An intangible asset is an identifiable non-monetary asset without physical substance. Examples include patents, licenses, software, customer relationships, acquired brands. Accounting for them turns on one question. How did you get it?
Purchased intangibles (bought outright or in an acquisition) are capitalized at cost or fair value. What happens next depends on their life:
- A finite-life intangible (patents, customer lists, licenses) is amortized over its useful life, usually straight-line, and tested for impairment when indicators arise.
- An indefinite-life intangible, such as some acquired brands and certain licenses, isn't amortized at all. It gets an impairment test at least annually instead. "Indefinite" here means no foreseeable limit, which isn't the same as "infinite". Kraft's and Oscar Mayer's trademarks were in this bucket, so remember them for goodwill impairment.
How are internally generated intangibles accounted for?
For internally generated intangibles, the frameworks split, and this is the biggest comparability distortion in all of accounting:
| US GAAP | IFRS / Ind AS (IAS 38 / Ind AS 38) | |
|---|---|---|
| Research | Expense as incurred (ASC 730) | Expense as incurred |
| Development | Expense as incurred — with two software exceptions below | Capitalize once six criteria are all met; expense before that point (no retro-capitalization) |
| Internally generated brands, mastheads, customer lists, goodwill | Never capitalized | Never capitalized |
Advertisement
Memorize the six IFRS/Ind AS development criteria with PIRATE: Probable future economic benefits · Intention to complete · Resources adequate to complete · Ability to use or sell it · Technical feasibility demonstrated · Expenditure measured reliably. All six have to be met at once, and until they are, you expense.
The US software nuances. You'll be asked about these in every interview.
- Software to be sold or marketed (ASC 985-20). Expense until technological feasibility (a working model or detailed program design), then capitalize from there to general release. In modern practice feasibility arrives so late that most product-software firms expense nearly everything.
- Internal-use software, including SaaS platforms (ASC 350-40). Expense the preliminary and post-implementation stages, but capitalize the application-development stage (coding, configuration, testing). So US SaaS companies do capitalize a slice of engineering payroll. Read the "capitalized software development" line in the notes and check the amortization period.
Does IFRS really let companies capitalise more R&D?
Reality check before you assume a gap. IFRS/Ind AS capitalization is conditional, and industries meet the criteria at very different points. Pharma is the cautionary case. "Probable benefits" and "technical feasibility" typically arrive only with late-stage or regulatory success. So Sun Pharma (Ind AS) expenses essentially all internal R&D, just like Pfizer under US GAAP, and their policy notes say so. At Tata Motors, though, the gap is real and large. Since the demerger effective 1 October 2025, the JLR-owning company is Tata Motors Passenger Vehicles Ltd. Its JLR unit has historically capitalized the majority of its product-engineering spend under IFRS/Ind AS. Sell-side analysts routinely deduct capitalized development from JLR's EBITDA before comparing it with Ford or GM, which expense it all under US GAAP. So here's the rule. Never adjust before reading the accounting-policy note. The standard tells you what's allowed, and the note tells you what's done.
How do analysts adjust for expensed R&D?
The analyst's fix: capitalize R&D yourself (Damodaran's adjustment). For research-driven firms, expensing R&D understates both profit and invested capital, which mangles ROIC and margins. You'll reuse the repair when you get to ROIC and again in Phase 3. Pick an amortizable life. Damodaran's rules of thumb are about 5 years for shorter-cycle businesses and generics, and about 10 for innovator pharma. Then rebuild an R&D asset from past spend and restate operating profit.
Worked example: Sun Pharma vs Pfizer, illustrative. These numbers are round and modeled on roughly FY24/FY25 scale. Sun Pharma's consolidated revenue is roughly ₹48,000 crore, with R&D near ₹3,000 crore (about 6%). Pfizer's is roughly $64bn, with about $11bn of R&D (about 17%). Pull the actual filings. Take Sun Pharma with a 5-year amortizable life, and start amortization the year after the spend. Here's an illustrative R&D history in ₹ crore: FY20 1,750 · FY21 2,000 · FY22 2,250 · FY23 2,500 · FY24 2,750 · FY25 3,000.
- Research asset, end-FY25 = 3,000×100% + 2,750×80% + 2,500×60% + 2,250×40% + 2,000×20% = 3,000 + 2,200 + 1,500 + 900 + 400 = ₹8,000 crore.
- FY25 amortization = 20% × (1,750 + 2,000 + 2,250 + 2,500 + 2,750) = 20% × 11,250 = ₹2,250 crore.
- Adjusted EBIT = reported EBIT + current R&D − R&D amortization = 10,000 + 3,000 − 2,250 = ₹10,750 crore, assuming reported EBIT of ₹10,000 crore for the illustration.
- Adjusted invested capital = 40,000 + 8,000 = ₹48,000 crore, assuming reported IC of ₹40,000 crore.
- ROIC (25% tax): unadjusted = 10,000×0.75/40,000 = 18.8%; adjusted = 10,750×0.75/48,000 = 16.8%.
Notice the direction. With a growing R&D budget, adjusted EBIT rises a little but invested capital rises a lot, so measured ROIC usually falls. R&D-heavy "asset-light" compounders are less magically profitable than their raw ratios suggest. On the other hand, their earnings are more durable than the P&L implies, because a third of "cost" is actually investment. Run Pfizer through the same machinery with a 10-year life and about $11bn of annual spend, and you build a research asset in the $50 to 60bn range. That's a real asset, and it's on no balance sheet. You get comparability by adjusting both companies with the same method. Trusting either book number won't get you there.