IFRS 15 Stand-Alone Selling Price and Allocating the Transaction Price: Step 4 in Full

1. What does Step 4 of IFRS 15 actually require, and why is the contract's own price list not the answer?

The fault this unit exists to fix

The prior version of this material allocated a bundled contract at the prices stated in the contract schedule and never mentioned the relative stand-alone selling price rule. That is audit fault F-06. It is worth naming plainly because it is the single most common Step 4 error in practice, and because it does not look like an error. The stated prices foot to the contract total. The revenue recognised in total over the life of the contract is right. Nothing in the trial balance is out of place. What is wrong is the pattern: revenue lands in the wrong period, in the wrong amount, for the wrong obligation.

The reason it happens is commercial. Sales teams price bundles to win them. The internal margin on a piece of hardware may be nil while the margin on the attached three year support contract is fifty per cent, or the reverse. Contract schedules are drafted to make a discount look like it applies to the item the customer cares about, or to keep a headline unit price intact for the next negotiation. None of that is improper. It simply means the contract schedule is a commercial artefact, and IFRS 15.77 refuses to let a commercial artefact settle an accounting question by default.

What Step 4 does and does not decide

Step 4 decides amount by obligation. It does not decide timing, and it does not decide whether an obligation exists. Those are Step 2 and Step 5 questions. The order matters because the three steps are frequently collapsed into one conversation, and the collapse hides errors.

The last row is worth pausing on. Step 4 is performed at contract inception under IFRS 15.76 and it does not care whether an obligation is satisfied over time or at a point in time. That independence is exactly why a Step 4 error is so damaging. The allocated amount attaches to an obligation, and that obligation then releases to revenue on its own timetable under Step 5. Move CU 1,000 from an obligation satisfied on delivery to one satisfied over three years and you have moved CU 1,000 of revenue from this year into the next three, without touching a single recognition judgement. The full five step sequence, including how Steps 2 and 3 feed this one, is set out in the complete guide to IFRS 15 revenue recognition.

Local FAQs

Can the contract price ever be the stand-alone selling price under IFRS 15? Yes. IFRS 15.77 says a contractually stated price "may be" the stand-alone selling price. What it removes is the presumption. Where an entity has a published price list, sells at that list price to the same customer class in the same market on a stand-alone basis with reasonable frequency, and the contract prices equal that list, the list price is observable evidence under IFRS 15.77 and the allocation will follow it. The point is that this is a conclusion supported by evidence, not a starting position.

Does Step 4 apply to a contract with a single performance obligation? No, subject to one qualification. IFRS 15.75 disapplies paragraphs 76 to 86 where there is only one performance obligation. The qualification is the second sentence: where the single obligation is a series under IFRS 15.22(b) and the consideration includes variable amounts, IFRS 15.84 to 15.86 continue to apply within the obligation.

Does the allocation change if a performance obligation is satisfied over time rather than at a point in time? No. IFRS 15.76 fixes the determination at contract inception and says nothing about the pattern of transfer. Timing is decided in Step 5. The allocated amount is the input to that timing decision, not an output of it.

Potential risks

The risk in this unit is a silent one. Because a stated-price allocation foots to the contract total and reverses over the contract life, it produces no reconciling difference, no unexplained balance and no failed control. It shows up only as revenue in the wrong period, which is precisely the assertion an auditor is required to test. Where the misallocation moves consideration between an obligation satisfied on delivery and one satisfied over several years, the current period effect can be material even though the cumulative effect is nil, and a cut-off or occurrence conclusion drawn from the contract schedule will not detect it.

2. What is a stand-alone selling price under IFRS 15, and what counts as directly observable evidence?

The three qualifiers in the "best evidence" sentence

IFRS 15.77 says the best evidence is "the observable price of a good or service when the entity sells that good or service separately in similar circumstances and to similar customers". Each qualifier disqualifies a category of price that finance teams routinely treat as observable.

The last row is where most of the argument happens. A list price is data. Whether it is observable evidence of the stand-alone selling price depends entirely on whether the entity transacts at it. Where the standard discount off list runs at thirty to forty per cent across the whole customer base, the list price is not the price at which the entity would sell separately, and using it as the stand-alone selling price will systematically over-allocate to whichever obligation carries the highest list price. That is not a theoretical concern. It is the mechanism by which perpetual software licences absorbed too much of a bundled contract for years before IFRS 15 forced the arithmetic into view.

Building an SSP that will survive review

The practical output of this unit is not a definition, it is a schedule. A defensible stand-alone selling price file has four columns and most files have one. The one they have is the allocated amount. The three that go missing are the SSP itself, the evidence supporting it, and the date it was set.

Most groups run this at portfolio level rather than contract level, and that is generally appropriate. IFRS 15.4 permits a portfolio approach where the entity reasonably expects the effects on the financial statements would not differ materially from applying the standard to individual contracts. In an SSP context that usually means setting a standard SSP, or a narrow SSP range, per product per customer tier per market, refreshing it periodically, and applying it to every contract signed in the period. The judgement then sits in two places: whether the population inside each band really is homogeneous, and how often the band is refreshed.

