Why do IFRS companies look more profitable when their R&D gets capitalised?
The same dollar spent on research and development tells two different profit stories depending on which accounting framework a company follows. Under US GAAP, research and development costs are almost always expensed immediately, flowing straight to the income statement as a cost that reduces earnings. Under IFRS, the development phase of R&D can be capitalised to the balance sheet as an intangible asset, expensed over its useful life, and therefore deferred from profit. The gap between these approaches is not small, and it is one of the most durable red flags when comparing a US-listed company to an international competitor or when evaluating a non-US corporation that reports under IFRS.
This article explains the mechanics of R&D capitalisation under IFRS, why it matters for profit comparison, and how to spot when a company uses this rule as an earnings lever.
Quick definition
Under IFRS (specifically IAS 38, Intangible Assets), costs incurred in the research phase of a project must be expensed as incurred. But costs in the development phase—once technical feasibility is demonstrated and the company intends to complete and use the asset—can be capitalised as an intangible asset and amortised over its useful life. Under GAAP, nearly all R&D costs, including development, are expensed immediately and cannot be capitalised.
Key takeaways
- IFRS allows development costs to be capitalised; GAAP requires them to be expensed, creating a permanent structural difference in how earnings are reported.
- A capitalised development cost reduces current-year net income by zero, but spreads the expense over future years via amortisation; expensing it reduces current income immediately.
- This difference inflates IFRS-reported earnings and depresses operating cash flow (because capitalised costs still consume cash upfront) for companies with large development programs.
- International and multi-national companies can use this discretion to smooth earnings or mask high R&D intensity; the choice of useful life for amortisation adds another layer of manipulation.
- Investors comparing an IFRS reporter to a GAAP reporter, or assessing an IFRS company's true profitability, must recast R&D capitalisation to expensing to see earnings on a like-for-like basis.
The IFRS framework: research phase vs development phase
Under IAS 38, the treatment of R&D costs hinges on a hard line: research phase versus development phase.
Research phase costs are inherently uncertain. A company exploring new technologies, running early experiments, or investigating market viability cannot know whether a project will succeed. IFRS mandates that research costs be expensed in the period incurred. There is no debate here.
Development phase costs are different. IAS 38 allows capitalisation if all of the following criteria are met:
- Technical feasibility of completing the asset has been demonstrated.
- Management intends to complete and use or sell the asset.
- The company can demonstrate a future economic benefit (market or internal use).
- Adequate technical, financial, and other resources exist to complete development.
- The company can measure the development costs reliably.
Once these conditions are satisfied, the development costs are capitalised as an intangible asset and amortised over the asset's useful life—typically 3 to 10 years for software, longer for some industrial applications.
The critical word is "demonstrated." A company cannot capitalise on hope or intent alone. But in practice, the line between research and development is blurry, and management has room to argue that a project has crossed the threshold from exploratory to development.
How capitalisation inflates earnings relative to expensing
To see the profit impact, consider a simple example.
A software company spends €100 million on R&D in Year 1. Assume €40 million is classified as research (expensed) and €60 million as development (capitalised). The development asset has a 5-year useful life.
Under GAAP:
- Year 1 net income: reduced by €100 million.
- Years 2–5: no further charge (all costs were expensed in Year 1).
Under IFRS:
- Year 1 net income: reduced by €40 million (research) + €12 million (amortisation of €60 million / 5 years) = €52 million.
- Years 2–5: reduced by €12 million each year (amortisation).
Year 1 net income under IFRS is €48 million higher than under GAAP, even though both firms spent the same cash. An investor reading the IFRS company's Year 1 earnings would see it as 48% more profitable than the GAAP company, when in reality both consumed the same economic resources.
Over five years, cumulative earnings are identical (€100 million in charges either way), but the timing is radically different. IFRS front-loads profitability; GAAP front-loads the cost. If a company is growing R&D spending every year (which most tech and pharma companies do), the effect compounds: every year's capitalised development becomes a balance-sheet asset, and the company's income statement benefits from deferral.
The cash flow misalignment: another red flag
A sophisticated investor will spot a mismatch that reveals this game immediately: operating cash flow divergence.
When a company capitalises development costs, those costs still consume cash in the year they are incurred. The cash outflow hits the cash flow statement as part of operating activities (they are payments for salaries, contractors, equipment rentals used in development). But under IFRS, the income statement does not fully reflect the cost, only the amortisation.
The result is a widening gap between accrual-based net income and operating cash flow. On the income statement, the company reports a development cost of €12 million (one year's amortisation). But the cash flow statement shows the full €60 million was paid for development in Year 1. This gap—where net income is high but operating cash flow lags—is a classic warning that earnings are being flattered by capitalisation.
A GAAP-based company in the same situation would show the full €100 million R&D cost on the income statement and an equal reduction in net income, aligned with the cash outflow. There is no mystery.
