In this issue of the Financial Reporting Hot Topics newsletter we look at the implications of the novel coronavirus (COVID-19) outbreak from a financial instruments accounting perspective. We also briefly highlight some other financial instruments accounting matters that are expected to be topical for entities with a 31 December 2019 financial year-end.

Below is a one-page summary for your consideration, followed by more detailed information covering the matters section by section. The insights are meant to help you tackle the issues.


The main question is whether the novel coronavirus outbreak affects the measurement of expected credit losses (ECLs) at the reporting date, or whether it is ‘non-adjusting’.

  • While the outbreak started in December 2019, detailed information about the virus and actions taken by the governments and private sector only emerged in January 2020. This means the outbreak would practically not have had an impact on ECL measurement.
  • If an obligor went bankrupt between the reporting date and the date the financial statements were authorised for issue, the question is whether such bankruptcy would confirm that the obligor was in troublesome conditions already at the reporting date. This depends on which factors eventually led to the bankruptcy, and given there were economic developments in 2019 (e.g. US-China trade tensions or social unrest in Hong Kong) in addition to the outbreak that affected entities from January 2020, this assessment can be judgemental.
  • Entities should disclose information that enables users of the financial statements to understand the impact of the outbreak on their financial position and performance up to when the statements are authorised for issue. There is not much detailed guidance on how to provide such disclosures. It is important they are clear, and entities need to be careful to avoid providing disclosures in ways that could be accidentally misleading.

Other topics that are addressed:

  • The effect of other detrimental developments for the economy (such as US-China trade tensions or social unrest in Hong Kong) on forward-looking adjustments for ECL measurement already in 2019.
  • Whether entities should update their risk disclosures in light of the developments in 2019 and after the reporting date.
  • How fair-value measurement for unlisted equities might be affected by these developments.
  • How financial guarantees affect measuring ECL given the related IFRIC discussion in 2019.
  • Considerations for whether to early adopt the IBOR reform-related changes for financial instruments standards.

Novel coronavirus outbreak: adjusting vs non-adjusting ‘event’?

The issue

Given the rapid development of the novel coronavirus outbreak in January 2020 and how it now affects the economy and the current economic outlook, the question for accountants is: how does this affect financial reporting for periods ended on 31 December 2019?

Accounting implications for expected credit losses (ECLs) in accordance with IFRS 9

The main question from the perspective of IFRS 9 Financial Instruments is how this development affects the accounting for ECLs. It is best to look at it in terms of the ‘stages’ that the ECL accounting model uses, i.e. a financial asset not yet having a significant increase in credit risk (‘Stage 1’), having experienced such an increase (‘Stage 2’), and ‘credit-impaired’ (‘Stage 3’).

Implications for ‘Stages 1 and 2’

ECL is a measurement of loss based on a probability-weighted amount in combination with forward-looking information. Hence it anticipates credit losses that are not linked to particular loss events that occurred up to the reporting date. For this part of the ECL exercise, the notion of ‘adjusting versus non-adjusting event’ does not have any relationship with the loss estimate. This is because adjusting versus non-adjusting event is a binary issue whereas a probability-weighted estimate includes different outcomes of a spectrum (instead of just one out of a ‘spectrum’ of two).

Consequently, for ‘Stages 1 and 2’ the equivalent of the adjusting versus non-adjusting ‘event’ debate is the question of what is:1

“reasonable and supportable information that is available without undue cost or effort at the reporting date about past events, current conditions and forecasts of future economic conditions” (emphasis added)

The salient aspects are that the information to be used must be:

a.  available at the reporting date, i.e. 31 December 2019
b. (also) about future economic conditions

This means that a forecast of future economic development has to be made – but without using hindsight. Therefore, only the information that was available up to 31 December 2019 can be considered in forecasting the economic conditions for 2020 (and beyond). That information then affects the economic scenarios used for the forecasting and the probability assigned to them.

The ITG2 discussed how to do forecasts of future economic conditions and observed (emphasis added):3

a. “reasonable and supportable new information that becomes available before the reporting date is required to be taken into consideration when applying the impairment requirements”
b. “expected credit losses are a probability-weighted estimate of credit losses at the reporting date. Accordingly, the determination of expected credit losses should take into consideration relevant possible future scenarios based on a range of expectations at the reporting date, using the information available at that date.”