The contract asset that emerges from this pattern is one of the more visible balance sheet consequences of Step 4, and it is dealt with in the note on contract assets and contract liabilities. It is worth flagging here because it is the second-order effect that betrays whether an entity has actually applied the relative stand-alone selling price basis. An entity that allocates at contract prices in a subsidised handset model will report no contract asset at all, because billing and revenue will track each other exactly. The absence of the balance is the evidence.

Local FAQs

What is the difference between the stand-alone selling price and fair value? They are different measurement objectives and the difference is not cosmetic. Fair value under IFRS 13 is an exit price in an orderly transaction between market participants at the measurement date. The stand-alone selling price under Appendix A of IFRS 15 is an entity-specific entry price: the price at which this entity would sell the item separately. Two entities selling the same item can have different stand-alone selling prices and both can be right. IFRS 15.79(a) allows market and competitor prices to be used as an input, but requires them to be adjusted "to reflect the entity's costs and margins", which is exactly the entity-specific adjustment IFRS 13 would not permit.

Do we need a stand-alone selling price for every performance obligation? Conceptually yes, in a multi-obligation contract, because IFRS 15.76 allocates in proportion to those prices and a proportion needs a denominator. Practically, the residual approach in IFRS 15.79(c) and the combination approach in IFRS 15.80 exist precisely so that an entity can reach an answer where one or more items have no discernible price. Neither method removes the need for an SSP; both are ways of estimating one.

Can we use the price in a comparable customer's contract as the observable SSP? Only if that comparable contract was for the item sold separately. A price extracted from another bundled contract is a derived amount, not an observable price, because it was itself the product of an allocation. Using it circularly imports the other contract's allocation judgement into this one.

Potential risks

Two risks sit in this unit. The first is definitional drift: a group that starts with "SSP equals list price less standard discount" and gradually stops testing whether the standard discount is still standard. The second is stale evidence. Because IFRS 15.76 sets the SSP at inception and IFRS 15.88 forbids reallocation afterwards, an SSP error is locked into the contract for its whole life. Unlike a variable consideration estimate, which is reassessed each period under IFRS 15.59, there is no self-correcting mechanism. The only correction is a prior period restatement.

3. Worked example: what does allocating at stated prices instead of relative stand-alone selling price actually do to the numbers?

The facts

What follows is an IFRS 15 allocation example built to isolate one variable. Everything about the contract is held constant except the basis on which the transaction price is split, so the whole of the difference in reported revenue is attributable to Step 4.

All amounts are in currency units, CU. This is a constructed arithmetic example, not any company's figures.

On 1 January 20X1 an entity enters into a contract with a customer for the supply of a piece of equipment, its installation, and three years of technical support running from 1 January 20X1 to 31 December 20X3. The total consideration is CU 12,000, fixed, payable in full on signature. There is no financing component because payment and the first significant transfer are contemporaneous, so IFRS 15.60 does not bite. The three promises have been assessed under IFRS 15.27 and IFRS 15.29 and are three separate performance obligations: the equipment is transferred at a point in time on delivery, the installation is a routine service completed on 31 January 20X1 and satisfied at a point in time, and the support is satisfied over time under IFRS 15.35(a) and recognised on a straight-line basis under IFRS 15.39.

Note what the contract schedule is doing. It carries the full discount, and more, on the support line, taking a service with a CU 6,000 stand-alone selling price down to CU 2,100, while the equipment is billed CU 1,400 above its own stand-alone selling price. That is a rational commercial structure. It protects the headline equipment price for the next negotiation and it makes the support look cheap. It is also exactly the structure IFRS 15.77 was drafted to stop being used as the accounting answer.

The wrong answer: allocate at the stated prices

This is audit fault F-06. The entity takes the contract schedule as the allocation and never determines a stand-alone selling price at all.

The right answer: allocate on a relative stand-alone selling price basis

IFRS 15.76 requires the transaction price to be allocated in proportion to the stand-alone selling prices determined at contract inception. The weights are each obligation's SSP over the aggregate SSP of CU 15,000.

Because there is no observable evidence under IFRS 15.82 that the CU 3,000 discount belongs to any particular obligation, IFRS 15.81 requires it to be allocated proportionately to all three. That is not a separate calculation. As IFRS 15.81 puts it, the proportionate allocation of the discount "is a consequence of the entity allocating the transaction price to each performance obligation on the basis of the relative stand-alone selling prices of the underlying distinct goods or services". Each obligation takes a 20 per cent haircut on its stand-alone selling price, which is what the 0.8 multiplier does.

Look at the total row. The cumulative revenue is identical and the difference is nil. That is the reason this error is not caught by a revenue completeness procedure, a billing-to-revenue reconciliation, or a contract-level margin review. It is caught only by asking what each obligation would have sold for on its own, which is the question IFRS 15.77 asks and the contract schedule does not answer.

The journals, correct method

Consideration of CU 12,000 is received in full on 1 January 20X1. The equipment has a carrying amount in inventory of CU 3,500. Because the cash is received before any obligation is satisfied, a contract liability arises on receipt under IFRS 15.106.