This misalignment is not forensic fraud. It is permitted under IFRS. But it is a red flag that earnings quality is lower, because the income statement does not reflect the full current-period economic cost of the company's operations.
Useful-life games: another discretionary lever
Once a development asset is capitalised, management must estimate its useful life. A 5-year life versus a 10-year life for the same asset cuts the annual amortisation in half.
There is no universal standard. A company developing a mobile app might argue for a 3-year life (because mobile OS updates break compatibility). The same company developing a server-side API might argue for 8 years. Both arguments are defensible; both change the profit imprint.
Some industries set norms (pharmaceutical companies often use 7–10 years for drug development costs, if they capitalise them at all). But companies have room to move within the range, and moving the estimate by a few years can swing earnings by millions.
Analysts watching for manipulation should monitor the useful-life assumptions in the notes to financial statements. A change from 5 to 7 years, or a lengthening of useful lives, can signal that management is seeking to slow down amortisation and boost reported earnings.
Capitalisation vs software development: the murky frontier
Software and cloud-based services sit at the fuzzy boundary between research and development.
Many SaaS and enterprise software companies capitalise some portion of software development costs under IFRS. The logic is that once a platform reaches a "stable" version and is being enhanced incrementally, those enhancements are development work—improving an existing asset, not researching a new one.
But this reasoning is easy to abuse. A company can argue that all development after the first release is development (not research), capitalise it, and report earnings far higher than a competitor that expenses all software costs.
A well-known example: a European SaaS company reports under IFRS and capitalises 30% of its engineering payroll as development costs (amortised over 5 years). A comparable US competitor expenses 100% of R&D. The European company's reported profit margin is 5 percentage points higher—purely from accounting choice, not operational performance.
When comparing software companies across borders, or assessing the earnings quality of an IFRS software reporter, a recast of capitalised development to immediate expensing is essential.
Pharmaceutical R&D: a special case
Pharmaceutical and biotech companies operate on a different R&D timeline. Drug development takes 10+ years and billions of dollars, with high failure rates and binary outcomes (drug is approved or it is not). Most pharma companies expense virtually all R&D under both GAAP and IFRS, treating the costs as period expenses because the success is so uncertain.
However, some companies—particularly European pharmaceutical firms—have experimented with capitalising late-stage development costs (Phase III trials, regulatory submissions) under IFRS, arguing that technical feasibility has been demonstrated.
This is rarer than in software because of the inherent regulatory risk: even in Phase III, a drug can fail. But it happens, and when it does, it signals management's confidence (or desperation) in particular assets. An investor should note any shift toward capitalisation of development costs in pharma as a potential earnings lever or sign of confidence in a pipeline.
Real-world example: the IFRS vs GAAP earnings gap
Consider two hypothetical semiconductor design firms, one US (GAAP) and one German (IFRS), with identical operating models:
- Annual revenue: $500 million
- Total R&D spend: $120 million per year
- R&D as % of revenue: 24%
US GAAP company:
- Net income: $60 million (after expensing all $120 million R&D)
- Net margin: 12%
German IFRS company (capitalises 40% of R&D as development):
- R&D expensed immediately: $72 million
- R&D capitalised: $48 million
- Annual amortisation of capitalised assets (average): $12 million (assuming 4-year life)
- Total R&D charge to income: $72 million + $12 million = $84 million
- Net income: $96 million (after $84 million R&D charge)
- Net margin: 19.2%
The German company's reported margin is 60% higher than the US peer, purely because of the accounting choice. An investor unfamiliar with this difference might believe the German company is more efficient. In reality, both companies spend the same R&D intensity and have the same economic profitability.
A correct comparison requires recasting the German company's capitalised R&D as expense (adding back $48 million capitalised + subtracting $12 million amortisation, net +$36 million R&D charge, reducing net income to ~$60 million and margin to 12%, matching the US peer).
Spot the warning signs
Watch for:
-
Increasing capitalised development assets on the balance sheet, year over year, without a corresponding increase in amortisation or write-offs. This suggests the company is growing the pool of deferred costs.
-
R&D intensity (R&D as % of revenue) stays constant, but reported R&D expense as % of revenue declines. This gap signals that more R&D is being capitalised than amortised away.
-
Operating cash flow lags net income by a widening margin. If net income is growing but operating cash flow stagnates or grows slower, capitalisation is a likely culprit.
-
Management changes the useful life or capitalisation policy. The notes will disclose changes, and they often correlate with earnings guidance misses or management transition (new CFO, new accounting policy review).
-
Comparison against GAAP peers shows persistently higher margins for an IFRS company. If two companies are in the same industry with similar business models, large and persistent margin differences may signal that accounting policy, not operational excellence, is the source.
