1.2.1 IFRIC 1

IFRIC 1 – Changes in Existing Decommissioning, Restoration and Similar Liabilities – provides guidance on how to account for the effect of changes in the measurement of existing provisions for obligations to dismantle, remove or restore items of property, plant and equipment. This is discussed at 6.3 below.

1.2.2 IFRIC 3

Another issue considered by the Interpretations Committee was how to account for a ‘cap and trade’ emission rights scheme. In December 2004, the Interpretations Committee issued IFRIC 3 – Emission Rights – but this was later withdrawn in June 2005. This interpretation, inter alia, required that as emissions are made, a liability was to be recognised for the obligation to deliver allowances equal to the emissions that had been made by the entity. Such a liability was a provision within the scope of IAS 37, and was to be measured at the present market value of the number of allowances required to cover emissions made up to the end of the reporting period. Accounting for liabilities associated with emissions trading schemes is discussed at 6.5 below.

1.2.3 IFRIC 5

IFRIC 5 – Rights to Interests arising from Decommissioning, Restoration and Environmental Rehabilitation Funds – deals with the accounting by an entity when it participates in a ‘decommissioning fund’, the purpose of which is to segregate assets to fund some or all of the costs of its decommissioning or environmental liabilities for which it has to make a provision under IAS 37. This is discussed at 6.3.3 below.

1.2.4 IFRIC 6

IFRIC 6 – Liabilities arising from Participating in a Specific Market – Waste Electrical and Electronic Equipment – provides guidance on the accounting for liabilities for waste management costs. This clarifies when certain producers of electrical goods will need to recognise a liability for the cost of waste management relating to the decommissioning of waste electrical and electronic equipment (historical waste) supplied to private households. This is discussed at 6.7 below.

1.2.5 IFRIC 21

IFRIC 21 addresses the recognition of a liability to pay a levy imposed by government if that liability is within the scope of IAS 37. It also addresses the accounting for a liability to pay a levy whose timing and amount is certain (see 6.8 below).

Although they are not interpretations of the standard as such, the IAS 37 guidance on non-financial liabilities is also referred to in IFRIC 12 – Service Concession Arrangements (see CHAPTER 25); and IFRIC 14 – IAS 19 – The Limit on a Defined Benefit Asset, Minimum Funding Requirements and their Interaction (see Chapter 35).

2.2.1 Items outside the scope of IAS 37

2.2.2 Provisions compared to other liabilities

IAS 37 states that the feature distinguishing provisions from other liabilities, such as trade payables and accruals, is the existence of ‘uncertainty about the timing or amount of the future expenditure required in settlement’. [IAS 37.11]. The standard compares provisions to:

1. trade payables – liabilities to pay for goods or services that have been received or supplied and have been invoiced or formally agreed with the supplier; and
2. accruals – liabilities to pay for goods or services that have been received or supplied but have not been paid, invoiced or formally agreed with the supplier, including amounts due to employees (for example, amounts relating to accrued vacation pay). Although it is sometimes necessary to estimate the amount or timing of accruals, the uncertainty is generally much less than for provisions.

IAS 37 also notes that accruals are often reported as part of trade and other payables whereas provisions are reported separately. [IAS 37.11].

For trade payables and their associated accruals, there is little uncertainty regarding either the amount of the obligation (which would be determined by the contracted price for the goods and services being provided) or of the timing of settlement (which would normally occur within an agreed period following transfer of the goods and services in question and the issue of an invoice). In practice, however, contracts can be more complex and give rise to a wide range of possible outcomes in terms of the amount or timing of payment. In these circumstances, the difference between provisions and other liabilities is less obvious and judgement may be required to determine where the requirement to make an estimate of an obligation indicates a level of uncertainty about timing or amount that is more indicative of a provision. Such judgements, if significant to the amounts recognised in the financial statements, would merit disclosure (see Chapter 3 at 5.1.1.B). [IAS 1.122].

One reason why this distinction matters is that provisions are subject to narrative disclosure requirements regarding the nature of the obligation and the uncertainties over timing and amount; and to quantitative disclosures of movements arising from their use, remeasurement or release that do not apply to other payables (see 7.1 below). In fact, although questions of recognition and measurement are important, transparency of disclosure is also a very significant matter in relation to accounting for provisions and ensuring that their effect is properly understood by users of the financial statements.

2.2.3 Distinction between provisions and contingent liabilities

There is an area of overlap between provisions and contingent liabilities. Although contingent liabilities are clearly not as likely to give rise to outflows, similar judgements are made in assessing the nature of the uncertainties, the need for disclosures and ultimately the recognition of a liability in the financial statements. The standard notes that in a general sense, all provisions are contingent because they are uncertain in timing or amount. However, in IAS 37 the term ‘contingent’ is used for liabilities and assets that are not recognised because their existence will be confirmed only by the occurrence of one or more uncertain future events not wholly within the entity's control. In addition, the term ‘contingent liability’ is used for liabilities that do not meet the recognition criteria for provisions. [IAS 37.12].

Accordingly, the standard distinguishes between:

1. provisions – which are recognised as liabilities (assuming that a reliable estimate can be made) because they are present obligations and it is probable that an outflow of resources embodying economic benefits will be required to settle the obligations; and
2. contingent liabilities – which are not recognised as liabilities because they are either:
   1. possible obligations, as it has yet to be confirmed whether the entity has a present obligation that could lead to an outflow of resources embodying economic benefits; or
   2. present obligations that do not meet the recognition criteria in the standard because either it is not probable that an outflow of resources embodying economic benefits will be required to settle the obligation, or a sufficiently reliable estimate of the amount of the obligation cannot be made. [IAS 37.13].

3.1.1 ‘An entity has a present obligation (legal or constructive) as a result of a past event’

The standard defines both legal and constructive obligations. The definition of a legal obligation is fairly straightforward and uncontroversial; it refers to an obligation that derives from a contract (through its explicit or implicit terms), legislation or other operation of law. [IAS 37.10].

The definition of a constructive obligation, on the other hand, may give rise to more problems of interpretation. A constructive obligation is defined as an obligation that derives from an entity's actions where:

1. by an established pattern of past practice, published policies or a sufficiently specific current statement, the entity has indicated to other parties that it will accept certain responsibilities; and
2. as a result, the entity has created a valid expectation on the part of those other parties that it will discharge those responsibilities. [IAS 37.10].

The following examples in IAS 37 illustrate how a constructive obligation is created.

These examples demonstrate that the essence of a constructive obligation is the creation of a valid expectation that the entity is irrevocably committed to accepting and discharging its responsibilities.

The standard states that in almost all cases it will be clear whether a past event has given rise to a present obligation. However, it acknowledges that there will be some rare cases, such as a lawsuit against an entity, where this will not be so because the occurrence of certain events or the consequences of those events are disputed. [IAS 37.16]. When it is not clear whether there is a present obligation, a ‘more likely than not’ evaluation (taking into account all available evidence) is deemed to be sufficient to require recognition of a provision at the end of the reporting period. [IAS 37.15].

The evidence to be considered includes, for example, the opinion of experts together with any additional evidence provided by events after the reporting period. If on the basis of such evidence it is concluded that a present obligation is more likely than not to exist at the end of the reporting period, a provision will be required (assuming that the other recognition criteria are met). [IAS 37.16]. This is an apparent relaxation of the standard's first criterion for the recognition of a provision as set out at 3.1 above, which requires there to be a definite obligation, not just a probable one. It also confuses slightly the question of the existence of an obligation with the probability criterion, which strictly speaking relates to whether it is more likely than not that there will be an outflow of resources (see 3.1.2 below). However, this interpretation is confirmed in IAS 10 – Events after the Reporting Period, which includes ‘the settlement after the reporting period of a court case that confirms that the entity had a present obligation at the end of the reporting period’ as an example of an adjusting event. [IAS 10.9].

The second half of this condition uses the phrase ‘as a result of a past event’. This is based on the concept of an obligating event, which the standard defines as ‘an event that creates a legal or constructive obligation and that results in an entity having no realistic alternative to settling that obligation’. [IAS 37.10]. The standard says that this will be the case only:

1. where the settlement of the obligation can be enforced by law; or
2. in the case of a constructive obligation, where the event (which may be an action of the entity) creates valid expectations in other parties that the entity will discharge the obligation. [IAS 37.17].

This concept of obligating event is used in the standard when discussing specific examples of recognition, which we consider further at 6 below. However, it is worth mentioning here that this concept, like that of a constructive obligation, is open to interpretation and requires the exercise of judgement, as the obligating event is not always easy to identify.

The standard emphasises that the financial statements deal with the financial position of an entity at the end of its reporting period, not its possible position in the future. Accordingly, no provision should be recognised for costs that need to be incurred to operate in the future. The only liabilities to be recognised are those that exist at the end of the reporting period. [IAS 37.18]. It is not always easy to distinguish between the current state at the reporting date and the entity's future possible position, especially where IAS 37 requires an assessment to be made based on the probability of obligations and expectations as to their outcome. However, when considering these questions it is important to ensure that provisions are not recognised for liabilities that arise from events after the reporting period (see Chapter 38 at 2).