The contract liability at 31 December 20X1 is CU 12,000 less CU 6,000 less CU 1,200 less CU 1,600, which is CU 3,200. That equals two remaining years of support at CU 1,600 each, so the balance proves itself. Under the stated-price allocation the closing contract liability would have been CU 12,000 less CU 10,600, which is CU 1,400, and that also proves itself against two years at CU 700. Both balances are internally consistent. Only one of them is right.

The correcting entry when the error is found

Suppose the misallocation is identified during the 20X1 audit, before the financial statements are issued. The adjustment is a reallocation between revenue and the contract liability. No cash, no receivable and no cost of sales is affected.

Local FAQs

Is this an error or a policy choice? It is an error. IFRS 15.74 uses "shall" and names only two exceptions, neither of which is "the contract says so". IFRS 15.77 removes the presumption that the contract price is the stand-alone selling price. There is no accounting policy election available at this point in the standard.

Does it matter if the contract prices are close to the stand-alone selling prices? Materiality applies as it does anywhere else. The point is that closeness has to be demonstrated, not assumed. Where a group has run the comparison, documented that stated prices sit within a tolerance of stand-alone selling prices, and applies the tolerance consistently, it has performed the IFRS 15.76 allocation and concluded it produces the same answer. That is a supportable position. Assuming closeness without measuring it is not.

Where does the difference show up on the balance sheet? In the contract liability, or in a contract asset where billing lags transfer. In the example the closing contract liability changes from CU 1,400 to CU 3,200. In a subsidised-handset model the direction reverses and a contract asset appears at activation. Either way, a Step 4 error is visible as a balance sheet difference as well as a revenue difference, which is often the more efficient place to test it.

Potential risks

The reporting risk is a systematic revenue timing misstatement that no total-based control detects. The disclosure risk is separate and is created by IFRS 15.126(c), which requires information about the methods, inputs and assumptions used in allocating the transaction price, including estimating stand-alone selling prices. An entity that allocates at contract prices has no method to disclose, so the note either says nothing or describes a method the entity has not applied. The second is the more serious of the two.

4. How do you estimate a stand-alone selling price under IFRS 15 when it is not directly observable?

The IFRS 15 adjusted market assessment approach

The expected cost plus a margin approach

A worked estimate using both methods

All amounts in CU. A software vendor's contracts include a bespoke integration service. The service is never sold separately, so the stand-alone selling price is not directly observable and must be estimated under IFRS 15.78. The entity runs both permitted estimation methods and compares them.

Check the arithmetic both ways. CU 61,200 multiplied by 0.70 is CU 42,840, which is the cost. CU 61,200 less CU 42,840 is CU 18,360, which is 30.0 per cent of CU 61,200. The margin is drawn from the entity's own separately sold professional services, not from its group gross margin, which is the IFRS 15.79(b) requirement for "an appropriate margin for that good or service".

The two methods produce CU 61,200 and CU 61,000, a spread of CU 200 on a CU 61,000 estimate, or 0.3 per cent. That convergence is the useful output. It is not required by the standard, which asks for one estimate rather than a corroborated one, but a second method run as a check is the cheapest possible piece of audit evidence and it turns a judgement into a triangulated judgement. Where the two methods diverge materially, the divergence is itself the finding, and it usually means either the cost model is missing a cost or the market comparators are not comparable.

Local FAQs

Can we use a single group margin in the expected cost plus a margin approach? Not without support. IFRS 15.79(b) asks for "an appropriate margin for that good or service". A blended margin applied to every obligation makes the relative stand-alone selling prices a pure function of relative cost, which will systematically under-allocate to high-margin, low-cost obligations such as software and over-allocate to low-margin, high-cost obligations such as hardware and field labour.

Do we have to use more than one method? No. IFRS 15.79 offers methods and IFRS 15.80 contemplates a combination where two or more items are highly variable or uncertain. A single method properly applied satisfies IFRS 15.78. Running a second method as corroboration is good practice rather than a requirement.

How often should estimated SSPs be refreshed? The standard sets no interval. What it sets is a measurement date, contract inception, in IFRS 15.76, and a consistency requirement in IFRS 15.78. The practical answer is that the refresh cycle has to be short enough that the SSP applied to a contract signed today is still an estimate of what the entity would sell that item for today. For most groups that is annual, with an out-of-cycle refresh on repricing or launch. The corroborating evidence for pricing decisions often sits in the same data used to set commission plans, which is why the analysis in the note on contract costs and sales commissions tends to draw on the same source systems.

Potential risks

The risk is a model that is never challenged because it is never presented. Estimated stand-alone selling prices frequently live in a pricing team spreadsheet rather than in the finance record, and they enter the accounts as a rate in an ERP revenue engine. When that happens the IFRS 15.78 requirements become untestable in practice: nobody can demonstrate that all reasonably available information was considered, that observable inputs were maximised, or that the method was applied consistently, because the file that would show it does not exist in finance. IFRS 15.126(c) then requires disclosure of methods, inputs and assumptions that the finance function cannot describe.