Applying this guidance to the timeline of the outbreak, at the relevant cut-off date of 31 December 2019 for information that can be considered in the estimate as ‘known’ (i.e. the basis for projections) there was not yet knowledge about:

a. The kind of virus – the pneumonia was still of an ‘unknown’ cause.
b. That the infection involved human-to-human transmission.
c. That the infection could lead to death.
d. Quarantine measures that have not yet been taken or announced.
e. Later measures such as transport restrictions and extended holidays.

Any such aspects, therefore, could only represent estimates of possible scenarios, which are subject to probability weighting. The modelling of such scenarios and assigning probabilities must also not be coloured by the use of hindsight. The timeline above suggests that until the reporting date of 31 December 2019 very little was known about the virus, and the expectation of the impact of the virus infections would have been far less pessimistic than the picture that emerged during January 2020. So when avoiding undue hindsight, the outbreak would practically not have had an impact even for a probability-weighted estimate.

In summary, the distinction that IFRS 9 requires for the forward-looking information4 is about ‘adjusting versus non-adjusting’ information (which is subtly different from the distinction between adjusting and non-adjusting events in IAS 10 Events after the Reporting Period).

Implications for ‘Stage 3’

For ‘Stage 3’ there is also guidance in IAS 105, which includes the bankruptcy of a customer after the reporting date as an example of an adjusting event that “usually confirms” that the financial asset was credit-impaired at the reporting date. This example has caused some confusion as to how it interacts with ECL as a probability-weighted measurement. Without going into details on that, the important aspect to note is that IAS 10 says that such a bankruptcy usually confirms that the financial asset was already credit-impaired at the end of the reporting period. This qualification – in conjunction with the definition of “[e]vents after the reporting period”, which refers to events that provide information about conditions that existed at the end of the reporting period – is relevant in the context of the outbreak:

a. Even though the virus already existed at the reporting date, it had not changed economic conditions for companies yet. The business of companies was not yet affected as there were no restrictions on activities yet, nor even changes to consumer behaviour. But it is those economic conditions that affect ECLs, not the mere presence of a virus itself (of which there are plenty, like the seasonal flu).
b. In contrast, in a typical bankruptcy the event of ‘going bankrupt’ is the culmination of a severe deterioration of the entity’s economic substance that leads to insolvency. This is a development that typically takes at least months, if not longer, as the entity normally takes measures to attempt avoiding insolvency. This means that in contrast to the situation of the virus, the company in a typical bankruptcy situation would have been in a very poor economic situation already months earlier – at the reporting date.

On this basis one could argue that in the context of the outbreak even a bankruptcy of an obligor between the reporting date and the date the financial statements were authorised for issue would not be the usual case in which it confirms that the obligor was in troublesome conditions already at the reporting date, but instead a non-adjusting event. However, there were other detrimental developments for the economy already in 2019 (e.g. US-China trade tensions or social unrest in Hong Kong) so that the bankruptcy of a customer should be analysed taking those developments into consideration as well. In other words, the outbreak is not a ‘free pass’ to regard all bankruptcies after the 31 December 2019 year-end as ‘non-adjusting’ solely on the grounds that the outbreak had not yet had a significant economic impact by then.


Without going into detail as to whether the IFRS 9 requirements regarding the cut-off date to be used for forward-looking information represent a concept of adjusting versus non-adjusting events, given the general disclosure requirements in IAS 1 Presentation of Financial Statements6, one way or another disclosures must be provided that enable users of the financial statements to understand the impact of the outbreak on the entity’s financial position and financial performance. This leads to the need to provide disclosures required for non-adjusting events by IAS 10.21, namely:

a. the nature of the event – the novel coronavirus outbreak (and how it affects the reporting entity); and
b. an estimate of the financial effect of that event, or a statement that such an estimate cannot be made.

In the context of the accounting for ECLs, the disclosure of the estimated financial effect would be quantitative information about how the outbreak is estimated to affect the ECL allowance based on the information available after the reporting date but before the financial statements are authorised for issue. From a practical perspective, this would often mean running the ECL exercise again with, or adjusting it for, the latest available information on a date close to when the financial statements are authorised for issue.

Making such an estimate gives rise to additional questions:

a. Should the estimate be based on the financial assets recognised on the date that the estimate is based on or those recognised on 31 December 2019?