IAS 37 prohibits certain provisions that might otherwise qualify to be recognised by stating that it ‘is only those obligations arising from past events existing independently of an entity's future actions (i.e. the future conduct of its business) that are recognised as provisions'. In contrast to situations where the entity's past conduct has created an obligation to incur expenditure (such as to rectify environmental damage already caused), a commercial or legal requirement to incur expenditure in order to operate in a particular way in the future, will not of itself justify the recognition of a provision. It argues that because the entity can avoid the expenditure by its future actions, for example by changing its method of operation, there is no present obligation for the future expenditure. [IAS 37.19].

The standard expects strict application of the requirement that, to qualify for recognition, an obligation must exist independently of an entity's future actions. Even if a failure to incur certain costs would result in a legal requirement to discontinue an entity's operations, no provision can be recognised. As discussed at 5.2 below, IAS 37 considers the example of an airline required by law to overhaul its aircraft once every three years. It concludes that no provision is recognised because the entity can avoid the requirement to perform the overhaul, for example by replacing the aircraft before the three year period has expired. [IAS 37 IE Example 11B].

Centrica plc describes its interpretation of the group's obligations under UK legislation to install energy efficiency improvement measures in domestic households in a manner consistent with Example 26.3 above.

Significantly, however, such considerations do not apply in the case of obligations to dismantle or remove an asset at the end of its useful life, where an obligation is recognised despite the entity's ability to dispose of the asset before its useful life has expired. Such costs are required to be included by IAS 16 – Property, Plant and Equipment – as part of the measure of an asset's initial cost. [IAS 16.16]. Decommissioning provisions are discussed at 6.3 below. Accordingly, the determination of whether an obligation exists independently of an entity's future actions can be a matter of judgement that depends on the particular facts and circumstances of the case.

There is no requirement for an entity to know to whom an obligation is owed. The obligation may be to the public at large. It follows that the obligation could be to one party, but the amount ultimately payable will be to another party. For example, in the case of a constructive obligation for an environmental clean-up, the obligation is to the public, but the liability will be settled by making payment to the contractors engaged to carry out the clean-up. However, the principle is that there must be another party for the obligation to exist. It follows from this that a management or board decision will not give rise to a constructive obligation unless it is communicated in sufficient detail to those affected by it before the end of the reporting period. [IAS 37.20]. The most significant application of this requirement relates to restructuring provisions, which is discussed further at 6.1 below.

The standard discusses the possibility that an event that does not give rise to an obligation immediately may do so at a later date, because of changes in the law or an act by the entity (such as a sufficiently specific public statement) which gives rise to a constructive obligation. [IAS 37.21]. Changes in the law will be relatively straightforward to identify. The only issue that arises will be to determine exactly when that change in the law should be recognised. IAS 37 states that an obligation arises only when the legislation is virtually certain to be enacted as drafted and suggests that in many cases, this will not be until it is enacted. [IAS 37.22].

The more subjective area is the possibility that an act by the entity will give rise to a constructive obligation. The example given is of an entity publicly accepting responsibility for rectification of previous environmental damage in a way that creates a constructive obligation. [IAS 37.21]. This seems to introduce a certain amount of flexibility to management when reporting results. By bringing forward or delaying a public announcement of a commitment that management had always intended to honour, it can affect the reporting period in which a provision is recognised. Nevertheless, the existence of a public announcement provides a more transparent basis for recognising a provision than, for example, a decision made behind the closed doors of a boardroom.

3.1.2 ‘It is probable that an outflow of resources embodying economic benefits will be required to settle the obligation’

This requirement has been included as a result of the standard's attempt to incorporate contingent liabilities within the definition of provisions. This is discussed at 3.2 below.

The meaning of probable in these circumstances is that the outflow of resources is more likely than not to occur; that is, it has a probability of occurring that is greater than 50%. [IAS 37.23]. The standard also makes it clear that where there are a number of similar obligations, the probability that an outflow will occur is based on the class of obligations as a whole. This is because in the case of certain obligations such as warranties, the possibility of an outflow for an individual item may be small (likely to be much less than 50%) whereas the possibility of at least some outflow of resources for the population as a whole will be much greater (almost certainly greater than 50%). [IAS 37.24]. With regard to the measurement of a provision arising from a number of similar obligations, the standard refers to the calculation of an ‘expected value’, whereby the obligation is estimated by weighting all the possible outcomes by their associated probabilities. [IAS 37.39]. Where the obligation being measured relates to a single item, the standard suggests that the best estimate of the liability may be the individual most likely outcome. [IAS 37.40]. For the purposes of recognition, a determination that it is more likely than not that any outflow of resources will be required is sufficient. The measurement of provisions is discussed at 4 below.

3.1.3 ‘A reliable estimate can be made of the amount of the obligation’

The standard takes the view that a sufficiently reliable estimate can almost always be made for a provision where an entity can determine a range of possible outcomes. Hence, the standard contends that it will only be in extremely rare cases that a range of possible outcomes cannot be determined and therefore no sufficiently reliable estimate of the obligation can be made. [IAS 37.25]. In these extremely rare circumstances, no liability is recognised. Instead, the liability should be disclosed as a contingent liability (see disclosure requirements at 7.2 below). [IAS 37.26]. Whether such a situation is as rare as the standard asserts is open to question, especially for entities trying to determine estimates relating to potential obligations that arise from litigation and other legal claims (see 6.11 below).

3.2.1 Contingent liabilities

A contingent liability is defined in the standard as:

1. a possible obligation that arises from past events and whose existence will be confirmed only by the occurrence or non-occurrence of one or more uncertain future events not wholly within the control of the entity; or
2. a present obligation that arises from past events but is not recognised because:
   1. it is not probable that an outflow of resources embodying economic benefits will be required to settle the obligation; or
   2. the amount of the obligation cannot be measured with sufficient reliability. [IAS 37.10].

At first glance, this definition is not easy to understand because a natural meaning of ‘contingent’ would include any event whose outcome depends on future circumstances. The meaning is perhaps clearer when considering the definition of a liability and the criteria for recognising a provision in the standard. A possible obligation whose existence is yet to be confirmed does not meet the definition of a liability; and a present obligation in respect of which an outflow of resources is not probable, or which cannot be measured reliably does not qualify for recognition. [IAS 37.14]. On that basis a contingent liability under IAS 37 means one of the following:

1. an obligation that is estimated to be less than 50+% likely to exist (i.e. it does not meet the definition of a liability). Where it is more likely than not that a present obligation exists at the end of the reporting period, a provision is recognised. [IAS 37.16(a)]. Where it is more likely than not that no present obligation exists, a contingent liability is disclosed (unless the possibility is remote); [IAS 37.16(b)]
2. a present obligation that has a less than 50+% chance of requiring an outflow of economic benefits (i.e. it meets the definition of a liability but does not meet the recognition criteria). Where it is not probable that there will be an outflow of resources, an entity discloses a contingent liability (unless the possibility is remote); [IAS 37.23] or
3. a present obligation for which a sufficiently reliable estimate cannot be made (i.e. it meets the definition of a liability but does not meet the recognition criteria). In these rare circumstances, a liability cannot be recognised and it is disclosed as a contingent liability. [IAS 37.26].

The term ‘possible’ is not defined, but literally it could mean any probability greater than 0% and less than 100%. However, the standard effectively divides this range into four components, namely ‘remote’, ‘possible but not probable’, ‘probable’ and ‘virtually certain’. The standard requires a provision to be recognised if ‘it is more likely than not that a present obligation exists at the end of the reporting period’. [IAS 37.16(a)]. Therefore, IAS 37 distinguishes between a ‘probable’ obligation (which is more likely than not to exist and, therefore requires recognition as a provision) and a ‘possible’ obligation (for which either the existence of a present obligation is yet to be confirmed or where the probability of an outflow of resources is 50% or less). [IAS 37.13]. Appendix A to IAS 37, in summarising the main requirements of the standard, uses the phrase ‘a possible obligation … that may, but probably will not, require an outflow of resources’. Accordingly, the definition restricts contingent liabilities to those where either the existence of the liability or the transfer of economic benefits arising is less than 50+% probable (or where the obligation cannot be measured at all, but as noted at 3.1.3 above, this would be relatively rare).

The standard requires that contingent liabilities are assessed continually to determine whether circumstances have changed, in particular whether an outflow of resources embodying economic benefits has become probable. Where this becomes the case, then provision should be made in the period in which the change in probability occurs (except in the rare circumstances where no reliable estimate can be made). [IAS 37.30]. Other changes in circumstances might require disclosure of a previously remote obligation on the grounds that an outflow of resources has become possible (but not probable).

3.2.2 Contingent assets

A contingent asset is defined in a more intuitive way. It is ‘a possible asset that arises from past events and whose existence will be confirmed only by the occurrence or non-occurrence of one or more uncertain future events not wholly within the control of the entity’. [IAS 37.10]. In this case, the word ‘possible’ is not confined to a level of probability of 50% or less, which may further increase the confusion over the different meaning of the term in the definition of contingent liabilities.

Contingent assets usually arise from unplanned or other unexpected events that give rise to the possibility of an inflow of economic benefits to the entity. An example is a claim that an entity is pursuing through legal process, where the outcome is uncertain. [IAS 37.32].

The standard states that a contingent asset should not be recognised, as this could give rise to recognition of income that may never be realised. However, when the realisation of income is virtually certain, then the related asset is no longer regarded as contingent and recognition is appropriate. [IAS 37.33].