5. When is the residual approach available under IFRS 15, and what happens when the gate fails?

Testing the gate on the IFRS 15 residual approach, limb by limb

The two limbs are the entire eligibility test for the residual approach. IFRS 15 does not offer a materiality override, a size threshold or a "where practicable" relief here, so whether the IFRS 15 residual approach is available to a particular entity turns wholly on which limb it can evidence.

The two limbs are alternatives, so meeting either is enough. They are also mutually exclusive in practice. Limb (i) is about an item the entity sells constantly at wildly different prices. Limb (ii) is about an item the entity has never sold at all.

Limb (ii) has an "and" in it that is easy to miss. The entity must have "not yet established a price for that good or service" and the good or service must not "previously been sold on a stand-alone basis". An entity that has approved a price for a new product and published it, but has not yet made a stand-alone sale, fails limb (ii) because the first condition is not satisfied. It may still fail limb (i), because with no transactions there is no range to be broad. In that position the entity is out of the residual approach entirely and must estimate under IFRS 15.79(a) or 15.79(b).

Worked example: the same contract, three ways

All amounts in CU. A software vendor sells a perpetual licence, twelve months of post-contract support, and an onboarding training package. The support and training are directly observable: support is sold separately at CU 180,000 for twelve months and training at CU 60,000, both consistently and to the same customer tier. The question is the licence.

The residual is CU 1,000,000 less CU 180,000 less CU 60,000, which is CU 760,000. Note that the allocated column equals the SSP column exactly. That is not a coincidence and it is not a choice. The residual forces the sum of SSPs to equal the transaction price, so the relative allocation step becomes an identity. Every unit of discount or premium in the contract, whatever its commercial origin, attaches to the residual item.

Variant B changes only the reason the gate opens. Suppose the licence has been sold stand-alone forty times in the last nine months at prices ranging from CU 300,000 to CU 900,000, depending on the customer's negotiating position and deal size, with no discernible central tendency. Limb (i) is met: the entity sells the same good to different customers at or near the same time for a broad range of amounts, so a representative stand-alone selling price is not discernible from past transactions. The arithmetic is identical to Variant A. What differs is the evidence on file, which in Variant B has to be the price distribution itself.

The mechanics are worth spelling out. Aggregate SSP is CU 980,000 against a transaction price of CU 882,000, so the contract carries a CU 98,000 discount, exactly ten per cent. Relative allocation applies that ten per cent haircut to every obligation, which is what IFRS 15.81 requires in the absence of IFRS 15.82 evidence. The residual approach instead subtracts the two observable SSPs at full value and dumps the entire CU 98,000 discount on the licence, taking it to CU 642,000. Check: CU 882,000 less CU 180,000 less CU 60,000 is CU 642,000, and CU 740,000 less CU 642,000 is CU 98,000, the whole discount. That is the structural bias of the residual approach and it is why IFRS 15.79(c) gates it.

Proving the 20X1 figures. Correct: licence CU 666,000 plus training CU 54,000 plus six months of support at CU 162,000 times six twelfths, which is CU 81,000, giving CU 801,000. Residual: licence CU 642,000 plus training CU 60,000 plus CU 180,000 times six twelfths, which is CU 90,000, giving CU 792,000. The difference of CU 9,000 is the licence understatement of CU 24,000 less the training overstatement of CU 6,000 less the support overstatement of CU 9,000. Both columns total CU 882,000.

The contract liability at 31 December 20X1 is CU 216,000 less CU 54,000 less CU 81,000, which is CU 81,000, equal to six remaining months of support at CU 162,000 for twelve months. The balance proves. Where the licence is a right to use rather than a right to access, the point in time recognition follows the analysis in the note on licensing intellectual property, and that timing conclusion is what makes the Step 4 misallocation visible in the period rather than merely in the disclosure.

The IFRS 15.83 ordering rule

Local FAQs

Can the residual approach produce a stand-alone selling price of zero or a negative number? Arithmetically yes, where the observable SSPs of the other obligations exceed the transaction price. That outcome is a strong signal that the method is being misapplied, because a distinct good or service with a stand-alone selling price of nil is difficult to reconcile with the conclusion in Step 2 that it was distinct and had value to the customer on its own under IFRS 15.27(a). The correct response is to revisit whether the gate in IFRS 15.79(c) was properly met, and whether the observable prices subtracted really are observable, rather than to record a nil allocation.

Is the residual approach the same as the residual method under legacy software guidance? No, and treating them as equivalent is the historical source of most of the misuse. The legacy US approach permitted a residual as a general answer where vendor-specific objective evidence of fair value existed for undelivered elements. IFRS 15.79(c) permits it only through the two-limb gate and only where the amounts subtracted are observable stand-alone selling prices. The change in wording between IFRS 15 and ASC 606 in this area is minimal, so an entity reporting under both frameworks should reach the same answer, a point covered in the IFRS 15 against ASC 606 comparison.

Can the residual approach be used for more than one item in a contract? Not directly. A residual returns one number. Where two or more items have highly variable or uncertain stand-alone selling prices, IFRS 15.80 provides the route: use the residual to estimate the aggregate stand-alone selling price for that group, then use a second method to split the aggregate between the individual items, then perform the back-test in the last sentence of IFRS 15.80.

