Every pharma licensing deal I have worked on contains variable consideration. Sometimes it is explicit — a Phase II completion milestone, a first commercial sale payment, a tiered royalty. Sometimes it is buried inside a headline number that a business development team has described as "committed." It rarely is.
IFRS 15 requires you to include variable consideration in the transaction price only to the extent that it is highly probable a significant reversal will not occur when the uncertainty is subsequently resolved. That sentence sounds straightforward until you are sitting in front of a drug candidate with a 15% probability of reaching approval and a €200 million regulatory milestone attached to it.
This article sets out the practical steps for estimating and constraining variable consideration in milestone-based deals, the judgements that create the most risk, and the disclosure obligations that follow once you have made those judgements.
What counts as variable consideration in a pharma deal
Variable consideration under IFRS 15 is broader than most finance teams initially assume. It covers any amount that is contingent on a future event or outcome. In a typical pharma licensing transaction, the following forms arise routinely.
Development milestones are payments triggered by clinical progression — IND filing, Phase I completion, Phase II data readout, Phase III enrolment, regulatory submission, approval. Each is contingent on a binary event that may or may not occur, often years into the future.
Commercial milestones are triggered by sales thresholds — first commercial sale, cumulative net sales exceeding a defined level in a defined territory. These depend on both regulatory success and commercial performance, compounding the uncertainty.
Royalties are a usage-based form of variable consideration, though IFRS 15 applies a specific exception for royalties on licences of intellectual property, which I will address separately.
Price adjustments — net sales adjustments, government pricing rebates, co-promotion payments — are common in deals where the licensed product enters managed markets with significant payer pressure.
Practical note: Upfront licence fees are not automatically fixed consideration. If any portion of an upfront fee is refundable on a future event — a regulatory failure, a change of control, a co-development withdrawal — that portion is variable. Always read the full contract, not the deal summary.
The two estimation methods
IFRS 15 permits two methods for estimating variable consideration, and the standard is clear that you should use whichever better predicts the amount to which you will be entitled.
Expected value is the probability-weighted sum of possible outcomes. It is appropriate when a range of outcomes is possible and you have a reasonable basis for assigning probabilities — for example, a royalty stream where net sales could be €50m, €150m, or €300m depending on market penetration, and your commercial team has modelled all three scenarios with supportable assumptions.
Most likely amount is the single most probable outcome from the range. It is appropriate when the contract has essentially binary outcomes — either the milestone is achieved or it is not. Most likely amount is almost always the right method for clinical development milestones.
A biotech licenses a Phase I oncology asset to a large pharma group. The deal includes a €150 million milestone payable on Phase III initiation. The biotech's internal modelling assigns a 40% probability of reaching Phase III within the contract term.
Using expected value: 40% × €150m = €60m estimate. But IFRS 15 does not ask what the expected value is — it asks whether including any amount in the transaction price would result in a significant revenue reversal. Even at 40% probability, including €60m now and reversing it if Phase III does not start would be a significant reversal. The constraint applies.
Using most likely amount: the most likely outcome is binary — €150m is either received or it is not. At 40% probability of success, zero is the most likely single outcome. The milestone is constrained to nil until the probability profile changes materially.
A five-step process for finance teams
The concepts above need to translate into a repeatable process that finance teams can apply consistently across a portfolio of deals. The following five steps cover the full lifecycle of variable consideration from contract inception through to each reporting date.
| Step | What to do | Pharma-specific note |
|---|---|---|
| 1. Classify each payment | Read the full contract — not the deal summary — and identify every payment contingent on a future event. Separate fixed from variable. Flag any upfront fee with a refund mechanism. | Development milestones, commercial milestones, royalties, and price adjustments each have different accounting treatments. Do not group them. |
| 2. Assign probabilities | For each variable payment, document the basis for your probability assessment. Use internal pipeline data, published industry phase transition rates, and input from R&D and regulatory affairs. | For novel mechanisms or first-in-class assets, published industry benchmarks provide a starting point but must be adjusted for asset-specific factors. Document all adjustments. |
| 3. Select the estimation method | Apply expected value where a range of outcomes is possible and supportable. Apply most likely amount where the outcome is binary. Most clinical milestones are binary. | Do not use expected value simply because it produces a higher recognised amount. The method must genuinely better predict the amount to which you will be entitled. |
| 4. Apply the constraint | Test whether including any estimated amount would create a highly probable significant reversal. Where it would, constrain to nil or to the amount that passes the test. Document the reasoning. | At contract signing, nearly all clinical development milestones will be constrained. The constraint analysis is not a one-time exercise — it is re-run at every reporting date. |
| 5. Reassess each period | Update your probability and constraint assessment at every period end. Trigger events — interim data readouts, regulatory feedback, safety signals — must feed into the reassessment promptly. | Build a standing process with R&D and regulatory affairs to surface trigger events in time for period-end close. A pipeline review the week before close is not sufficient for material balances. |
The constraint test in practice
The constraint is not a bright-line probability threshold. IFRS 15 does not say "include consideration above 50% probability" or "exclude below 75%." It requires judgement, documented with reference to the factors the standard identifies as increasing the likelihood of a significant reversal.