Virtual certainty is not defined in the standard, but it is certainly a much higher hurdle than ‘probable’ and indeed more challenging than the term ‘highly probable’, defined in IFRS 5 as ‘significantly more likely than probable’. [IFRS 5 Appendix A]. We think it reasonable that virtual certainty is interpreted as being as close to 100% as to make any remaining uncertainty insignificant. What this means in practice requires each case to be decided on its merits and any judgement should be made in the knowledge that, in any event, it is rarely possible to accurately assess the probability of the outcome of a particular event.

The standard requires disclosure of the contingent asset when the inflow of economic benefits is probable. [IAS 37.34]. For the purposes of the standard ‘probable’ means that the event is more likely than not to occur; that is, it has a probability greater than 50%. [IAS 37.23]. The disclosure requirements are detailed at 7.3 below.

As with contingent liabilities, any contingent assets should be assessed continually to ensure that developments are appropriately reflected in the financial statements. If it has become virtually certain that an inflow of economic benefits will arise, the asset and the related income should be recognised in the period in which the change occurs. If a previously unlikely inflow becomes probable, then the contingent asset should be disclosed. [IAS 37.35].

The requirement to recognise the effect of changing circumstances in the period in which the change occurs extends to the analysis of information available after the end of the reporting period and before the date of approval of the financial statements. In our view, such information would not give rise to an adjusting event after the reporting period. In contrast to contingent liabilities (in respect of which IAS 10 includes as a specific example of an adjusting event ‘the settlement after the reporting period of a court case that confirms that the entity had a present obligation at the end of the reporting period’ [IAS 10.9]), no adjustment should be made to reflect the subsequent settlement of a legal claim in favour of the entity. In this instance, the period in which the change occurs is subsequent to the reporting period. There is also no suggestion that the example in IAS 10 is referring to anything but liabilities. An asset could only be recognised if, at the end of the reporting period, the entity could show that it was virtually certain that its claim would succeed.

3.2.3 How probability determines whether to recognise or disclose

The following matrix summarises the treatment of contingencies under IAS 37:

The standard does not put a numerical measure of probability on either ‘virtually certain’ or ‘remote’. In our view, the use of such measures would downgrade a process requiring the exercise of judgement into a mechanical exercise. It is difficult to imagine circumstances when an entity could reliably determine an obligation to be, for example, 92%, 95% or 99% likely, let alone be able to compare those probabilities objectively. Accordingly, we think it reasonable to regard ‘virtually certain’ as describing a likelihood that is as close to 100% as to make any remaining uncertainty insignificant; to see ‘remote’ as meaning a likelihood of an outflow of resources that is not significant; and for significance to be a matter for judgement and determined according to the merits of each case.

4.3.1 Real versus nominal rate

IAS 37 does not indicate whether the discount rate should be a real discount rate or a nominal discount rate (although a real discount rate is referred to in Example 2 in Appendix D which illustrates the narrative disclosure for decommissioning costs). The discount rate to be used depends on whether:

1. the future cash flows are expressed in current prices, in which case a real discount rate (which excludes the effects of general inflation) should be used; or
2. the future cash flows are expressed in expected future prices, in which case a nominal discount rate (which includes a return to cover expected inflation) should be used.

Either alternative is acceptable, and these methods may produce the same figure for the initial present value of the provision. However, the effect of the unwinding of the discount may be different in each case (see 4.3.5 below).

4.3.2 Adjusting for risk and using a government bond rate

IAS 37 also requires that risk is taken into account in the calculation of a provision, but gives little guidance as to how this should be done. Where discounting is concerned, it merely says that the discount rate should not reflect risks for which the future cash flow estimates have been adjusted. [IAS 37.47]. One may use a discount rate that reflects the risk associated with the liability (a risk-adjusted rate). The following example, taken from the UK Accounting Standards Board's (ASB) Working Paper – Discounting in Financial Reporting,⁷ shows how an entity might calculate such a risk adjusted rate.

As can be seen from this example, the risk-adjusted discount rate is a lower rate than the unadjusted (risk-free) discount rate. This may seem counter-intuitive initially, because the experience of most borrowers is that banks and other lenders will charge a higher rate of interest on loans that are assessed to be higher risk to the lender. However, in the case of a provision a risk premium is being suffered to eliminate the possibility of the actual cost being higher (thereby capping a liability), whereas in the case of a loan receivable a premium is required to compensate the lender for taking on the risk of not recovering its full value (setting a floor for the value of the lender's financial asset). In both cases the actual cash flows incurred by the paying entity are higher to reflect a premium for risk. In other words, the discount rate for an asset is increased to reflect the risk of recovering less and the discount rate for a liability is reduced to reflect the risk of paying more.

A problem with changing the discount rate to account for risk is that this adjusted rate is a theoretical rate, as it is unlikely that there would be a market assessment of the risks specific to the liability alone. [IAS 37.47]. However the lower discount rate in the above example is consistent with the premise that a risk-adjusted liability should be higher than a liability without accounting for the risk that the actual settlement amount is different to the estimate. [IAS 37.43]. It is also difficult to see how a risk-adjusted rate could be obtained in practice. In the above example, it was obtained only by reverse-engineering; it was already known that the net present value of a risk-adjusted liability was £138, so the risk-adjusted rate was just the discount rate applied to unadjusted cash flow of £150 to give that result.

IAS 37 offers an alternative approach – instead of using a risk-adjusted discount rate, the estimated future cash flows themselves can be adjusted for risk. [IAS 37.47]. This does of course present the problem of how to adjust the cash flows for risk (see 4.2 above). However, this may be easier than attempting to risk-adjust the discount rate.

For the purposes of discounting post-employment benefit obligations, IAS 19 requires the discount rate to be determined by reference to market yields at the end of the reporting period on high quality corporate bonds (although in countries where there is no deep market in such bonds, the market yields on government bonds should be used). [IAS 19.83]. Although IAS 19 indicates that this discount rate reflects the time value of money (but not the actuarial or investment risk), [IAS 19.84], we do not believe it is appropriate to use the yield on a high quality corporate bond for determining a risk-free rate to be used in discounting provisions under IAS 37. Accordingly, in our view, where an entity is using a risk-free discount rate for the purposes of calculating a provision under IAS 37, that rate should be based on a government bond rate with a similar currency and remaining term as the provision. It follows that because a risk-adjusted rate is always lower than the risk-free rate, an entity cannot justify the discounting of a provision at a rate that is higher than a government bond rate with a similar currency and term to the provision.

Whichever method of reflecting risk is adopted, IAS 37 emphasises that care must be taken that the effect of risk is not double-counted by inclusion in both the cash flows and the discount rate. [IAS 37.47].

In recent years, government bond rates have been more volatile as markets have changed rates to reflect (among other factors) heightened perceptions of sovereign debt risk. In some cases, government bond yields may be negative. The question has therefore arisen whether government bond rates, at least in certain jurisdictions, should continue to be regarded as the default measure of a risk-free discount rate. Whilst the current volatility in rates has highlighted the fact that no debt (even government debt) is totally risk free, the challenge is to find a more reliable measure as an alternative. Any adjustment to the government bond rate to ‘remove’ the estimate of sovereign debt risk is conceptually flawed, as it not possible to isolate one component of risk from all the other variables that influence the setting of an interest rate. Another approach might be to apply some form of average bond rate over a period of 3, 6 or 12 months to mitigate the volatility inherent in applying the spot rate at the period end. However, this is clearly inappropriate given the requirements in IAS 37 to determine the best estimate of an obligation by reference to the expenditure required to settle it ‘at the end of the reporting period’, [IAS 37.36], and to determine the discount rate on the basis of ‘current market assessments’ of the time value of money. [IAS 37.47].

With ‘risk’ being a measure of potential variability in returns, it remains the case that in most countries a government bond will be subject to the lowest level of variability in that jurisdiction. As such, in most countries it remains the most suitable of all the observable measures of the time value of money in a particular country. Where government bond rates are negative or, more likely, result in a negative discount rate once adjusted for risk, we believe that it is not appropriate to apply a floor of zero to the discount rate as this would result in an understatement of the liability. As discussed above, and at 4.3.1 above, IAS 37 offers various approaches to determining an appropriate discount rate. It may sometimes be the case that one or more of the allowed approaches result in a negative discount rate whereas the application of an alternative permitted approach would not. In order to avoid some of the presentational difficulties associated with a negative discount rate, entities faced with a negative real discount rate before risk adjustment may wish to consider the alternative approach of discounting expected future cash flows expressed in future prices, at a nominal discount rate (see 4.3.1 above), if the nominal rate is not negative. Similarly, entities that are faced with a negative risk-adjusted discount rate only because of risk adjustment (i.e. where risk free rates themselves are not negative) may wish to adopt the alternative approach of adjusting the estimated future cash flows to reflect the risks associated with the liability, rather than risk-adjusting the discount rate.

A difficulty that can arise in certain countries is finding a government bond with a similar term to the provision, for example when measuring a decommissioning provision expected to be settled in 30 years in a country where there are no government bonds with a term exceeding 10 years. In such cases, the government bond rate might be adjusted and the techniques adopted by actuaries for measuring retirement obligations with long maturities, that involve extrapolating current market rates along a yield curve, [IAS 19.86], might be considered. The difficulties of finding an appropriate discount rate in the context of retirement benefit obligations are discussed in Chapter 35 at 7.6.

