When investors talk about “quality of earnings,” they’re usually probing one core question: how much of this quarter’s profit reflects the underlying economics of the business—versus the pliability of the rulebook. Few places test that boundary like research and development. The treatment of R&D costs can turn a loss into a profit, lift EBITDA without a euro of extra cash, and reshape return metrics in ways that flatter or obscure performance. Under IFRS, management has far more room to maneuver than under U.S. GAAP. And, yes, many investors see aggressive capitalization as a negative signal—because it tends to overstate current earnings relative to cash reality.

Two Playbooks, Two Very Different Outcomes

U.S. GAAP keeps things simple: except for a handful of software-related exceptions, R&D is expensed as incurred. That pushes the cost into today’s income statement and forces managers to justify spending in real time. It isn’t perfect, but it is hard to game.

IFRS takes a split view. “Research” is expensed, but “development” must be capitalized once a project satisfies a list of feasibility and commercialization hurdles—technical feasibility, intention and ability to complete and use or sell, probable future benefits, adequate resources, and reliable measurement of cost. Those criteria sound rigid; in practice they leave considerable room for judgment about timing and probability. The result is an accounting asset on the balance sheet and amortization later, instead of a hit to today’s P&L.

Even within IFRS, there is more nuance. Internally generated brands and similar items are barred from recognition, but acquired in-process R&D is capitalized and then tested for impairment. For software and games, IFRS reporters commonly capitalize significant development outlays once the criteria are met. The line between “research” and “development,” and when a project crosses it, is where the judgment—and the flexibility—lives.

Why Capitalizing R&D Can Inflate Today’s Earnings

Capitalizing development costs pushes expenses into the future via amortization. That has three predictable effects that equity analysts care about:

1. EBITDA gets a tailwind. Amortization is excluded from EBITDA by convention, while expensing would hit operating costs today. Capitalization therefore boosts EBITDA without adding a cent of incremental cash. It can also improve debt covenants and valuation multiples keyed to EBITDA. (Standards don’t define EBITDA; market practice does, which is precisely the point.)
2. Operating profit and margins look better now, worse later. Amortization trickles in over a useful life the company chooses—another judgment call. Front-loading benefits and back-loading costs is classic earnings management territory, even if it’s all within the rules.
3. Return metrics can be engineered. Capitalizing increases invested capital, which can depress ROIC; but by lifting NOPAT today (no expense now), the net effect can still flatter trends—especially if amortization schedules are long or impairment testing is slow to catch bad projects. Impairments tend to show up late and lumpy.

In short, capitalization smooths the optics and decouples reported performance from cash outlay. That’s why many fundamental investors treat large swings in capitalized development as a quality-of-earnings flag and rebuild the income statement to a “cash R&D” basis.

The Flexibility Under IFRS—And How It’s Used

IFRS doesn’t give management a free pass, but it does give them levers:

• When feasibility is “achieved.” Crossing the line from research to development is inherently subjective. Move that point earlier and more spend gets capitalized. Move it later and more is expensed. The standard lists necessary conditions; it doesn’t stamp a date.
• What counts as directly attributable. Capitalized cost can include salaries of engineers, testing, and certain overhead allocations—areas where policy choices and time-tracking rigor matter.
• Useful lives and impairment cadence. Longer useful lives mean smaller annual amortization charges. Impairment testing relies on management cash-flow forecasts and discount rates; optimism delays pain.

Under U.S. GAAP, flexibility shows up mainly in software: after “technological feasibility” (for software to be sold) or once a project enters the application-development stage (for internal-use software), certain costs are capitalized. The rules are narrower and more prescriptive than IFRS’ development criteria, but they still require judgment—especially on the feasibility crossing point for sellable software.

Why Many Investors Haircut Capitalization

Investors aren’t opposed to recognizing durable assets. In valuation, plenty of practitioners do build R&D assets and amortize them to estimate economic profit. But in financial reporting, the asymmetry between optics and cash creates skepticism. Capitalization can:

• Obscure cash burn. Operating cash flow doesn’t get better when you capitalize—it just looks like it does if you focus on EBITDA or operating profit. Seasoned readers migrate to cash-flow statements and capex footnotes to track what’s really leaving the building.
• Encourage “policy drift.” The slope is gradual: a tweak to feasibility timing this year, a nudge to useful life next year. Each move is defensible; together they compound.
• Complicate comparability. Two IFRS peers can apply the same standard and arrive at very different capitalized R&D balances. Compare them to a U.S. GAAP competitor that expenses nearly everything, and headline margins can mislead. That’s why cross-border analysis often requires normalizing adjustments.

Put plainly: when the same cash spending can yield two very different earnings profiles, the market discounts the rosier one until proven otherwise.

What To Watch (and How To Adjust)

If you want an apples-to-apples view, rebuild the numbers:

• R&D-adjusted EBIT(DA). Start with reported operating profit (or EBITDA). Add back amortization of previously capitalized development costs. Then expense current-period capitalized development (treat it as if it went through the P&L). The result is closer to a “cash R&D” income statement.
• R&D asset roll-forward. Track openings, additions, amortization, and impairments. Sudden growth in additions without a commensurate payoff in product revenue is a yellow flag.
• Useful life sanity checks. Compare amortization periods to product cycles and competitive churn. If the policy assumes benefits last seven years but the market turns every three, you know where the pressure will land.
• Software carve-outs. Under both frameworks, software has bespoke rules. Understand whether big capitalization lines are software for sale (GAAP: after feasibility) or internal-use builds (GAAP: application stage), or IFRS development under IAS 38. Different rule sets, similar incentives.

The Bottom Line

Standards are supposed to force clarity. In R&D, they also open doors. U.S. GAAP’s default—expense it—keeps earnings closer to cash. IFRS’ split—expense research, capitalize qualifying development—seeks to match costs with benefits but hands managers real discretion over timing and magnitude. That discretion can be used wisely; it can also burnish short-term optics. The market knows this. Which is why, when an IFRS filer leans hard into R&D capitalization, many investors don’t cheer—they reach for their models and take some of the shine off the headline numbers.