if relevant to an understanding of the financial statements.

If it is not already clear from a qualitative assessment that a modification has resulted in a substantial change in a financial asset, it is appropriate to apply a quantitative assessment based on the guidance for financial liabilities. When doing so, a modification is generally deemed to be substantial if the net present value of the cash flows under the modified terms, including any fees paid or received, is at least 10 per cent different from the net present value of the remaining cash flows of the financial asset prior to the modification, both discounted at the original effective interest rate of the financial asset prior to the modification.

In the case of a modification or renegotiation of a credit‑impaired financial asset or a purchased or originated credit‑impaired financial asset that was subject to a write‑off, the gross carrying amount no longer represents the full contractual cash flows because the entity has no reasonable expectation of recovering those cash flows and so IFRS 9 deems them as being derecognised. In such circumstances, it may be appropriate to consider the expected cash flows (i.e. the contractual cash flows less the written off cash flows rather than the full contractual cash flows) before modification or renegotiation and compare those with the contractual cash flows after modification or renegotiation, in particular, when the modification or renegotiation can be seen as a concession to the borrower that in substance modifies the contract to reflect those expected cash flows.

Modifications that result in derecognition
If an amortised cost financial asset is modified leading to derecognition then the new financial asset is initially recognised at fair value and a derecognition gain or loss is recognised in profit or loss being the difference between the amortised cost of the old instrument (including the updated ECL) at the date of derecognition and the fair value of the new instrument at that date. IAS 1:82(aa) requires separate presentation in profit or loss of that gain or loss.

When a credit‑impaired financial asset is modified and the modification results in derecognition of the original financial asset, the new financial asset recognised is not automatically deemed to be originated credit‑impaired simply because the asset was credit‑impaired before modification. Credit‑impaired financial assets may be restructured so as to provide the borrower with relief through partial forgiveness of amounts due and so reduce the borrower’s PD. However, the borrower’s financial difficulty may remain after restructuring resulting in a large ECL relative to the ECL when the financial asset was first originated prior to it being restructured.

The requirement in IFRS 9:5.5.3 to assess whether credit risk has increased significantly since initial recognition will be based on comparing the credit risk at the reporting date and at the date of initial recognition, which in this case is the date of modification because that is the date the financial asset was initially recognised.

However, in some unusual circumstances following a modification that results in derecognition of the original financial asset, there may be evidence that the modified financial asset is credit‑impaired at initial recognition, and thus, the financial asset should be recognised as an originated credit‑impaired financial asset.

Modifications that do not result in derecognition
If the modification does not result in the asset being derecognised, the terms of the modified asset are discounted by the original EIR to derive the new gross carrying amount and a modification gain or loss is recognised in profit or loss for the difference between the old gross carrying amount and the new gross carrying amount. For example, a bank may agree with a borrower to permit the borrower to defer payments over a specific period with those deferred payments not incurring additional interest to compensate the lender for the time value of money and credit risk. Where that payment holiday is regarded as a modification not leading to substantially different terms given the original cash flows on the loan in absolute amounts are unchanged, but the timing of them is deferred, the loss arising from the absence of compensation for the time value of money and credit risk will be recognised in profit or loss as a modification loss as the gross carrying value of the asset is reduced.

If the asset is not considered credit‑impaired it will be necessary to determine whether the asset is in stage 1 or 2 of the impairment model. The original credit risk of the loan before modification is used to assess whether there has been a SICR since initial recognition. At the reporting date the original credit risk is compared to the credit risk of the modified loan to determine whether there has been a SICR since initial recognition and which stage of the model the asset is classified in. The staging of the asset following modification could be different from the staging that would have applied absent the modification. For example, the modified cash flows may lead to a reduced PD compared to what would have been the case under the original terms; conversely, banks should consider what prompted the modification which may indicate a higher PD. Any difference between the opening ECL and closing ECL is recognised in profit or loss as an impairment gain/loss separately from any modification gain/loss.

Not all modifications that reduce the cash flows the lender is entitled to will necessarily result in a change in the staging for impairment purposes. A thorough understanding of the credit quality of the borrower both before and after the modification is needed. Some governments have proposed payment holidays for individual borrowers in response to the financial stress placed on households due to potential loss of income. Whether such payment holidays lead to a change in staging for ECL, for example, moving from stage 1 to stage 2 because that is evidence of a SICR, will depend on the specific circumstances of the borrowers. IFRS 9:5.5.4 permits a collective assessment for ECL measurement, including allocating exposures in stages, where the change in credit risk of borrowers cannot be individually identified. Further, banks may need to segregate existing portfolios to reflect that groups of borrowers within a portfolio no longer have the same credit risk.

Forbearance
When the terms of a loan are modified in response to the borrower’s financial difficulty this will often represent a concession or forbearance provided to the borrower which if not provided would have resulted in default and a credit shortfall for the lender. If this is the case, the asset meets the definition of credit impaired (i.e. it is in stage 3). However, not all loan modifications, such as payment holidays, will result in a SICR (movement to stage 2 or 3) or in the loan being credit‑impaired (stage 3). Where modifications are offered as a matter of policy to all borrowers in a particular country, region or affected industry, it will be necessary to assess the specific facts and circumstances of the borrower to determine whether this reflects an expected cash shortfall. In situations where relief through modified contractual terms is available to all borrowers, but must be applied subject to meeting certain credit quality deterioration criteria, there is a greater likelihood that use of the relief is evidence of a SICR or meeting the definition of credit‑impaired. In some cases a borrower may take advantage of a payment holiday that provides a liquidity benefit, but does not result in the lifetime PD of the borrower increasing significantly. Determining whether a change in the timing of contractual cash flows is a SICR or evidence of a credit‑impaired financial asset, requires careful consideration of the specific facts and circumstances.