4.3.3 Own credit risk

IAS 37 does not address whether an entity's own credit risk should be considered a risk specific to a liability when determining a risk-adjusted discount rate. In March 2011, the Interpretations Committee decided not to take to its agenda a request for interpretation of the phrase ‘the risks specific to the liability’ and whether this means that an entity's own credit risk should be excluded from any adjustments made to the discount rate used to measure liabilities. In doing so, the Interpretations Committee acknowledged that IAS 37 is not explicit on the question of own credit risk; but understood that the predominant practice was to exclude it for the reason that credit risk is generally viewed as a risk of the entity rather than a risk specific to the liability.⁹

Whether own credit risk is considered a risk specific to a liability may depend, at least in part, on the way in which an entity measures a provision (i.e. how an entity determines the best estimate of the expenditure required to settle the obligation). As discussed at 4.1 above, the measurement objective in IAS 37 is not precise, and the term ‘best estimate’ may be interpreted in different ways. For example, an entity may determine the best estimate by assuming that the obligation will be cash settled by negotiation with the party to whom the obligation is owed. In this case, own credit risk might be considered relevant if the party to whom the obligation is owed would be willing to negotiate a lower settlement amount if the entity risks defaulting in the future. If an entity has exercised significant judgement in determining the best estimate of the expenditure required to settle an obligation, and whether own credit risk is taken into account, consideration should be given to the disclosure requirements in paragraph 122 of IAS 1. See Chapter 3 at 5.1.1.B.

4.3.4 Pre-tax discount rate

Since IAS 37 requires provisions to be measured before tax, it follows that cash flows should be discounted at a pre-tax discount rate. [IAS 37.47]. No further explanation of this is given in the standard.

This is probably because, in reality, the use of a pre-tax discount rate will be most common. Supposing, for example, that the risk-free rate of return is being used, then the discount rate used will be a government bond rate. This rate will be obtained gross. Thus, the idea of trying to determine a pre-tax rate (for example by obtaining a required post-tax rate of return and adjusting it for the tax consequences of different cash flows) will seldom be relevant.

The calculation is illustrated in the following example.

4.3.5 Unwinding of the discount

IAS 37 indicates that where discounting is used, the carrying amount of a provision increases in each period to reflect the passage of time, and that this increase is recognised as a borrowing cost. [IAS 37.60]. This is the only guidance that the standard gives on the unwinding of the discount. IFRIC 1 in relation to provisions for decommissioning, restoration and similar liabilities requires that the periodic unwinding of the discount is recognised in profit or loss as a finance cost as it occurs. The Interpretations Committee concluded that the unwinding of the discount is not a borrowing cost for the purposes of IAS 23 – Borrowing Costs – and thus cannot be capitalised under that standard. [IFRIC 1.8]. It noted that IAS 23 addresses funds borrowed specifically for the purpose of obtaining a particular asset and agreed that a decommissioning liability does not fall within this description since it does not reflect funds borrowed. Accordingly, the Interpretations Committee concluded that the unwinding of the discount is not a borrowing cost as defined in IAS 23. [IFRIC 1.BC26].

However, there is no discussion of the impact that the original selection of discount rate can have on its unwinding, that is, the selection of real versus nominal rates, and risk-free versus risk-adjusted rates and the fact that these different discount rates will unwind differently. This is best illustrated by way of an example.

Although the total expense in each year is the same under either method, what will be different is the allocation of the change in provision between operating costs (assuming the original provision was treated as an operating expense) and finance charges. It can be seen from the second table in the above example that using the real discount rate will give rise to a much lower finance charge each year. However, this does not lead to a lower provision in the statement of financial position at the end of each year. Provisions have to be revised annually to reflect the current best estimate of the obligation. [IAS 37.59]. Thus, the provision in the above example at the end of each year needs to be adjusted to reflect current prices at that time (and any other adjustments that arise from changes in the estimate of the provision), as well as being adjusted for the unwinding of the discount. For example, the revised provision at the end of Year 1 is €100,173, being €105,000 discounted for two years at 2.381%. After allowing for the unwinding of the discount, this required an additional provision of €4,770. Alternatively, entities may take the view that the amount included as a borrowing cost should be based on the nominal discount rate. The basis for this argument is that IAS 37 requires the increase in a provision due to the passage of time to be recognised as a borrowing cost [IAS 37.60] and inflation, which is reflected in the nominal rate but not the real rate, is part of the time value of money.

A more significant difference will arise where the recognition of the original provision is included as part of the cost of property, plant or equipment, rather than as an expense, such as when a decommissioning provision is recognised. In that case, using a real discount rate will result initially in a lower charge to the income statement, since under IFRIC 1 any revision to the estimate of the provision is not taken to the income statement but is treated as an adjustment to the carrying value of the related asset, which is then depreciated prospectively over the remaining life of the asset (see 6.3.1 below).

A similar issue arises with the option of using the risk-free or the risk-adjusted discount rate. However, this is a more complex problem, because it is not clear what to do with the risk-adjustment built into the provision. This is illustrated in the following example.

In this example, the unwinding of different discount rates gives rise to different provisions. The difference of £10 (£22 – £12) relates to the risk adjustment that has been made to the provision. However, as the actual date of settlement comes closer, the estimates of the range of possible outcomes (and accordingly the expected value of the outflow) and the premium the entity would accept for certainty would need to be re-estimated. As such, the effect of any initial difference related to the decision to apply a risk-free or risk-adjusted rate will be lost in the other estimation adjustments that would be made over time.

In circumstances where a provision has been discounted using a negative discount rate (see 4.3.2 above), entities will need to give some consideration to how the unwinding of the discount should be presented in the statement of comprehensive income. The presentation of income and expense resulting from negative interest has been discussed by the Interpretations Committee in the context of financial instruments. Whilst that topic was not taken on to the Interpretations Committee's agenda, the Committee noted in their agenda decision that the expense arising on a financial asset because of a negative interest rate should not be presented as interest revenue, but in an appropriate expense classification.¹⁰ Whilst negative interest rates on financial liabilities were not explicitly addressed in the agenda decision, the Interpretations Committee noted during discussions that an inflow of economic benefits arising from a liability cannot be characterised as revenue.¹¹ In our view, similar considerations are appropriate to the presentation of the unwinding of the discount on a provision that has been discounted using a negative nominal discount rate.

4.3.6 The effect of changes in interest rates on the discount rate applied

The standard requires the discount rate to reflect current market assessments of the time value of money. [IAS 37.47]. This means that where interest rates change, the provision should be recalculated on the basis of revised interest rates (see Example 26.11 below).

Any revision in the interest rate will give rise to an adjustment to the carrying value of the provision in addition to the unwinding of the previously estimated discount. The standard requires these movements to be disclosed in the notes to the financial statements (see 7.1 below). [IAS 37.84(e)]. However, the standard does not explicitly say how the effect of changes in interest rates should be classified in the income statement. We believe that this element should be treated separately from the effect of the passage of time, with only the charge for unwinding of the discount being classified as a finance cost. Any adjustment to the provision as a result of revising the discount rate is a change in accounting estimate, as defined in IAS 8 – Accounting Policies, Changes in Accounting Estimates and Errors (see Chapter 3 at 4.5). Accordingly, it should be reflected in the line item of the income statement to which the expense establishing the provision was originally taken and not as a component of the finance cost. Indeed, this is the approach required by IFRIC 1 in relation to provisions for decommissioning, restoration and similar liabilities in relation to assets measured using the cost model (see 6.3.1 below). However, in that case the original provision gives rise to an asset rather than an expense, so any subsequent adjustment is not included in profit or loss, [IAS 8.36], but added to or deducted from the cost of the asset to which it relates. [IFRIC 1.5]. The adjusted depreciable amount of the asset is then depreciated prospectively over its remaining useful life. [IFRIC 1.7]. Nevertheless, the effect is distinguished from the unwinding of the discount.

In addition, the standard gives no specific guidance on how or when this adjustment should be made. For example, it is unclear whether the new discount rate should be applied during the year or just at the year-end, and whether the rate should be applied to the new estimate of the provision or the old estimate. IFRIC 1 implies that the finance cost is adjusted prospectively from the date on which the liability is remeasured. Example 1 to IFRIC 1 states that if the change in the liability had resulted from a change in discount rate, instead of a change in the estimated cash flows, the change would still have been reflected in the carrying value of the related asset, but next year's finance cost would have reflected the new discount rate. [IFRIC 1.IE5]. This conclusion is consistent with the requirement in IAS 37 for the value of a provision to reflect the best estimate of the expenditure required to settle the obligation as at the end of the reporting period, [IAS 37.36], as illustrated in the following example.

In Year 2, the finance charge is based on the previous estimate of the discount rate and the revision to the estimate of the provision would be charged to the same line item in the income statement that was used to establish the provision of €93,184 at the start of Year 1.

Where market rates of interest are more volatile, entities may decide to reassess the applicable discount rate for a provision during an annual reporting period. Equally, it would be appropriate to revise this assessment as at the end of any interim reporting period during the financial year to the extent that the impact is material.

6.1.1 Definition

IAS 37 defines a restructuring as ‘a programme that is planned and controlled by management, and materially changes either:

1. the scope of a business undertaken by an entity; or
2. the manner in which that business is conducted’. [IAS 37.10].

This is said to include:

1. the sale or termination of a line of business;
2. the closure of business locations in a country or region or the relocation of business activities from one country or region to another;
3. changes in management structure, for example, eliminating a layer of management; and
4. fundamental reorganisations that have a material effect on the nature and focus of the entity's operations. [IAS 37.70].