| Factor | What it means for a clinical milestone |
|---|---|
| Amount highly susceptible to factors outside the entity's influence | Regulatory outcomes, patient recruitment, safety signals — none of which the licensor controls once the asset is out-licensed |
| Uncertainty not expected to resolve for a long period | Phase III programmes routinely run three to seven years; approval timelines add further uncertainty |
| Limited experience with similar contracts | First-in-class assets in novel mechanisms have no direct historical comparator |
| Contract has a large number and broad range of possible amounts | Multi-territory deals with territory-specific milestones create compounding uncertainty |
| Practice of offering concessions or changing terms | Renegotiation on programme failure is commercially common and creates precedent |
In practice, almost every clinical development milestone will be constrained at the point of deal signing. The question is not whether to apply the constraint but when the probability profile has improved sufficiently to include some or all of the milestone in the transaction price.
When to release the constraint
Reassessment is required at each reporting date. The constraint does not mean the milestone stays at nil permanently — it means you re-examine the evidence at each period end and update your estimate.
The evidence base for releasing a constraint on a Phase III milestone typically includes: completion of Phase II with clean data, a clear regulatory pathway agreed with the relevant authority, a fully funded development plan in place with the licensee, and absence of material safety signals. When the probability of a significant reversal has become remote rather than merely possible, the milestone can be included in the transaction price — recognised as revenue if the associated performance obligation has been satisfied, or as a contract liability if it has not.
Where teams get this wrong: Finance functions often release the constraint at the point the milestone becomes contractually due — when the triggering event occurs — rather than progressively as the evidence base strengthens. This produces a lumpy, event-driven revenue recognition pattern that does not reflect the economics of the arrangement and creates restatement risk.
Common pitfalls and how to avoid them
Over-optimism in probability assessment. Commercial teams have an incentive to present the pipeline positively. When probability estimates are based on press releases, analyst consensus, or verbal assurances from R&D rather than documented clinical data, they will not withstand audit scrutiny. The fix is to involve biostatisticians and regulatory affairs in the formal probability assessment, and to retain the supporting documentation in the deal file alongside the accounting memo.
Premature constraint release. Releasing the constraint only when the milestone triggers — rather than as the probability profile improves — creates recognition events that are too large and too late. Where the evidence base genuinely supports partial inclusion before the triggering event, that is both permitted and preferable to a cliff-edge recognition.
Siloed teams. Finance cannot assess clinical probability without R&D input, and R&D rarely understands what the accounting team needs or when it needs it. The result is a period-end scramble that produces poorly documented estimates. A cross-functional revenue working group — meeting monthly, with a standing agenda item for pipeline events affecting revenue recognition — eliminates this structural gap.
Documentation gaps. Auditors will ask for the contemporaneous basis for every constraint decision. "Management judgement" is not a basis. The deal file should contain the probability assessment with its sources, the constraint test with its reasoning, the estimation method selection and why it was chosen, and the reassessment at each period end.
The royalty exception
Royalties on licences of intellectual property sit outside the general variable consideration guidance. Under IFRS 15.B63, a sales-based or usage-based royalty on a licence of IP is recognised only when the subsequent sale or usage occurs — not when it becomes highly probable. This is an exception to the constraint guidance, not an application of it.
The practical implication is that royalty streams from pharma licences are recognised as the licensee sells the product, regardless of how probable those future sales might appear at contract inception. No estimate of future royalties is included in the transaction price at the outset.
The exception applies only where the royalty relates solely to a licence of IP or where the licence is the predominant element of the arrangement. Where a royalty is partly attributable to a service — an ongoing co-promotion, a supply arrangement, a manufacturing licence — the exception may not apply in full, and careful allocation across performance obligations is required.
Disclosure obligations
Variable consideration creates substantial disclosure obligations under IFRS 15.120–122, and these are an area where auditors have historically found significant gaps in pharma entities' financial statements.
You are required to disclose the aggregate amount of transaction price allocated to remaining performance obligations, including an explanation of when you expect to recognise it as revenue. For constrained variable consideration, this means explaining not just the amount constrained but the nature of the uncertainty, the methods used, and why the amount constrained is consistent with the constraint guidance.
Qualitative disclosure is not sufficient on its own for material balances. If a €500 million development milestone is constrained to nil, that fact — and the basis for that conclusion — should be visible in the notes. Investors and counterparties reading the accounts deserve to understand the gap between the contractual entitlement and what has been included in the transaction price.
What this means at the deal stage
The accounting analysis should not begin at the point of signing. Finance teams brought into deal structuring early can influence whether milestones are designed in a way that aligns commercial intent with accounting outcome. A single €200 million regulatory milestone may produce a different accounting result from four €50 million milestones spread across the regulatory pathway — not because the economics differ, but because the probability assessment at each trigger point differs.
Similarly, the distinction between a milestone that triggers on initiation of a study versus completion versus receipt of data versus regulatory acceptance is not merely legal drafting. Each trigger point has a different probability profile at the time the contract is signed, and that profile drives the constraint analysis from day one.
Getting variable consideration right in a milestone-based deal requires the accounting team to understand the science, the regulatory pathway, and the commercial structure — not just the contract schedule. That is precisely the kind of cross-functional involvement that reduces restatement risk and produces financial statements that reflect what the deal actually means.