  1. This might not affect the ‘big picture’ anyway, depending on factors like how long after the reporting date the estimate is made, and how dynamic the entity’s financial asset portfolio is. But in case it does, IAS 10 does not delve into that level of detail for how to develop the information to be disclosed.
  2. On the one hand, IAS 10.22 includes situations that involve new transactions after the reporting date (i.e. the disclosable events are not limited to those in relation to transactions or items already existing at the reporting date7). This demonstrates that the disclosure does not use a ‘cut-off’ whereby only the effect in relation to items that already existed at the reporting date is considered. This could be taken to support a view that the estimate may appropriately be based on the financial assets recognised on the date of making the estimate. For example, that would allow an entity that does monthly ‘ECL runs’ for its internal reporting to use that information if the month-end is sufficiently close to when the financial statements are authorised for issue. This would also better serve the information needs of those users of the financial statements, who would want to know ‘how bad it is’ by the time the financial statements become available to them (‘it’ being the situation of the entity).
  3. On the other hand, if an entity wanted to give a narrower scope ‘update’ on how the ECL in relation to what was recognised at the reporting date has developed since, then the ‘broad’ reference in IAS 10.21 to ‘the financial effect of the event’ would not preclude that either.
  4. In any case, the basis on which the estimate is made should be disclosed if it has an effect that is large enough to influence how the user of that information would view it.

b. What should be disclosed in addition to such an ECL estimate?

Depending on the entity’s situation, accounting consequences that are not obvious from the disclosure of an updated ECL estimate might be relevant to the understanding by a user of the financial statements.  This could include write-offs, modifications of financial instruments (whether resulting in derecognition or not), transfers resulting in derecognition of financial instruments and similar events and transactions since the reporting date, if material. For example, a write-off after the reporting date would reduce the ECL allowance measured at a later date and, if the write-off was not apparent to a user, might result in drawing wrong conclusions (e.g. underestimating an increase in the ECL allowance). Appropriate disclosure about such events and transactions should be made so that such misinterpretations of the information given are avoided. This applies irrespective of the basis used for the disclosure. To illustrate: even if the narrower scope update only for items recognised on the reporting date is used, e.g. if after the reporting date that ECL amount would have been materially affected by sales or modification of receivables (realisation of credit losses), that could be important context for that updated estimate to avoid it being misleading.

Entities that contemplate making a statement that such an estimate of the impact on ECL ‘cannot be made’ should consider the impression that would leave on users of the financial statements, including securities and prudential regulators. Especially for financial institutions, that is an important consideration. That said, there is significant uncertainty over the financial effect from the outbreak, so instead of giving a point estimate it might be appropriate instead to disclose a range, e.g. based on different scenarios how long the outbreak and the related business interruption will last. Also, the sooner after the reporting date an entity authorises financial statements for issue, the more plausible it would be that such an estimate is difficult or not feasible and vice versa. 

Other aspects to consider

There was not only the novel coronavirus outbreak...

With the outbreak grabbing all the headlines and attention, it is easy to forget that there were other developments too. As mentioned, in 2019 there were challenges posed to the economy such as US-China trade tensions and social unrest in Hong Kong. This is why when entities consider their forward-looking adjustments for their ECL measurement exercise they should not simply assume it is ‘business as usual’. Even without taking into account the outbreak, the economic situation at the end of last year was generally not as stable as the situation at 31 December 2018 year-end.

Anything to update for risk disclosures (IFRS 7 Financial Instruments: Disclosures)?

IFRS 7 requires an entity to disclose information about the nature and extent of risks arising from financial instruments to which it is exposed at the reporting date (‘risk disclosures’).8

Entities should revisit whether their risk disclosures provide sufficient information that enables users of financial statements to evaluate the risks to which the entity was exposed at 31 December 2019. In light of the developments of the novel coronavirus outbreak, US-China trade tensions, and social unrest in Hong Kong, additional information about risk exposures may have to be given, or in a different form than before. For example, concentrations of risk9 could be very relevant as those developments affect different geographical areas, economic sectors or customer segments differently. Also, in light of those developments, the exposure to liquidity risk and how it is managed10 might be more important and relevant to users of the financial statements now than it was for past reporting periods.