This definition is very wide and whilst it may be relatively straightforward to establish whether an operation has been sold, closed or relocated, the determination of whether an organisational change is fundamental, material or just part of a process of continuous improvement is a subjective judgement. Whilst organisational change is a perennial feature in most business sectors, entities could be tempted to classify all kinds of operating costs as restructuring costs and thereby invite the user of the financial statements to perceive them in a different light from the ‘normal’ costs of operating in a dynamic business environment. Even though the requirements in IAS 37 prevent such costs being recognised too early, the standard still leaves the question of classification open to judgement. As such there can be a tension between the permitted recognition of expected restructuring costs, subject to meeting the criteria set out at 6.1.2 below, and the general prohibition in IAS 37 against provision for future operating losses, which is discussed at 5.1 above.

IAS 37 emphasises that when a restructuring meets the definition of a discontinued operation under IFRS 5, additional disclosures may be required under that standard (see Chapter 4 at 3). [IAS 37.9].

6.1.2 Recognition of a restructuring provision

IAS 37 requires that restructuring costs are recognised only when the general recognition criteria in the standard are met, i.e. there is a present obligation (legal or constructive) as a result of a past event, in respect of which a reliable estimate can be made of the probable cost. [IAS 37.71]. The standard's specific requirements for the recognition of a provision for restructuring costs seek to define the circumstances that give rise to a constructive obligation and thereby restrict the recognition of a provision to cases when an entity:

1. has a detailed formal plan for the restructuring identifying at least:
   1. the business or part of a business concerned;
   2. the principal locations affected;
   3. the location, function, and approximate number of employees who will be compensated for terminating their services;
   4. the expenditures that will be undertaken; and
   5. when the plan will be implemented; and
2. has raised a valid expectation in those affected that it will carry out the restructuring by starting to implement that plan or announcing its main features to those affected by it. [IAS 37.72].

The standard gives examples of the entity's actions that may provide evidence that the entity has started to implement a plan, quoting the dismantling of plant or selling of assets, or the public announcement of the main features of the plan. However, it also emphasises that the public announcement of a detailed plan to restructure will not automatically create an obligation; the important principle is that the announcement is made in such a way and in sufficient detail to give rise to valid expectations in other parties such as customers, suppliers and employees that the restructuring will be carried out. [IAS 37.73].

The standard also suggests that for an announced plan to give rise to a constructive obligation, its implementation needs to be planned to begin as soon as possible and to be completed in a timeframe that makes significant changes to the plan unlikely. Any extended period before commencement of implementation, or if the restructuring will take an unreasonably long time, will mean that recognition of a provision is premature, because the entity is still likely to have a chance of changing the plan. [IAS 37.74].

In summary, these conditions require the plan to be detailed and specific, to have gone beyond the directors' powers of recall and to be put into operation without delay or significant alteration.

The criteria set out above for the recognition of provisions mean that a board decision, if it is the only relevant event arising before the end of the reporting period, is not sufficient. This message is reinforced specifically in the standard, the argument being made that a constructive obligation is not created by a management decision. There will only be a constructive obligation where the entity has, before the end of the reporting period:

1. started to implement the restructuring plan; or
2. announced the main features of the restructuring plan to those affected by it in a sufficiently specific manner to raise a valid expectation in them that the entity will carry out the restructuring. [IAS 37.75].

If the restructuring is not started or announced in detail until after the end of the reporting period, no provision is recognised. Instead, the entity discloses a non-adjusting event after the reporting period. [IAS 37.75, IAS 10.22(e)].

The following examples in IAS 37 illustrate how a constructive obligation for a restructuring may or may not be created.

The standard acknowledges that there will be circumstances where a board decision could trigger recognition, but not on its own. Only if earlier events, such as negotiations with employee representatives for termination payments or with purchasers for the sale of an operation, have been concluded subject only to board approval would the decision of the board create an obligation. In such circumstances, it is reasoned that when board approval has been obtained and communicated to the other parties, the entity is committed to restructure, assuming all other conditions are met. [IAS 37.76].

There is also discussion in the standard of the situation that may arise in some countries where, for example, employee representatives may sit on the board, so that a board decision effectively communicates the decision to them, which may result in a constructive obligation to restructure. [IAS 37.77].

In practice it can be very difficult to determine whether it is appropriate to recognise a provision for the future costs of a restructuring programme. The determination of whether an organisational change is fundamental, material or just part of a process of continuous improvement is a subjective judgement. Once it has been established that the activities in question constitute a restructuring rather than an ongoing operating cost, it can be difficult to determine whether management's actions before the reporting date have been sufficient to have ‘raised a valid expectation in those affected’. [IAS 37.72]. Even if a trigger point is easily identifiable, such as the date of an appropriately detailed public announcement, it might not necessarily commit management to the whole restructuring, but only to specific items of expenditure such as redundancy costs. When the announcement is less clear, referring for example to consultations, negotiations or voluntary arrangements, particularly with employees, judgement is required. Furthermore, taken on its own, the ‘valid expectation’ test is at least as open to manipulation as one based on the timing of a board decision. Entities anxious to accelerate or postpone recognition of a liability could do so by advancing or deferring an event that signals such a commitment, such as a public announcement, without any change to the substance of their position.

In these situations it is important to consider all the related facts and circumstances and not to ‘home in’ on a single recognition criterion. The objective of the analysis is to determine whether there is a past obligating event at the reporting date. The guidance in the standard about restructuring, referring as it does to constructive obligations and valid expectations is ultimately aimed at properly applying the principle in IAS 37 that only those obligations arising from past events and existing independently of an entity's future actions are recognised as provisions. [IAS 37.19]. In essence, a restructuring provision qualifies for recognition if, as at the reporting date, it relates to a detailed plan of action from which management cannot realistically withdraw.

6.1.3 Recognition of obligations arising from the sale of an operation

IAS 37 has some further specific rules governing when to recognise an obligation arising on the sale of an operation, stating that no obligation arises for the sale of an operation until the entity is committed to the sale, i.e. there is a binding sale agreement. [IAS 37.78]. Thus a provision cannot be made for a loss on sale unless there is a binding sale agreement by the end of the reporting period. The standard says that this applies even when an entity has taken a decision to sell an operation and announced that decision publicly, it cannot be committed to the sale until a purchaser has been identified and there is a binding sale agreement. Until there is such an agreement, the entity will be able to change its mind and indeed will have to take another course of action if a purchaser cannot be found on acceptable terms. [IAS 37.79].

Even in cases where it is part of a larger restructuring that qualifies for recognition under IAS 37, an obligation arising from the sale is not recognised until there is a binding sale agreement. Instead, the assets of the operation must be reviewed for impairment under IAS 36. This may therefore mean that an expense is recorded in the income statement; it is just that the expense gives rise to a reduction of the carrying amount of assets rather than the recognition of a liability. The standard also recognises that where a sale is only part of a restructuring, the entity could be committed to the other parts of restructuring before a binding sale agreement is in place. [IAS 37.79]. Hence, the costs of the restructuring will be recognised over different reporting periods.

6.1.4 Costs that can (and cannot) be included in a restructuring provision

Having met the specific tests in the standard for the recognition of a restructuring provision at the end of the reporting period, IAS 37 imposes further criteria to restrict the types of cost that can be provided for. Presumably these additional restrictions are intended to ensure that the entity does not contravene the general prohibition in IAS 37 against provision for future operating losses. [IAS 37.63].

A restructuring provision should include only the direct expenditures arising from the restructuring, which are those that are both:

1. necessarily entailed by the restructuring; and
2. not associated with the ongoing activities of the entity. [IAS 37.80].

The standard gives specific examples of costs that may not be included within the provision, because they relate to the future conduct of the business. Such costs include:

1. retraining or relocating continuing staff;
2. marketing; or
3. investment in new systems and distribution networks. [IAS 37.81].

Because these costs relate to the future conduct of the business, they are recognised on the same basis as if they arose independently of a restructuring. [IAS 37.81]. In most cases, this means that the costs are recognised as the related services are provided.

This example shows that individual classes of expenditure should be disaggregated into components that distinguish those elements associated with ongoing activities. Even if expenditure would not have been incurred without the restructuring activity, its association with ongoing activities means that it is ineligible for inclusion in a provision. IAS 37 requires the cost to be both necessarily entailed by the restructuring and not associated with the ongoing activities of the entity. [IAS 37.80].

For that reason, whilst the cost of making employees redundant is an eligible restructuring cost, any incremental amounts paid to retain staff to ensure a smooth transition of operations from one location to another are not eligible because they are incurred to facilitate ongoing activities. IAS 19 requires these to be treated as short-term employee benefits to the extent that they are expected to be settled within 12 months after the end of the reporting period.¹⁶ Similarly, whilst the costs of dismantling plant and equipment intended to be scrapped is an eligible restructuring cost, the costs of dismantling plant and equipment intended to be relocated and installed at the new site is ineligible, because it is associated with ongoing activities.

A further rule in IAS 37 is that the provision should not include identifiable future operating losses up to the date of the restructuring, unless they relate to an onerous contract. [IAS 37.82]. This means that even if the operation being reorganised is loss-making, its ongoing costs are not provided for. This is consistent with the general prohibition against the recognition of provisions for future operating losses. [IAS 37.63].