Potential risks

The residual approach carries a directional bias that makes it an attractive answer for the wrong reasons. It concentrates every discount on the item with no observable price, which in software, technology and equipment contracts is usually the licence or the hardware, and those are typically the obligations satisfied at a point in time at the start of the contract. So an unjustified residual usually defers revenue rather than accelerating it, which is the opposite of the usual incentive and is often why nobody challenges it. That does not make it right. IFRS 15.79(c) sets a condition, and an entity that cannot evidence the condition has used a method the standard did not make available to it.

6. How is a discount allocated under IFRS 15, and when can it go entirely to one performance obligation?

Worked example: the same contract, allocated both ways

All amounts in CU. An entity contracts to supply three distinct goods and services, A, B and C, for total fixed consideration of CU 120,000. A is a product transferred at a point in time on 1 February 20X1. B is a twelve month service delivered evenly through 20X1 and satisfied over time under IFRS 15.35(a). C is a product transferred at a point in time on 1 March 20X2.

Case 1: no IFRS 15.82 evidence. Proportionate allocation.

The entity sells A, B and C separately, so criterion (a) is met. It does not regularly sell any sub-bundle of them on a stand-alone basis at a discount, so criterion (b) fails. Criterion (c) cannot be tested because there is no bundle discount to compare. IFRS 15.81 therefore requires proportionate allocation, which the relative stand-alone selling price arithmetic delivers automatically.

Each obligation takes an eighty per cent multiplier, which is CU 120,000 over CU 150,000. Each therefore bears twenty per cent of its own stand-alone selling price as discount: CU 10,000, CU 12,000 and CU 8,000, summing to the CU 30,000 discount. The allocated column sums to CU 120,000.

Case 2: the IFRS 15.82 criteria are met.

Change one fact. The entity also regularly sells B and C together as a standing bundle, on a stand-alone basis, for CU 70,000 against aggregate stand-alone selling prices of CU 100,000. That is a bundle discount of CU 30,000. Test the criteria.

Two steps are running here and it is worth separating them. First, the CU 30,000 discount is assigned entirely to the B and C group under IFRS 15.82, leaving that group with CU 70,000. Second, that CU 70,000 is allocated between B and C on their relative stand-alone selling prices under IFRS 15.76, because IFRS 15.82 identifies the group the discount belongs to and not the split within it. B takes sixty per cent and C takes forty per cent, giving CU 42,000 and CU 28,000. Adding A at its full CU 50,000 gives CU 120,000.

Proving 20X1: Case 1 is A at CU 40,000 plus B at CU 48,000, which is CU 88,000. Case 2 is A at CU 50,000 plus B at CU 42,000, which is CU 92,000. The CU 4,000 swing comes from A carrying CU 10,000 of discount in Case 1 and none in Case 2, offset by B carrying CU 12,000 in Case 1 and CU 18,000 in Case 2. Both columns total CU 120,000, and the timing difference sits entirely in whether C, satisfied in 20X2, absorbs CU 8,000 or CU 12,000 of the discount.

The contract liability closes 20X1 at CU 70,000 less CU 42,000, which is CU 28,000, being the amount allocated to C and released in March 20X2. Under Case 1 the same balance would have been CU 32,000. The difference of CU 4,000 is the same CU 4,000 in the revenue table, appearing on the other side of the entry. This is the general rule for Step 4 errors and it is the reason a contract balance walk is often a faster test than a revenue walk: an allocation difference always has a balance sheet twin. That relationship is set out more fully in the note on contract assets and contract liabilities.

Local FAQs

Can a discount be allocated to all but one obligation? Yes. IFRS 15.82 says "one or more, but not all", which permits any subset short of the whole. The worked example above allocates to two of three. What it does not permit is allocating the entire discount to all obligations in some non-proportionate ratio, because that is neither the IFRS 15.81 proportionate outcome nor the IFRS 15.82 exception.

What does "substantially the same" mean in IFRS 15.82(c)? The standard does not quantify it and no interpretation has. It is a judgement about whether the observed bundle discount explains the contract discount. Where the contract discount is CU 30,000 and the standing bundle discount is CU 30,000, the answer is clear. Where the contract discount is CU 45,000 and the bundle discount is CU 30,000, the analysis does not explain CU 15,000 of it, and the second limb of criterion (c) fails because the evidence no longer identifies where the whole discount belongs.

Does a customer option or loyalty credit change the discount analysis? It changes the population before the discount question is reached. Where an option confers a material right under IFRS 15.B40, it is itself a performance obligation and takes an allocation, and IFRS 15.B42 explains how to estimate its stand-alone selling price where it is not directly observable. That is a Step 2 and Step 4 interaction rather than a discount question, and it is worked through in the note on warranties, returns and customer options.