Measures to support borrowers
Measures that support the borrower to pay the amounts due under the original terms of their loan (e.g. government provided income support) can have an effect on the staging of a loan asset because it will affect the borrower’s ability to repay its obligations. Such support can also affect the measurement of 12‑month or lifetime expected losses as the borrower’s capacity to repay is enhanced relative to the financial support not being available. It will be important for banks to assess carefully the support measures in place for borrowers and whether they will be available to its customers. Where this is not certain, an assessment of the likelihood of the availability of the support and quantum of such support will be required based on reasonable and supportable information available at the reporting date.

Measures that support borrowers may improve the PD of borrowers relative to the measure not being available. However, this does not necessarily mean that all measures improve the PD of borrowers overall given the PD used for staging reflects the overall lifetime PD reflecting the broader ability of the borrower to meet its obligations when they fall due. A borrower may experience an increase in PD arising from current economic conditions that is only partly offset by government measures that are available to the borrower.

Measures to support banks
Measures that support banks to cover any losses suffered due to non‑payment by the borrower will need to be assessed carefully to determine whether they are integral to the contractual terms of the loans to which they relate. Where such support is integral, the cash flows are included in the measurement of ECL. However, they are not included in the assessment of the probability of default, in the same way that other credit loss mitigants, such as collateral, which are considered as part of the loss given default but not the probability of default.

Where the credit enhancement is not integral, the cash flows expected to be received are not included in the probability of default assessment nor the measurement of ECL. Instead, a separate reimbursement asset is recognised to the extent the conditions in IAS 37 Provisions, Contingent Liabilities and Continent Assets are met.

Whether credit enhancements are integral to the contractual terms of financial instruments is not defined in IFRS 9, although the IASB’s Transition Resource Group for Impairment of Financial Instruments acknowledged that some guarantees that are not part of the contractual loan agreement may be deemed integral. Where recent measures have been introduced in response to the economic impact of COVID‑19 which were not contemplated when the loan was originated, such measures to support banks would generally not be deemed integral to the loan. Where banks originate new loans as part of a government assistance programme for borrowers affected by the economic impact of COVID‑19, paying the government for the guarantee to reimburse it in part or in full for credit losses it may suffer, then the guarantee may be integral. Determining what is integral may be a significant judgement that warrants disclosure.

Disclosure

IFRS 7 Financial Instruments: Disclosures requires extensive credit risk disclosures which are not reproduced in this publication. The credit risk disclosures required by IFRS 7 shall enable users of financial statements to understand the effect of credit risk on the amount, timing and uncertainty of future cash flows (IFRS 7:35B). IFRS 7 includes many mandatory quantitative disclosures, though the need to expand on qualitative disclosures will be critical in explaining how the current economic environment has impacted the amount, timing and uncertainty of future cash flows. In particular, disclosures should help users understand the movement of exposures through the three stages of the general ECL model and the resulting impact on recorded amounts. Given the areas discussed earlier in this publication particular consideration should be given to the disclosure requirements in IFRS 7 pertaining to how forward looking information is incorporated into the determination of ECL (IFRS 7:35G(b)), effect of modifications of contractual cash flows that have not resulted in derecognition (IFRS 7:35J) and the effect of collateral and other credit enhancements (IFRS 7:35L).

IFRS 7 applies in the annual financial statements, not in interim financial statements. However, for interim financial statements, IAS 34 Interim Financial Reporting requires in the interim financial report an explanation of events and transactions that are significant to an understanding of the changes in financial position and performance of the entity since the end of the last annual reporting period (IAS 34:15). For banks, the decline in economic activity and resulting credit deterioration since the last reporting date would be such an event that would warrant disclosure. Consideration should be given to whether the IFRS 7 disclosures required in the annual financial statements should be partly reproduced in the interim financial report in order to comply with the explanation needed by IAS 34.

Entities should reconsider the disclosure of judgements that management has made in the process of applying the entity’s accounting policies that have the most significant effect on the amounts recognised in the financial statements (IAS 1:122). Management may have formed new judgements in light of the current economic environment that warrant disclosure that in previous financial statements did not. Similarly, reconsideration of disclosure of sources of estimation uncertainty (IAS 1:125) will be needed given the uncertainties may have changed, or been exacerbated, since the last reporting period. ECL accounting often includes a significant degree of estimation uncertainty, but particularly so when the amount of uncertainty is higher given the various economic paths that could follow from the immediate outbreak of COVID‑19.

Disclosure of non‑adjusting post balance sheet events is referred in the section in Application and timing of recognition above.

Further information

For more information, please contact Kerry Danyluk, Diana De Acetis, Maryse Vendette, Joyce Lam, Alexia Donoghue or An Lam.

---------------------------------

¹ An exception applies for purchased or originated credit impaired assets (POCI) subject to IFRS 9:5.5.13‑14.

² IFRS 9:B5.5.22 explains that an asset displays low credit risk when it has a low risk of default; the borrower has a strong capacity to meet its
contractual cash flow obligations in the near term; and adverse changes in economic and business conditions in the longer term, may, but will
not necessarily, reduce the ability of the borrower to fulfil its contractual cash flow obligations.