The general requirement in the standard that gains from the expected disposal of assets cannot be taken into account in the measurement of provisions, [IAS 37.51], is also relevant to the measurement of restructuring provisions, even if the sale of the asset is envisaged as part of the restructuring. [IAS 37.83]. Whilst the expected disposal proceeds from asset sales might have been a significant element of the economic case for a restructuring, the income from disposal is not anticipated just because it is part of a restructuring plan.

6.2.1 Onerous leases

As discussed at 2.2.1.B above, IAS 37 specifically applies to leases only in limited circumstances. IAS 37 specifically applies only to:

• leases that become onerous before the commencement date of the lease; and
• short-term leases (as defined in IFRS 16) and leases for which the underlying asset is of low value that are accounted for in accordance with paragraph 6 of IFRS 16 and that have become onerous. [IAS 37.5(c)].

The IASB noted in paragraph BC 72 of IFRS 16 that it ‘…decided not to specify any particular requirements in IFRS 16 for onerous contracts. The IASB made this decision because …(a) for leases that have already commenced, no requirements are necessary. After the commencement date, an entity will appropriately reflect an onerous lease contract by applying the requirements of IFRS 16. For example, a lessee will determine and recognise any impairment of right-of-use assets applying IAS 36 Impairment of Assets.’ However, this raises the question how an entity should account for onerous variable lease payments that do not depend on an index or a rate and are not included in the measurement of right-of-use assets or lease liabilities under IFRS 16 (e.g. variable payments related to property taxes or non-lease components that have not been separated from the lease components, in accordance with paragraph 15 of IFRS 16). It is not immediately obvious whether such onerous variable lease payments fall within the scope of IFRS 16 or IAS 37.

As discussed in Chapter 20, at 7.1.8.A, we believe that contractual variable lease payments that are not included in the lease liability (such as those that are not based on an index or rate) should be reflected in cash flow forecasts used for value-in-use calculations for impairment testing purposes. Once the right-of-use is fully impaired, entities may need to use their judgement to determine whether any further onerous lease provisions should be recognised under the requirements of IAS 37. If an entity exercises significant judgement in determining the most appropriate approach, consideration should be given to the disclosure requirements in paragraph 122 of IAS 1. See Chapter 3 at 5.1.1.B.

6.2.2 Contracts with customers that are, or have become, onerous

As IFRS 15 contains no specific requirements to address contracts with customers that are, or have become, onerous, IAS 37 applies to such contracts. [IAS 37.5(g)]. In assessing whether a contract is onerous, an entity should compare the unavoidable costs of meeting the obligations under the contract to the economic benefits expected to be received under it. The unavoidable costs under the contract are the lower of the cost of fulfilling the contract and any compensation or penalties arising from failure to fulfil the contract. [IAS 37.68]. As discussed at 6.2 above, in December 2018, the IASB issued an exposure draft that proposed to clarify that the costs of fulfilling a contract comprise the costs that relate directly to the contract, rather than only the incremental costs of the contract.²¹

One question that may arise, is how an entity should account for an onerous contract with a customer when the contract includes several ‘over time’ performance obligations that are satisfied consecutively. Since the requirements for onerous contracts are outside the scope of IFRS 15, an entity's accounting for onerous contracts does not affect the accounting for its revenue from contracts with customers in accordance with IFRS 15. Therefore, an entity must use an ‘overlay’ approach, which consists of two steps:

1. apply the requirements of IFRS 15 to measure progress in satisfying each performance obligation, and account for the related costs when incurred in accordance with the applicable standards; and
2. at the end of each reporting period, apply IAS 37 to determine if the remaining contract as a whole is onerous. If the entity concludes that the remaining contract as a whole is onerous, it recognises a provision only to the extent that the amount of the remaining unavoidable costs under the contract exceed the remaining economic benefits to be received under it.

This approach is illustrated below.

The effect of an onerous contract provision, or a change in the provision, is recognised as an expense in profit or loss and not as an adjustment to revenue.

As provisions for onerous contracts with customers are in the scope of IAS 37, they should be classified as provisions in the balance sheet and disclosed in accordance with IAS 37 (see 7.1 below). Onerous contract provisions are outside the scope of IFRS 15 and should not be included within contract liabilities.

Since the definition of an onerous contract in IAS 37 refers only to a contract, the unit of account to determine whether an onerous contract exists is the contract itself, rather than the performance obligations identified in accordance with IFRS 15. [IAS 37.10]. As a result, the entity must consider the entire remaining contract, including remaining revenue to be recognised for unsatisfied, or partially unsatisfied, performance obligations and the remaining costs to fulfil those performance obligations.

As discussed at 6.2 above, there is diversity in practice in how entities determine the economic benefits expected to be received under a contract for the purposes of assessing whether a contract is onerous. Where a contract with a customer contains an element of variable consideration, it is likely that judgement will be required in assessing the economic benefits expected to be received under a contract. This may particularly be the case where an entity is required to constrain the estimate of variable consideration to be included in the transaction price for revenue recognition purposes under IFRS 15. Variable consideration under IFRS 15 is discussed further in Chapter 29 at 2.2.

6.3.1 Changes in estimated decommissioning costs (IFRIC 1)

IAS 37 requires provisions to be revised annually to reflect the current best estimate of the provision. [IAS 37.59]. However, the standard gives no guidance on accounting for changes in the decommissioning provision. Similarly, IAS 16 is unclear about the extent to which an item's carrying amount should be affected by changes in the estimated amount of dismantling and site restoration costs that occur after the estimate made upon initial measurement. This was addressed by the IASB with the publication of IFRIC 1 in May 2004. [IFRIC 1.1].

IFRIC 1 applies to any decommissioning, restoration or similar liability that has been both included as part of the cost of an asset measured in accordance with IAS 16 or as part of the cost of a right-of-use asset in accordance with IFRS 16 and recognised as a liability in accordance with IAS 37. [IFRIC 1.2]. It addresses how the effect of the following events that change the measurement of an existing decommissioning, restoration or similar liability should be accounted for:

1. a change in the estimated outflow of resources embodying economic benefits (e.g. cash flows) required to settle the obligation;
2. a change in the current market-based discount rate (this includes changes in the time value of money and the risks specific to the liability); and
3. an increase that reflects the passage of time (also referred to as the unwinding of the discount). [IFRIC 1.3].

IFRIC 1 requires that (c) above, the periodic unwinding of the discount, is recognised in profit or loss as a finance cost as it occurs. [IFRIC 1.8]. The Interpretations Committee concluded that the unwinding of the discount is not a borrowing cost as defined in IAS 23, and thus cannot be capitalised under that standard. [IFRIC 1.BC26‑27].

For a change caused by (a) or (b) above, however, the adjustment is taken to the income statement only in specific circumstances. Any revision to the provision (other than to reflect the passage of time) is first recognised in the carrying value of the related asset or in other comprehensive income, depending on whether the asset is measured at cost or using the revaluation model. [IFRIC 1.4‑7].

If the related asset is measured using the cost model, the change in the liability should be added to or deducted from the cost of the asset to which it relates. Where the change gives rise to an addition to cost, the entity should consider the need to test the new carrying value for impairment. This is particularly relevant for assets approaching the end of their useful life, as their remaining economic benefits are often small compared to the potential changes in the related decommissioning liability. Reductions over and above the remaining carrying value of the asset are recognised immediately in profit or loss. [IFRIC 1.5]. The adjusted depreciable amount of the asset is then depreciated prospectively over its remaining useful life. [IFRIC 1.7]. IFRIC 1 includes the following illustrative example.

In illustrating the requirements of the Interpretation, the example in IFRIC 1 reduces the carrying value of the whole asset (comprising its construction cost and decommissioning cost) by the reduction in the present value of the decommissioning provision. The solution set out in the example does not treat the decommissioning element as a separate component of the asset. Had this been the case, the component would have had accumulated depreciation as at 31 December 2020 of $2,500,000 ($10,000,000 × 10/40), giving a carrying amount of $7,500,000 at that date and a gain of $500,000 when reduced by the decrease in the provision of $8,000,000. Accordingly, we believe that the example in IFRIC 1 indicates that it would not be appropriate to recognise any gain until the carrying value of the whole asset is extinguished.

If the related asset is measured using the revaluation model, changes in the liability alter the revaluation surplus or deficit previously recognised for that asset. Decreases in the provision are recognised in other comprehensive income and increase the value of the revaluation surplus in respect of the asset, except that:

1. a decrease in the provision should be recognised in profit or loss to the extent that it reverses a previous revaluation deficit on that asset that was recognised in profit or loss; and
2. if a decrease in the provision exceeds the carrying amount of the asset that would have been recognised under the cost model, the excess should be recognised in profit or loss. [IFRIC 1.6].

Increases in the provision are recognised in profit or loss, except that they should be recognised in other comprehensive income, and reduce the revaluation surplus, to the extent of any credit balance existing in the revaluation surplus in respect of that asset. Changes in the provision might also indicate the need for the asset (and therefore all assets in the same class) to be revalued. [IFRIC 1.6].

The illustrative examples in IFRIC 1 address this alternative.

