Section 174 capitalization produced one of the least expected tax outcomes of recent years: companies losing money found themselves owing federal tax.
The change was mechanical rather than conceptual. Research and experimental spending that had been deducted in the year incurred must now be amortised across several years, which increases taxable income without changing cash spending at all.
This guide covers what the rule requires, why it hits loss-making companies hardest, and seven practical effects.
Research and experimental expenditure must be capitalised and amortised rather than deducted immediately. Domestic spending amortises over five years, foreign spending over fifteen.
The definition is broad and captures most software development salaries that had been claimed as ordinary startup tax deductions, which is why technology companies felt the change more sharply than most sectors.
| Treatment | Before | Under Section 174 |
|---|---|---|
| Domestic R and E spending | Deducted in year incurred | Amortised over 5 years |
| Foreign R and E spending | Deducted in year incurred | Amortised over 15 years |
| First-year deduction | 100% | 10%, half-year convention |
| Effect on taxable income | Matches cash | Increases in early years |
| Cash spending | Unchanged | Unchanged |
| Effect on loss-making companies | None | Can create taxable income |
A company spending $2m on engineering and generating $1.5m revenue previously showed a loss. Under Section 174 capitalization only a fraction of that spending is deductible in year one, so the same company can show taxable profit.
The cash position has not changed. The money left the business exactly as before. What changed is how much of it the tax code allows you to deduct this year.
This is why the rule surprised so many founders. Nothing about the business was different and a tax liability appeared where none had existed.
Domestic spending amortises across five years with a half-year convention, meaning only a tenth is deducted in the first year. The remainder follows across the subsequent years.
The effect reduces over time as earlier years’ amortisation stacks. A company with stable R and E spending reaches a steady state after five years where deductions approximate spending again.
The first three affect cash directly. The rest affect planning and reporting.
The headline effect. A company with genuine losses can owe federal tax because most of its largest expense is not currently deductible. This caught a substantial number of startups unprepared.
Quarterly estimated tax payments based on prior-year patterns will be wrong. A company that paid nothing in prior years may now need substantial quarterly payments, and the underpayment penalty applies regardless of the surprise.
Foreign research amortises over fifteen years rather than five. For companies with engineering teams abroad this materially increases the after-tax cost of that work.
An R and D tax credit guide is now essential reading rather than optional, because the credit directly offsets the liability Section 174 creates. The two provisions interact and should be planned together.
Models that assumed no tax liability while loss-making are wrong. The forecast needs a tax line driven by the amortisation schedule rather than by accounting profit.
The timing difference between book and tax treatment creates deferred tax assets requiring tracking and disclosure. This adds genuine complexity to financial statements for smaller companies.
What counts as research and experimental expenditure is broader than many assumed and includes most engineering salaries previously treated as startup tax deductions. Section 174 capitalization therefore requires a classification exercise most businesses had never done.
Getting that classification wrong in either direction creates exposure, which is why it is worth doing deliberately rather than by assumption.
A Raleigh analytics company spent $1.9m on US engineering salaries against $1.4m revenue, producing a book loss of $740,000 after other costs. The founder expected no federal tax, as in prior years.
Under Section 174 only $190,000 of the engineering spend was deductible in year one. Taxable income was positive.
| Line | Book treatment | Tax treatment |
|---|---|---|
| Revenue | $1,400,000 | $1,400,000 |
| Engineering salaries | $1,900,000 expensed | $190,000 amortised |
| Other operating costs | $240,000 | $240,000 |
| Result | Loss of $740,000 | Income of $970,000 |
| Federal tax before credit | $0 expected | $203,700 |
| R and D credit applied | – | $118,000 |
| Net liability | – | $85,700 |
The R and D credit absorbed more than half the liability, which is why the two provisions need planning together. The remaining $85,700 arrived with no estimated payments made against it, adding an underpayment penalty to a bill nobody had forecast.
A requirement that research and experimental expenditure be capitalised and amortised rather than deducted in the year incurred. Domestic spending amortises over five years, foreign spending over fifteen, with a half-year convention in year one.
Because only a fraction of research spending is deductible in the first year. A company with genuine book losses can show taxable income once its largest expense is substantially removed from the current year deduction.
Research and experimental expenditure broadly defined, including most software development salaries. The definition is wider than many founders assumed, which is why a classification exercise is usually needed.
They offset each other partially. An R and D tax credit guide is worth reading alongside this, because the credit directly reduces the liability Section 174 creates and the two should be planned as one exercise.
Almost certainly. Quarterly estimated tax payments based on prior years will be wrong if those years assumed full deduction of research spending. The underpayment penalty applies regardless of whether the liability was anticipated.
Yes, substantially. Foreign research amortises over fifteen years rather than five, which materially increases the after-tax cost of development performed outside the United States.
Section 174 changed nothing about how businesses spend and a great deal about what they owe. The companies that handled it well were those that modelled the liability before it arrived rather than discovering it at filing.
Explore our fractional CFO services, read our R and D tax credit guide, or how estimated payments work. Our financial reporting services cover deferred tax disclosure.
Ask us to model your Section 174 position.
Animesh Shah | Accounting Lead, Finkeepers
Chartered Accountant with 15+ years leading accounting delivery for US businesses across QuickBooks, NetSuite, Odoo and Sage. Tax guidance referenced here is published by the IRS.
This article provides general guidance for educational purposes only. Tax treatment, amortisation schedules and legislative provisions change. Always consult a qualified CPA before relying on this content for filing or planning decisions. Last updated October 2026.
Built by experienced professionals, Finkeepers delivers reliable accounting services for startups and SMBs, powered by proven processes and modern software.