In addition, if the entities have changed their objectives, policies and processes for managing risk, or the methods to measure it, that should be disclosed.11 Even if this occurred after the reporting date, that information might be relevant to users of the financial statements for the same reasons that underpin the disclosure regarding non-adjusting events in IAS 10.

Fair value measurement for unlisted equities

IFRS 9 requires all investments in equity instruments to be measured at fair value. If an entity used cost as an estimate of fair value for earlier reporting dates it should revisit whether this is still appropriate. IFRS 9 states that using cost in that way is only appropriate “in limited circumstances”12, and it provides indicators when cost might not be representative of fair value, including for example:13

“(a) a significant change in the performance of the investee compared with budgets, plans or milestones.
(c) a significant change in the market for the investee’s equity or its products or potential products.
(d) a significant change in the global economy or the economic environment in which the investee operates.
(e) a significant change in the performance of comparable entities, or in the valuations implied by the overall market.

The developments of US-China trade tensions and social unrest in Hong Kong would in many cases have affected the above indicators for fair values with a 31 December 2019 measurement date, in which case cost would not be an appropriate estimate of fair value. In contrast, the novel coronavirus outbreak was not affecting those indicators yet. But entities that need to measure fair value for investments in equity instruments should prepare for the situation that at the next measurement date for fair value the outbreak will have affected those indicators in many cases, and hence a valuation technique will have to be used.

Credit enhancements and their effect on measuring ECL

It is not uncommon for entities to purchase or originate a debt instrument whose terms include, or that is accompanied by, a guarantee contract or other credit enhancement. Examples include a guaranteed bond quoted in an active market, a trade receivable covered by credit insurance etc. The IFRS Interpretations Committee (the “IFRIC”) clarified that the cash flows expected from a credit enhancement are included in the measurement of ECLs if the credit enhancement is both:

a. part of the contractual terms; and
b. not recognised separately by the entity.14

Both these criteria have been subject to debate, and they each involve application challenges. For the first criterion this is in the form of difficult judgements because the reference to “part of the contractual terms” is not being applied strictly as worded. Instead, following some debate, this criterion has been applied in the notion of whether a guarantee is ‘integral’ to the debt instrument. This is a wider notion, but also a less clear one and therefore gives rise to difficult judgements.

For the second criterion, the difficulty arises from the issue being circular: if a guarantee is recognised separately it cannot be again taken into account in measuring the ECL of the guaranteed asset. It would create a double-counting issue because the same benefit of the credit protection would be included in the measurement of two items (an asset representing the guarantee and with the effect of also reducing ECL measurement). Conversely, if the guarantee is not recognised separately, it must be considered when measuring ECL as otherwise the benefit of the credit protection would not be accounted for at all. So the second ‘criterion’ is in effect a reminder to avoid double-counting or omitting accounting for the guarantee; but it does not help in deciding whether to take the benefit of the credit protection into account in measuring ECL or in accounting for a separate asset. To know the answer for whether to recognise the guarantee separately you already need to know the answer whether it is an integral part of the guaranteed asset – in short, a ‘chicken-and-egg’ situation.

Therefore it is not surprising that entities may have difficulties in determining whether the credit enhancement should be part of the ECL exercise for the debt instrument or accounted for as an item of its own.

Moreover, there is a lack of guidance in IFRSs for the accounting treatment of the credit enhancement if it is separately recognised (i.e. not part of the ECL exercise). In particular, when and how a reimbursement right representing the credit enhancement should be recognised and measured and its interaction with the accounting for the ECL on the debt instrument is unclear. This has resulted in different views evolving in practice.

For further KPMG insights on credit enhancements and their effect on measuring ECL, please visit our Global IFRS Institute’s webpage here.

Hedge accounting and IBOR reform

Again, with the novel coronavirus outbreak grabbing all the headlines and attention, it is easy to forget that there were developments in the reform of benchmark interest rates, and related IFRS standard setting. This affects entities that:

a. have derivatives on LIBOR or similar interest rate benchmarks (including cross-currency swaps with a LIBOR leg or similar); and
b. who apply hedge accounting for those in accordance with IFRSs (or HKFRSs, which have already been conformed to the changes made to IFRSs).