The depreciation expense for 2014 is therefore $3,420,000 ($126,600,000 ÷ 37) and the discount expense for 2013 is $580,000 (5% of $11,600,000). On 31 December 2014, the decommissioning liability (before any adjustment) is $12,180,000 and the discount rate has not changed. However, on that date, the entity estimates that, as a result of technological advances, the present value of the decommissioning liability has decreased by $5,000,000. Accordingly, the entity adjusts the decommissioning liability to $7,180,000. To determine the extent to which any of the change to the provision is recognised in profit or loss, the entity has to keep a record of revaluations previously recognised in profit or loss; the carrying amount of the asset that would have been recognised under the cost model; and the previous revaluation surplus relating to that asset. [IFRIC 1.6]. In this example, the whole of the adjustment is taken to revaluation surplus, because it does not exceed the carrying amount that would have been recognised for the asset under the cost model of $103,000 (see below). [IFRIC 1.IE6‑10].

In addition, the entity decides that a full valuation of the asset is needed at 31 December 2014, in order to ensure that the carrying amount does not differ materially from fair value. Suppose that the asset is now valued at $107,000,000, which is net of an allowance of $7,180,000 for the reduced decommissioning obligation. The valuation of the asset for financial reporting purposes, before deducting this allowance, is therefore $114,180,000. The effect on the revaluation reserve of the revision to the estimate of the decommissioning provision and the new valuation can be illustrated in the table below. [IFRIC 1.IE11‑12].

As indicated at 4.3.6 above, IAS 37 is unclear whether a new discount rate should be applied during the year or just at the year-end, and whether the rate should be applied to the new estimate of the provision or the old estimate. Although IFRIC 1 requires that changes in the provision resulting from a change in the discount rate is added to, or deducted from, the cost of the related asset in the current period, it does not deal specifically with these points. However, Example 1 in the illustrative examples to IFRIC 1 indicates that a change in discount rate would be accounted for in the same way as other changes affecting the estimate of a provision for decommissioning, restoration and similar liabilities. That is, it is reflected as a change in the liability at the time the revised estimate is made and the new estimate is discounted at the revised discount rate from that point on. [IFRIC 1.IE5].

When accounting for revalued assets to which decommissioning liabilities attach, the illustrative example in IFRIC 1 states that it is important to understand the basis of the valuation obtained. For example:

1. if an asset is valued on a discounted cash flow basis, some valuers may value the asset without deducting any allowance for decommissioning costs (a ‘gross’ valuation), whereas others may value the asset after deducting an allowance for decommissioning costs (a ‘net’ valuation), because an entity acquiring the asset will generally also assume the decommissioning obligation. For financial reporting purposes, the decommissioning obligation is recognised as a separate liability, and is not deducted from the asset. Accordingly, if the asset is valued on a net basis, it is necessary to adjust the valuation obtained by adding back the allowance for the liability, so that the liability is not counted twice. This is the case in Example 26.18 above;
2. if an asset is valued on a depreciated replacement cost basis, the valuation obtained may not include an amount for the decommissioning component of the asset. If it does not, an appropriate amount will need to be added to the valuation to reflect the depreciated replacement cost of that component. [IFRIC 1.IE7].

6.3.2 Changes in legislation after construction of the asset

The scope of IFRIC 1 is set out in terms of any existing decommissioning, restoration or similar liability that is both recognised as part of the cost of the asset under IAS 16 or as part of the cost of a right-of-use asset in accordance with IFRS 16; and recognised as a liability in accordance with IAS 37. [IFRIC 1.2]. The Interpretation does not address the treatment of obligations arising after the asset has been constructed, for example as a result of changes in legislation. [IFRIC 1.BC23]. Nevertheless, in our opinion the cost of the related asset should be measured in accordance with the principles set out in IFRIC 1 regardless of whether the obligation exists at the time of constructing the asset or arises later in its life.

As discussed at 3.3 in Chapter 18, IAS 16 makes no distinction in principle between the initial costs of acquiring an asset and any subsequent expenditure upon it. In both cases any and all expenditure has to meet the recognition rules, and be expensed in profit or loss if it does not. IAS 16 states that the cost of an item of property, plant and equipment includes ‘the initial estimate of the costs of dismantling and removing the item and restoring the site on which it is located, the obligation for which an entity incurs either when the item is acquired or as a consequence of having used the item during a particular period for purposes other than to produce inventories during that period.’ [IAS 16.16(c)]. For example, the introduction of new legislation to require the clean-up of sites that cease to be used as gasoline filling stations would give rise to the recognition of a decommissioning provision and, to the extent that the clean-up obligation arose as a result of the construction of the filling stations, an increase in the carrying value of the properties. Similarly, IFRS 16 states that the cost of a right-of-use asset includes ‘an estimate of costs to be incurred by the lessee in dismantling and removing the underlying asset, restoring the site on which it is located or restoring the underlying asset to the condition required by the terms and conditions of the lease, unless those costs are incurred to produce inventories. The lessee incurs the obligation for those costs either at the commencement date or as a consequence of having used the underlying asset during a particular period.’ [IFRS 16.24(d)].

Both IAS 16 and IFRS 16 require an entity to apply IAS 2 to the costs of obligations for dismantling, removing and restoring the site on which an item is located that are incurred during a particular period as a consequence of having used the item to produce inventories during that period. [IAS 16.18, IFRS 16.25]. For example, the cost of restoring the site of a quarry would be reflected as part of the cost of the aggregate extracted from it, and not added to the carrying value of the site. Accordingly, if an entity previously had no obligation to restore the site and new legislation was introduced after 25% of the site had been excavated and 80% of that output had been sold, then 80% of the new estimate of the restoration cost would be expensed; 20% added to the cost of inventory; and none added to the carrying value of the site.

When changes in legislation give rise to a new decommissioning, restoration or similar liability that is added to the carrying amount of the related asset, it would be appropriate to perform an impairment review in accordance with IAS 36 (see Chapter 20).

6.3.3 Funds established to meet an obligation (IFRIC 5)

Some entities may participate in a decommissioning, restoration or environmental rehabilitation fund, the purpose of which is to segregate assets to fund some or all of the costs of decommissioning for which the entity has to make a provision under IAS 37. IFRIC 5 was issued in December 2004 to address this issue, referring to decommissioning to mean not only the dismantling of plant and equipment but also the costs of undertaking environmental rehabilitation, such as rectifying pollution of water or restoring mined land. [IFRIC 5.1].

Contributions to these funds may be voluntary or required by regulation or law, and the funds may have one of the following common structures:

• funds that are established by a single contributor to fund its own decommissioning obligations, whether for a particular site, or for a number of geographically dispersed sites;
• funds that are established with multiple contributors to fund their individual or joint decommissioning obligations, where contributors are entitled to reimbursement for decommissioning expenses to the extent of their fund contributions plus any actual earnings on those contributions less their share of the costs of administering the fund. Contributors may have an obligation to make potential additional contributions, for example, in the event of the bankruptcy of another contributor;
• funds that are established with multiple contributors to fund their individual or joint decommissioning obligations when the required level of contributions is based on the current activity of a contributor, but the benefit obtained by that contributor is based on its past activity. In such cases there is a potential mismatch in the amount of contributions made by a contributor (based on current activity) and the value realisable from the fund (based on past activity). [IFRIC 5.2].

Such funds generally have the following features:

• the fund is separately administered by independent trustees;
• entities (contributors) make contributions to the fund, which are invested in a range of assets that may include both debt and equity investments, and are available to help pay the contributors' decommissioning costs. The trustees determine how contributions are invested, within the constraints set by the fund's governing documents and any applicable legislation or other regulations;
• the contributors retain the obligation to pay decommissioning costs. However, contributors are able to obtain reimbursement of decommissioning costs from the fund up to the lower of the decommissioning costs incurred and the entity's share of assets of the fund; and
• the contributors may have restricted or no access to any surplus of assets of the fund over those used to meet eligible decommissioning costs. [IFRIC 5.3].

IFRIC 5 applies to accounting in the financial statements of a contributor for interests arising from decommissioning funds that have both the following features:

• the assets are administered separately (either by being held in a separate legal entity or as segregated assets within another entity); and
• a contributor's right to access the assets is restricted. [IFRIC 5.4].

A residual interest in a fund that extends beyond a right to reimbursement, such as a contractual right to distributions once all the decommissioning has been completed or on winding up the fund, may be an equity instrument within the scope of IFRS 9, and is not within the scope of IFRIC 5. [IFRIC 5.5].

The issues addressed by IFRIC 5 are:

1. How should a contributor account for its interest in a fund?
2. When a contributor has an obligation to make additional contributions, for example, in the event of the bankruptcy of another contributor, how should that obligation be accounted for? [IFRIC 5.6]

6.3.4 Interaction of leases with asset retirement obligations

An entity may sometimes (expect to) use a leased asset to carry out decommissioning or remediation work for which it has recognised a decommissioning, remediation or asset retirement provision. For example, at the end of life of an oil field, an entity may use a leased ship or rig to undertake the plugging and abandonment of oil wells. The entity may have included the estimated cost of plugging and abandonment of the wells in the decommissioning provision set up at the commencement of oil production. If an entity uses leased assets to carry out decommissioning or remediation work for which it recognised a decommissioning, remediation, or asset retirement provision, the question arises as to whether, at lease commencement, the entity's recognition of a lease liability for the leased assets in accordance with IFRS 16 results in derecognition of the asset retirement obligation recognised in the balance sheet. Given that, prior to the commencement of any asset retirement obligation related activities, the entity still has an obligation to rehabilitate under IAS 37, it cannot derecognise the asset retirement obligation. Instead, it now has a separate lease liability for the financing of the lease of the asset. Accordingly, acquiring the right-of-use asset does not result in the derecognition of the asset retirement obligation liability, rather it would be the activity undertaken or output of the asset which would ultimately settle the asset retirement obligation. This is discussed further in Chapter 43 at 17.7.