Common mistakes
-
Assuming capitalisation is fraud. It is not. IFRS permits it, and many companies use it legitimately. The risk is not criminal but rather that earnings are less comparable and less cash-backed than reported.
-
Ignoring the research-phase costs. A company can still express high R&D intensity through research costs (which are always expensed). Capitalisation of development does not hide all R&D; it only defers part of it.
-
Not comparing apples to apples. Many investors compare a US company's reported earnings to an IFRS company's reported earnings without adjusting for capitalisation. This is a critical mistake. Always recast.
-
Forgetting the balance-sheet trail. If you recast capitalised development as immediate expense, you must also adjust the balance sheet (reduce intangible assets, reduce shareholders' equity). This recast affects downstream ratios like ROE, debt-to-equity, and book value per share.
-
Assuming longer useful lives are always red flags. A 10-year life for a core platform asset can be reasonable. The flag is not the absolute life, but changes to the life, and the gap between capitalisation and expensing amounts versus industry norms.
FAQ
Is capitalising development costs ever appropriate, even under GAAP?
Not in general. GAAP's standard is that R&D should be expensed. However, there is a narrow exception for software development costs (ASC 985), where some costs incurred after technological feasibility is reached can be capitalised. But this is much more restrictive than IFRS and applies to a smaller pool of costs.
How do I find capitalised development costs on the balance sheet?
Look for "Capitalised software" or "Capitalised development costs" under intangible assets (non-current assets). The amortisation expense is usually detailed in the notes, under a section on intangible assets or under "Summary of Significant Accounting Policies."
If a company amortises development costs over 10 years, does that mean the asset is good for 10 years?
Not necessarily. The useful life is management's estimate of how long the asset will generate economic benefits. It can be wrong, and it is sometimes chosen with an eye toward earnings smoothing. Tech products often become obsolete faster than management's stated lives. It is worth comparing stated lives to actual product lifespans (based on company announcements or industry knowledge).
Can a company move costs between research and development to manipulate earnings?
In theory, yes. The boundary is fuzzy, and moving $10 million from research (expensed) to development (capitalised) would save $2 million in Year 1 charge (assuming a 5-year amortisation life). In practice, auditors and the company's own review processes are supposed to prevent this. But it is a risk, especially in private companies or those with weak governance.
Should I always add back capitalised development costs when calculating free cash flow?
Not exactly. Capitalised development is already a cash outflow in the operating section of the cash flow statement (or sometimes the investing section, depending on classification). The recast is for earnings, not for cash flow. To compare apples to apples on cash flow, you compare the two companies on their actual reported operating cash flow, which already reflects the cash impact.
Does capitalisation affect the effective tax rate?
It can. The tax treatment of capitalised versus expensed R&D varies by jurisdiction. Some countries allow immediate write-offs (R&D credits), others amortise. This can cause the IFRS reporter's tax rate to differ from a GAAP reporter's even if both have the same pre-tax earnings. The notes should disclose this, but it adds another layer of complexity to cross-border comparison.
How do I adjust earnings for capitalisation if I do not know the full detail?
As a rough proxy, if the notes disclose total capitalised development and total amortisation, you can estimate the implied pool of deferred costs and the trend. If capitalised amounts are growing faster than amortisation, costs are accumulating (a warning sign). If the two are roughly balanced, the company is in a steady state. You can also compare the company's net income to operating cash flow; if net income is consistently 5–10 percentage points higher than the cash flow would suggest, capitalisation is likely a factor.
Related concepts
- Intangible assets and useful-life estimation (IAS 38 and ASC 805): How all intangible assets are valued and amortised, and how the choice of useful life is a discretionary lever.
- Capitalising vs expensing (Chapter 13, Article 05): A broader look at the mechanics of capitalisation decisions and where they hide earnings.
- Software development capitalisation games (Chapter 13, Article 06): Specific patterns in how SaaS and software firms use capitalisation to flatten reported costs.
- GAAP vs IFRS: development costs (Chapter 6, Article 07): The foundational comparison of how the two frameworks differ on this issue.
- Comparing international companies: How to recast financial statements when mixing GAAP and IFRS reporters.
Summary
IFRS's permission to capitalise development costs is a structural advantage for reported profitability that GAAP competitors do not enjoy. A company spending €100 million on R&D can report substantially higher earnings than a GAAP-based competitor spending the same amount. This difference is not fraud—it is a permitted accounting choice—but it is a red flag for earnings quality and comparability.
The key warning signs are a widening gap between net income and operating cash flow, a rising balance-sheet pool of capitalised assets, and persistent margin advantages over GAAP peers. When comparing international competitors or assessing an IFRS reporter's true profitability, recast capitalised development as immediate expense to see earnings on a level playing field.
Next
Read about Sudden changes in deferred revenue, another pattern where timing shifts hide the true trend in customer economics.