For such entities, recent standard setting by the IASB is relevant:

a. In September 2019, the IASB issued Interest Rate Benchmark Reform, which amends IFRS 9, IAS 39 and IFRS 7 (the “Amendments”). Equivalent amendments to HKFRS 9, HKAS 39 and HKIFRS 7 were issued by HKICPA in November 2019.
b. The Amendments provide far-reaching exceptions from normal hedge accounting requirements so that hedge accounting may continue despite the uncertainty resulting from regulatory-driven changes to many benchmark interest rates (mainly interbank offered rates, hence the label “IBOR”). Without the Amendments, IBOR reform creates at least a significant risk but, in most cases, even certainty regarding failing hedge accounting requirements and thus having to discontinue hedge accounting15.
c. The Amendments will be effective from annual periods beginning on or after 1 January 2020. Earlier application is permitted. This means companies can elect to apply the Amendments for their 2019 financial statements.

Early adoption of the Amendments gives companies the best chance to keep alive hedge accounting relationships directly affected by IBOR reform (e.g. a cash flow hedge of a LIBOR-based floating rate loan, a fair value hedge of the benchmark interest rate risk component of a fixed-rate bond etc). A summary of the Amendments can be found here.

An entity that chooses to early adopt the Amendments for the year ended 31 December 2019 must apply the Amendment retrospectively. However, for these particular changes the ‘retrospective’ application has no practical implication in that it would not change any numbers in an entity’s financial statements.

However, additional disclosures in IFRS 7.24H will need to be made, which include:

  • the significant interest rate benchmarks to which the entity’s hedging relationships are exposed;
  • the extent of the risk exposure that is affected by IBOR reform;
  • how the transition to an alternative interest rate benchmark is being managed;
  • a description of significant assumptions or judgements made in applying the Amendments; and
  • the nominal amount of the hedging instruments in those hedging relationships.

As a reminder, an entity will also need to provide the usual description of the change in accounting policies in accordance with IAS 8 Accounting Policies, Changes in Accounting Estimates and Errors. But given that the adoption of the Amendments will not change any numbers in an entity’s financial statements, the requirements to disclose quantitative information resulting from the retrospective application in IAS 8.28 will not be applicable.16

If you have any questions about the matters discussed in this publication, please feel free to contact the following partners or your KPMG contacts.

Paul Lau

Partner, Head of Professional Practice/Capital Markets
KPMG China
+852 2826 8010

Martin Friedhoff

Partner, Department of Professional Practice
KPMG China
+852 2913 2975

Tracy Chu

Senior Manager, Department of Professional Practice
KPMG China
+852 2685 7717

Connect with us

  1. FRS, which is about the measurement of ECL. The same type of information is also used to determine whether there has been a significant increase in credit risk, which drives the classification between Stages 1 and 2 (see IFRS While on the face of it this classification looks like a binary issue it is in actual effect not binary because of the requirement to use a so-called ‘top-down’ approach if using a ‘bottom-up’ approach (i.e. at the individual instrument level, including the use of groupings) a significant increase in credit risk cannot be assessed.
  2. Transition Resource Group for Impairment of Financial Instruments.
  3. See summary of the ITG meeting on 22 April 2015, paragraphs 11-18 (
  4. For ‘Stage 3’ there is still a notion of ‘adjusting events’; see section “Implications for ‘Stage 3’”.
  5. See IAS 10.9(b)(i).
  6. See IAS 1.17(c) and 31.
  7. For example, a business combination, an asset purchase, a fire destroying property, plant and equipment, foreign exchange rate movements.
  8. See IFRS 7.31-42.
  9. See IFRS 7.34(c).
  10. See IFRS 7.39 and the related application guidance (in particular IFRS 7.B11E-F).
  11. See IFRS 7.33.
  12. See IFRS 9.B5.2.3.
  13. See IFRS 9.B5.2.4.
  14. See IFRS 9.B5.5.55 and the IFRIC agenda decision in March 2019 (
  15. This is the result of the uncertain timing of the replacement for the existing benchmark interest rate, which then affects the qualifying criteria of hedging relationships. For example: interest rate benchmark-based cash flows being no longer highly probable, a non-contractually specified interest rate benchmark risk component being no longer separately identifiable, inability to demonstrate an economic relationship between the hedged item and hedging instrument or to conclude that the hedging relationship is expected to be highly effective, etc.
  16. The transition requirements provide an exemption to that effect too; see IFRS 7.44DF.