6.8.1 Scope of IFRIC 21

A levy is defined as an outflow of resources embodying economic benefits that is imposed by governments on entities in accordance with legislation, other than:

1. those outflows of resources that are within the scope of other Standards (such as income taxes that are within the scope of IAS 12); and
2. fines or other penalties that are imposed for breaches of the legislation. [IFRIC 21.4].

In addition to income taxes (see Chapter 33) and fines, the Interpretation does not apply to contractual arrangements with government in which the entity acquires an asset (see Chapters 17 and 18) or receives services; and it is not required to be applied to liabilities that arise from emission trading schemes (see 6.5 above). [IFRIC 21.5, 6]. Although IFRIC 21 does not apply to income taxes in scope of IAS 12, the Interpretations Committee concluded in 2006 that any taxes not within the scope of other standards (such as IAS 12) are within the scope of IAS 37. Therefore such taxes may be within the scope of IFRIC 21. [IFRIC 21.BC4].

IFRIC 21 was developed to address concerns over the timing of recognition for government-imposed levies in which the obligation to pay depended upon participation in a particular market on a specified date. However, the definition of levy in IFRIC 21 has resulted in the scope of the interpretation being broader than entities might have expected. The term ‘levy’ may not be widely used across jurisdictions, and may be referred to as a charge, duty or a tax, for example. However, it is not the terminology, but the nature of the payment, that should be considered when determining if it is in the scope of IFRIC 21.

Entities should consider all payments imposed by governments pursuant to legislation to determine whether they are in scope of IFRIC 21. The interpretation provides a broad definition of government, including municipal, provincial, state, federal or international governments or government agencies or organisations controlled or administered by government. [IFRIC 21.4].

IFRIC 21 clarifies that both levies that give rise to a liability under IAS 37, and levies whose timing or amounts are certain are within scope of the interpretation. [IFRIC 21.2]. Therefore, the scope of IFRIC 21 is broader than IAS 37. For example, a non-refundable fixed fee imposed by government payable at a specific date may be a levy within the scope of IFRIC 21.

In some cases, payments may pass through one or more non-governmental bodies or entities before being received by the government. In our view, IFRIC 21 does not distinguish between recipients of the payment; the key factor is whether the payment is required by law. Therefore, as long as the payments are required by law, they are generally considered to be imposed by the government. [IFRIC 21.4].

Some of the legislation relating to payments imposed by governments can be complex, so entities should carefully analyse the facts and circumstances to determine whether a payment falls within the scope of IFRIC 21. However, where entities are making payments for any of the following items, it may be necessary to assess for any potential IFRIC 21 impacts:

• taxes other than income taxes, e.g. property tax, land tax and capital-based tax;
• certain fees, concessions, contributions or royalty fees imposed on industries which are regulated by government, e.g. telecommunications, mining, airline, banking, insurance, dairy produce and energy and natural resources; and
• transaction taxes based on activity in a specified market, e.g. banking and insurance.

6.8.2 Recognition and measurement of levy liabilities

For levies within the scope of the Interpretation, the activity that creates the obligation under the relevant legislation to pay the levy is the obligating event for recognition purposes. [IFRIC 21.8]. In many cases this activity is related to the entity's participation in a relevant market at a specific date or dates. The Interpretation states that neither a constructive nor a present obligation arises as a result of being economically compelled to continue operating; or from any implication of continuing operations in the future arising from the use of the going concern assumption in the preparation of financial statements. [IFRIC 21.9, 10].

When a levy is payable progressively, for example as the entity generates revenues, the entity recognises a liability over a period of time on that basis. This is because the obligating event is the activity that generates revenues. [IFRIC 21.11]. If an obligation to pay a levy is triggered in full as soon as a minimum threshold is reached, such as when the entity commences generating sales or achieves a certain level of revenue, the liability is recognised in full on the first day that the entity reaches that threshold. [IFRIC 21.12]. If an entity pays over amounts to government before it is determined that an obligation to pay that levy exists, it recognises an asset. [IFRIC 21.14].

The table below summarises the illustrative examples that accompany IFRIC 21, which provide guidelines on how to account for the timing of the recognition for the various types of levies: [IFRIC 21.IE1]

As set out in the table above, when a levy is triggered progressively, for example, as the entity generates revenues, the entity recognises a liability over the period of time on that basis. Some examples of progressive-type levies are set out below.

When the legislation provides that a levy is triggered by an entity operating in a market only at the end of its annual reporting period, no liability is recognised until the last day of the annual reporting period. No amount is recognised before that date in anticipation of the entity still operating in the market. Accordingly, a provision would not be permitted to be recognised in interim financial statements if the obligating event occurs only at the end of the annual reporting period. [IFRIC 21.IE1 Example 3]. The accounting treatment in interim reports is discussed in Chapter 41 at 9.7.5.

6.8.3 Recognition of an asset or expense when a levy is recorded

IFRIC 21 only provides guidance on when to recognise a liability, which is the credit side of the journal entry. The interpretation specifically states that it does not address whether the debit side of the journal entry is an asset or an expense, [IFRIC 21.3], except in the case of prepaid levies. [IFRIC 21.14].

Prepayments may be fairly common in arrangements in which the legislation requires entities to pay levies in advance and where the obligating events for these levies are progressive. For example, property taxes that are paid in advance at a specified date (e.g. 1 January) for an obligating event that relates to future periods (e.g. 1 January to 31 December). In such instances, the entity would recognise the prepaid levy as an asset. [IFRIC 21.14]. In this scenario, the prepaid levy would then be amortised over the period.

Aside from prepaid levies, there are also instances when the assessment of expensing the liability or recognising a corresponding asset requires the application of other standards, such as IAS 2, IAS 16 or IAS 38. Given that levies are imposed by government and arise from non-exchange transactions, there would not typically be a clear linkage to future economic benefits. Consequentially, if the incurrence of the liability does not give rise to an identifiable future economic benefit to the entity, the recognition of an asset would be inappropriate as the definition of an asset would not be met. In such cases, the debit side would therefore be to an expense account.

In the case where asset recognition is appropriate under other IFRS standards, levies are generally not expected to give rise to a stand-alone asset in its own right, given that payments for the acquisition of goods or services are scoped out of IFRIC 21. [IFRIC 21.5]. However, a levy may form part of the acquisition costs of some other asset, provided it meets the asset recognition criteria in other IFRS standards. For example, an entity may be required to pay an import duty to the government under legislation for any large cargo trucks purchased from overseas. The entity uses the large cargo trucks as part of their operations to transport their goods to customers locally and therefore capitalises the trucks as part of property, plant and equipment under IAS 16. The import duty that is payable under the legislation may give rise to an IFRIC 21 levy, which would also be capitalised as part of the cost of the asset. [IAS 16.16(a)].

6.8.4 Payments relating to taxes other than income tax

As discussed at 6.8.1 above, IFRIC 21 does not apply to income taxes in scope of IAS 12. However, the Interpretations Committee concluded in 2006 that any taxes not within the scope of other standards (such as IAS 12) are within the scope of IAS 37. Therefore such taxes may be within the scope of IFRIC 21. [IFRIC 21.BC4].

In 2018, the Interpretations Committee discussed a fact pattern where an entity is in dispute with a tax authority in respect of a tax other than income tax. The entity determines that it is probable that it does not have an obligation for the disputed amount and consequently, it does not recognise a liability applying IAS 37. The entity nonetheless pays the disputed amount to the tax authority, either voluntarily or because it is required to do so. The entity has no right to a refund of the amount before resolution of the dispute. Upon resolution, either the tax authority returns the payment to the entity (if the outcome of the dispute is favourable to the entity) or the payment is used to settle the tax liability (if the outcome of the dispute is unfavourable to the entity).

In March 2018, the Interpretations Committee observed that the payment made by the entity gives rise to an asset as defined in the Conceptual Framework that was in issue at the time. The Conceptual Framework (2010) defines an asset as a resource controlled by the entity as a result of past events and from which future economic benefits are expected to flow to the entity. [CF(2010) 4.4(a)]. On making the payment, the entity has a right to receive future economic benefits either in the form of cash or by using the payment to settle the tax liability. The payment is not a contingent asset as defined in IAS 37 because it is an asset, and not a possible asset, of the entity. The entity therefore recognises an asset when it makes the payment to the tax authority.³⁴

Also in March 2018, the IASB issued a revised Conceptual Framework for Financial Reporting. The revised framework became effective immediately for the IASB and IFRS Interpretations Committee and is effective from 1 January 2020 for entities that use the Conceptual Framework to develop accounting policies when no IFRS standard applies to a particular transaction. Paragraph 4.3 of the revised Conceptual Framework defines an asset as a present economic resource controlled by the entity as a result of past events. In May 2018, the Interpretations Committee considered whether the new Conceptual Framework would change its observation that the payment made by the entity gives rise to an asset, and not a possible asset (contingent asset), of the entity. The Committee tentatively concluded that the new Conceptual Framework would not change its previous observation. The Committee also noted that this matter had raised questions about the role of the new Conceptual Framework and decided to consult the Board on these questions.³⁵