Potential risks

The exposure runs both ways and both are live. Applying IFRS 15.82 without the evidence produces an allocation that moves revenue between periods and is unsupported. Failing to apply IFRS 15.82 where the evidence does exist is equally a departure, because the paragraph says "shall". The second failure is much less often challenged and is common in groups with standing bundle offers, where finance runs a proportionate allocation for simplicity while the commercial data on file would compel a different answer. IFRS 15.126(c) then requires disclosure of the basis on which discounts have been allocated to a specific part of the contract, which puts the question in the notes whether or not the file has answered it.

7. How is variable consideration allocated, and when does it go entirely to one performance obligation?

Worked example: a fixed licence and a bonus-bearing service

All amounts in CU. On 1 January 20X1 an entity contracts to grant a right to use licence, transferred and controlled by the customer on that date, and to provide a managed service for two years to 31 December 20X2. The consideration is CU 500,000 fixed, plus a performance bonus of CU 30,000 for each of the two years in which platform availability exceeds 99.9 per cent. Maximum consideration is therefore CU 560,000. Stand-alone selling prices are directly observable: the licence at CU 200,000 and the two year managed service at CU 300,000.

At inception the entity estimates the variable consideration using the most likely amount method under IFRS 15.53(b) and then applies the constraint in IFRS 15.56. Its availability record supports one of the two bonus years but not both, and it concludes that including more than CU 30,000 would risk a significant reversal. The transaction price at inception is therefore CU 530,000. The estimation and constraint mechanics sit in the note on variable consideration and the constraint; what follows is the allocation.

The wrong answer: spreading the bonus on relative SSP

Suppose the entity ignores IFRS 15.85 and allocates the whole CU 530,000 on relative stand-alone selling prices, forty per cent to the licence and sixty per cent to the service. The licence takes CU 212,000 and the service takes CU 318,000.

Proving the correct column. In 20X1 the licence gives CU 200,000 and the service gives CU 330,000 divided by two, which is CU 165,000, for CU 365,000. During 20X2 availability is achieved in both years, the constraint no longer applies, and the transaction price rises by CU 30,000 to CU 560,000. Under IFRS 15.89 that change is allocated entirely to the service because the IFRS 15.85 criteria are met, so cumulative service consideration becomes CU 360,000 and 20X2 service revenue is CU 360,000 less the CU 165,000 already recognised, which is CU 195,000. Total CU 560,000.

Proving the incorrect column. In 20X1 the licence gives CU 212,000 and the service gives CU 318,000 divided by two, which is CU 159,000, for CU 371,000. On reassessment the transaction price of CU 560,000 is spread again at forty and sixty per cent, giving the licence CU 224,000 and the service CU 336,000. The licence therefore picks up a catch-up of CU 12,000 in 20X2, and the service records CU 336,000 less CU 159,000, which is CU 177,000. Total 20X2 is CU 189,000, and the two years sum to CU 560,000.

Proving the contract asset. It opens at nil, takes CU 115,000 at the end of 20X1, is reduced by CU 250,000 on the second invoice to negative CU 135,000, and is brought back to nil by CU 135,000 at the end of 20X2. Total revenue is CU 200,000 plus CU 165,000 plus CU 195,000, which is CU 560,000, and total amounts invoiced are CU 250,000 plus CU 250,000 plus CU 60,000, which is CU 560,000. The contract closes flat. The negative contract asset position during 20X2 is a contract liability in presentation terms and would be shown as such under IFRS 15.105.

Local FAQs

Is applying IFRS 15.85 optional? No. The paragraph reads "shall allocate ... entirely" where both criteria are met. It is not an election, and an entity that spreads a qualifying bonus proportionately has departed from the standard just as surely as one that allocates a non-qualifying bonus entirely.

Does the constraint in IFRS 15.56 apply before or after the allocation? Before. The constraint is a Step 3 measurement question that determines how much variable consideration enters the transaction price at all. Step 4 then allocates whatever survived. Running them in the other order produces a constrained allocation of an unconstrained price, which is not a calculation the standard describes.

How does a series under IFRS 15.22(b) interact with this? Directly, and it is the reason IFRS 15.75 preserves paragraphs 84 to 86 for single obligation contracts. Where the series is one performance obligation and the consideration for a particular period is variable, IFRS 15.85 allocates that variable amount to the distinct goods or services in that period rather than across the whole obligation. A usage-based fee in a hosted arrangement is the everyday case, and the analysis is developed in the note on SaaS and subscription revenue.

Potential risks

The recurring exposure is a contract where the variable element is large relative to the fixed element and the IFRS 15.85 test has never been documented. Usage fees, royalties, service credits, gain shares and volume rebates all fall here. The default in most ledgers is to recognise the variable amount against whatever was invoiced, which mimics the IFRS 15.85 answer often enough that nobody notices when it does not. It fails badly when a rebate is calculated on total contract spend but recognised against the product line that triggered it, or when a milestone bonus is credited to whichever obligation the billing system associates with the milestone code rather than the one whose efforts earned it.

