Key Takeaways
- Research spend is always expensed; IAS 38.54 permits no exceptions regardless of the project's eventual success.
- Capitalisation of development costs begins only on the date all six IAS 38.57 criteria are met simultaneously.
- Entities in pharma, software, automotive, and engineering sectors commonly spend 8%–15% of revenue on R&D, making the phase boundary a material judgment.
- If the entity cannot separate research from development, the entire amount is expensed under IAS 38.53.
What is IAS 38 Research Costs vs Development Costs?
| Dimension | Research costs | Development costs |
|---|---|---|
| Accounting treatment | Expensed in profit or loss as incurred (IAS 38.54) | Capitalised when all six IAS 38.57 criteria are satisfied |
| Typical activities | Searching for alternatives, evaluating formulations, laboratory studies, original investigation | Prototyping, pilot production, designing tooling, testing pre-production samples |
| Recognition test | None; expense is mandatory regardless of management's intent or the project's prospects | Six cumulative criteria: technical feasibility, intention to complete, ability to use or sell, probable future economic benefit, adequate resources, reliable cost measurement |
| Reclassification | Prohibited; IAS 38.71 bars reclassifying research costs as an asset even if the project later succeeds | Not applicable; capitalisation starts from the date the criteria are first met, not retrospectively |
| Phase boundary trigger | Ends when the entity demonstrates technical feasibility | Begins on the date technical feasibility is demonstrated and the remaining five criteria are also met |
| Common documentation gap | No contemporaneous project log separating research-phase activities from development-phase activities | No dated record of when all six criteria were first satisfied |
Decision rule: If the entity cannot demonstrate technical feasibility, expense the cost. If all six IAS 38.57 criteria are met and the entity can document when that threshold was crossed, capitalise from that date forward.
When the distinction matters on an engagement
The entire research-versus-development question hinges on one date: the day the entity first satisfies all six IAS 38.57 criteria. Move that date earlier by two months on a project burning EUR 150,000 per month and the entity capitalises an extra EUR 300,000, lifting both total assets and current-year profit by the same amount.
IAS 38.53 adds a failsafe. When a project blends research and development activities so tightly that the entity cannot distinguish between them, the standard treats all expenditure as research phase (and therefore expenses it). Auditors encounter this most often in software companies running agile sprints, where requirements gathering, coding, and testing overlap within the same iteration cycle. Under ISA 540.13(a), the auditor evaluates whether management's method for determining the transition point is appropriate. A board resolution approving commercialisation signals intention (criterion b) but does not by itself demonstrate technical feasibility (criterion a) or the ability to measure costs reliably (criterion f). Treating the board date as the capitalisation start without separate evidence for each criterion leaves the file exposed.
Worked example: Schaefer Elektrotechnik AG
Client: German electronics manufacturer, FY2025, revenue EUR 310M, IFRS reporter. Schaefer is developing a next-generation power inverter for industrial solar installations. Total project spend in FY2025 is EUR 4.6M.
Step 1 — Map project activities to IAS 38 phases
From January to July 2025, Schaefer's R&D team evaluated four semiconductor architectures and built disposable bench prototypes to compare efficiency curves. These activities fall under IAS 38.56(a) and (b). In August 2025, the team selected a gallium-nitride architecture after bench tests confirmed 97.8% conversion efficiency at rated load.
Step 2 — Test the six IAS 38.57 criteria at 1 August 2025
(a) Technical feasibility: bench tests confirm the selected architecture meets the product specification. (b) Intention to complete: the management board approved the development budget on 12 August 2025. (c) Ability to sell: Schaefer has framework agreements with two large EPC contractors covering the target market. (d) Probable future economic benefit: management's five-year forecast projects cumulative sales of EUR 48M. (e) Adequate resources: approved budget of EUR 6.2M funded from operating cash flow and an undrawn revolving facility. (f) Reliable cost measurement: the project uses a dedicated SAP cost centre with time-sheet allocation by engineer.
Step 3 — Split the FY2025 spend
January to July (research phase): EUR 2.85M expensed to profit or loss. August to December (development phase): EUR 1.75M capitalised as an internally generated intangible asset.
Step 4 — Subsequent measurement
The capitalised EUR 1.75M is not yet amortised because the inverter is still in development (product launch targeted for Q3 2026). IAS 38.97 requires an annual impairment test for intangible assets not yet available for use.
Conclusion: of the EUR 4.6M total R&D spend, EUR 2.85M is expensed and EUR 1.75M is capitalised. If the auditor accepted management's claim that technical feasibility was established in May (when early simulation results looked promising but bench testing had not yet been completed), an additional EUR 750,000 would have been capitalised, overstating the intangible asset and operating profit by that amount.
Why it matters in practice
Entities frequently treat the board approval date as the capitalisation trigger without verifying that all six IAS 38.57 criteria were met at that date. A board resolution demonstrates intention (criterion b) but provides no evidence of technical feasibility (criterion a) or reliable cost measurement (criterion f). ISA 500.9 requires the auditor to evaluate whether the evidence obtained supports the capitalisation start date, not merely the existence of a management decision.
When agile or iterative development methods blur the boundary between research and development, teams sometimes capitalise from an arbitrary "sprint zero" milestone rather than identifying the date each criterion was demonstrably satisfied. IAS 38.53 is explicit: if the entity cannot distinguish research from development, the entire spend is expensed.
Related terms
Frequently asked questions
Can development costs be capitalised under Dutch GAAP (RJ)?
RJ 210.223 permits but does not require capitalisation of development costs that satisfy criteria similar to IAS 38.57. Under Dutch GAAP, the entity has an accounting policy choice. If it elects to capitalise, BW2 Article 365.2 requires a legal reserve equal to the capitalised amount, restricting distributable profits by the same figure.
What happens if a capitalised development project is abandoned?
The entity writes the carrying amount down to zero through profit or loss as an impairment charge. IAS 38.111 and IAS 36.12(b) require the loss to be recognised in the period the abandonment decision is made. The auditor verifies the write-down is complete and that no residual balance remains on the statement of financial position.
How do I audit the transition date from research to development?
Obtain the entity's internal project records (stage-gate reports, trial results, feasibility assessments, board minutes) and map each of the six IAS 38.57 criteria to dated evidence. ISA 540.18 requires the auditor to evaluate whether management's assumptions, including the claimed transition date, are reasonable. If the entity cannot produce contemporaneous evidence for each criterion, IAS 38.53 requires the entire spend to be expensed.