8. What happens to the allocation when the transaction price changes after contract inception?

Worked example: a price change with obligations partly satisfied

All amounts in CU. Take the contract from unit 3. Transaction price CU 12,000 at inception, allocated on relative stand-alone selling prices at 50 per cent, 10 per cent and 40 per cent, giving equipment CU 6,000, installation CU 1,200 and support CU 4,800. Equipment and installation were satisfied in 20X1. The support is one year through a three year term at 31 December 20X1, with CU 1,600 recognised.

On 1 January 20X2 an uncertainty in the contract resolves and the consideration the entity expects to be entitled to increases by CU 1,500 to CU 13,500. This is not a modification: no promise has changed and no scope has been added, so IFRS 15.90 does not redirect the analysis. Suppose also that by this date the observable stand-alone selling price of three year support has risen from CU 6,000 to CU 7,500. IFRS 15.88 requires that movement to be ignored.

Note the weights. They are the inception weights of 50, 10 and 40 per cent, not weights recalculated from the current stand-alone selling prices. Had the entity reallocated on current SSPs of CU 7,500, CU 1,500 and CU 7,500, aggregate CU 16,500, the support would have taken 45.5 per cent of the increase rather than 40 per cent. IFRS 15.88 prohibits that recalculation in its second sentence.

The total column foots to CU 13,500, which is the original CU 12,000 plus the CU 1,500 increase. The obligation columns foot to 50, 10 and 40 per cent of CU 13,500, which is CU 6,750, CU 1,350 and CU 5,400. The inception ratio is preserved across the whole contract, which is exactly what "on the same basis as at contract inception" means.

Total support revenue in 20X2 is the CU 200 catch-up plus the CU 1,800 for the year, which is CU 2,000, agreeing to Table 31. Had the price change been a decrease of CU 1,500 rather than an increase, every entry reverses: CU 750 and CU 150 would be debited to revenue in 20X2 as reductions of revenue for the satisfied obligations, which IFRS 15.88 contemplates expressly in its third sentence.

Where a modification is accounted for as a separate contract under IFRS 15.20, the two contracts are simply separate and each carries its own allocation, which is why IFRS 15.90(b) is expressed as applying where the modification "was not accounted for as a separate contract". The full sequencing of IFRS 15.18 to 15.21, including the stand-alone selling price condition in IFRS 15.20(b), is worked through in the note on contract modifications. It is worth noting there that IFRS 15.20(b) is itself a stand-alone selling price test: a modification is a separate contract only if the price increase reflects the entity's stand-alone selling prices of the additional goods or services. So Step 4 thinking decides whether the modification rules or the allocation rules apply in the first place.

Local FAQs

If a stand-alone selling price changes during a long contract, do we reallocate? No. IFRS 15.88 prohibits it in terms. The inception allocation ratio applies for the life of the contract, and it applies to later changes in the transaction price as well.

Is the catch-up on a partly satisfied obligation a prior period adjustment? No. IFRS 15.88 requires amounts allocated to a satisfied obligation to be recognised "in the period in which the transaction price changes". It is a current period item. A prior period adjustment arises only where the original estimate was wrong at the time it was made, which is an IAS 8 error rather than an IFRS 15.87 change in circumstances.

What if the price change makes the allocated amount for a satisfied obligation negative? A large downward change can exceed the revenue previously recognised on an obligation. IFRS 15.88 contemplates recognition "as a reduction of revenue", and a reduction that exceeds cumulative revenue for that obligation results in a net debit to revenue for the period. That is an unusual outcome and it usually signals that the original transaction price failed the constraint in IFRS 15.56 rather than that the allocation is wrong.

Potential risks

The main risk is systems rather than judgement. Frozen allocation weights are an unglamorous control that almost every revenue engine can enforce and many are not configured to. The secondary risk is classification: treating a modification as a price change, or a price change as a modification. The two produce genuinely different numbers, because a modification under IFRS 15.21(a) re-strikes the allocation on the remaining obligations at their current stand-alone selling prices while a price change under IFRS 15.88 preserves the original ratio. Getting the gate wrong changes the answer even when the cash is identical.

9. How is the transaction price allocated to a customer option, and how is the option's stand-alone selling price estimated?

Worked example: allocating to a material right

All amounts in CU. An entity sells a product for CU 1,000, which is its directly observable stand-alone selling price. As part of the same contract the customer receives a voucher for 35 per cent off any purchase made in the following twelve months. The entity routinely offers customers of this class a 10 per cent discount, so the incremental discount is 25 percentage points. Based on redemption history the entity expects the voucher to be used on a purchase of approximately CU 1,250, and expects 80 per cent of customers to exercise.

Check: CU 1,250 aggregate SSP against a CU 1,000 transaction price is a CU 250 discount, or twenty per cent. Both obligations take a twenty per cent haircut, so CU 800 and CU 200, which sum to CU 1,000. This is IFRS 15.81 proportionate allocation operating exactly as unit 6 describes, and it is worth seeing that the option is not treated as a special case for the discount. It is a performance obligation and it takes its share.

The arithmetic on the second sale proves: CU 1,250 less 35 per cent is CU 812.50 of cash, plus the CU 200 released from the contract liability, giving CU 1,012.50 of revenue. Across both transactions the entity has recognised CU 800 plus CU 1,012.50, which is CU 1,812.50, against total cash of CU 1,000 plus CU 812.50, which is CU 1,812.50. The contract liability opens at nil, rises to CU 200 and returns to nil.

Local FAQs

Is a free trial a material right? It depends on whether it is granted in a contract. IFRS 15.B40 applies to an option granted "in a contract" to acquire additional goods or services. A free trial offered to the market at large before any contract exists is a marketing offer. A free extension granted inside a signed contract, on terms not available to that customer class generally, is capable of being a material right and needs the B40 test.

How do renewal options in subscription contracts work? Two routes are open. Estimate the option's stand-alone selling price under IFRS 15.B42, or use the IFRS 15.B43 practical alternative where the renewal is of similar services on the original terms. The choice materially affects the pattern of revenue in subscription businesses and interacts with the contract term assessment, which is developed in the note on SaaS and subscription revenue.

What happens to the contract liability if the option is never exercised? IFRS 15.B40 recognises the revenue "when the option expires". Where an entity has many similar options and expects a proportion never to be exercised, the breakage question interacts with IFRS 15.B46 on unexercised rights, and the exercise likelihood already sits inside the IFRS 15.B42(b) estimate, so care is needed not to reflect the same expectation twice.

Potential risks

The estimation inputs in IFRS 15.B42 live outside finance. Expected purchase value, standing discount ranges and redemption rates come from commercial, marketing and CRM systems that are not usually in scope for financial reporting controls. Where the material right balance is significant, the audit exposure is a data reliability exposure rather than a technical one, and the response is to bring those populations into the control environment rather than to argue about the paragraph. Options also sit at the boundary with the returns and warranty guidance, and the full set of adjacent balances is covered in the note on warranties, returns and customer options.

10. What does IFRS 15 require an entity to disclose about the allocation, and what should the file contain?

What a Step 4 note should actually say

The gap between a compliant note and a common one is narrow in length and wide in content. Both run to a paragraph. One of them tells a reader what happened.

The reason this matters commercially, rather than only technically, is that Step 4 judgements move revenue between periods without moving cash. A reader who can see the method can form a view on how sensitive reported revenue is to it. A reader who cannot, cannot. That is the point the Financial Reporting Council made in its review of first-year IFRS 15 disclosures.

Both the Financial Reporting Council and the European Securities and Markets Authority have commented on this specific disclosure, and their wording is set out in the regulatory guidance section below. The short version is that they ask for more than the literal text of IFRS 15.126(c): not only which method was used, but why that method is suitable, and where a discount has been directed at particular obligations, the reason.

The governance around a stand-alone selling price

Because IFRS 15.76 fixes the stand-alone selling price at contract inception and IFRS 15.88 forbids reallocation afterwards, an SSP is a number that has to be right at the moment a contract is signed. That is an unusual demand. Most accounting estimates are made at a reporting date by a finance team with time to think. An SSP is applied at the point of sale, in volume, usually by a system.

Row four is the one that most often does not exist and is the cheapest to build. Row five is the one that pays for itself fastest in an audit, because it turns a population of thousands of contracts into a shortlist. Where the allocated amount and the contract line price are within a tolerance, Step 4 has changed nothing and the file needs only the evidence that the tolerance was tested. Where they diverge sharply, the contract is worth a look. That is a straightforward analytical procedure and it is the practical answer to how a reviewer covers a large contract population without reading every schedule.

Local FAQs

Do we have to disclose our actual stand-alone selling prices? No. IFRS 15.126(c) asks for the methods, inputs and assumptions, not the price table. An entity can describe the method used for each category of obligation, the inputs it draws on and the key assumptions without publishing commercially sensitive prices. What it cannot do is describe none of them.

Is the allocation a significant judgement requiring disclosure under IAS 1? Frequently yes, and it is a separate requirement from IFRS 15.126. Where the allocation basis is a source of estimation uncertainty with a significant risk of material adjustment in the next year, IAS 1.125 applies in its own right, and IFRS 18 carries the equivalent requirement forward for periods beginning on or after 1 January 2027. In practice the IFRS 15.126(c) disclosure and the significant judgement disclosure are usually best given together.

How much of this applies to an entity with mostly single-obligation contracts? Very little, and that is worth establishing early. IFRS 15.75 disapplies paragraphs 76 to 86 where a contract has one performance obligation, subject to the series point. An entity whose contracts are genuinely single-obligation has no allocation to disclose under IFRS 15.126(c) beyond saying so. What it does need is the Step 2 analysis showing that conclusion, which is dealt with in the note on identifying performance obligations and in the complete guide to IFRS 15 revenue recognition.

Potential risks

The disclosure risk and the file risk are the same risk arriving at different times. An entity without a documented allocation method will write a generic note, and the note will pass until a regulator or an auditor asks what the method is. At that point the answer has to be constructed retrospectively for contracts whose stand-alone selling prices were required to be determined at inception under IFRS 15.76, which is not a gap that can be closed after the fact. The second risk is a stale note: a description written at transition in 2018 that no longer matches a business whose product mix, pricing model or customer tiering has since changed.