Intellectual property valuation rests on three approaches, and the choice between them depends on which right you are valuing.
Cost, market, income. Every method is a variant of one of those.
But a patent expires on a fixed date, a trademark can last forever, and a trade secret lasts only while nobody else works it out. Duration alone changes the arithmetic enough that one method cannot span them.
The common failure is applying one approach across a mixed portfolio and reporting a single number.
The three approaches
| Approach | Asks | Best for |
|---|---|---|
| Cost | What did it take to create, or to replace? | Early-stage, internally developed assets |
| Market | What have comparable assets sold for? | Where comparables exist |
| Income | What future benefit does it produce? | Most commercial situations |
Cost is the weakest for patents. Prosecution spend reflects effort, not coverage, and an expensive patent covering an approach nobody uses is worth less than a cheap one covering what everybody does.
Market suffers from thin comparables. Most IP transactions are private, and disclosed deals frequently bundle assets, which makes clean comparison difficult.
Income dominates in practice, and it depends entirely on assumptions.
Income methods
| Method | Detail |
|---|---|
| Relief from royalty | What you would pay to license it if you did not own it |
| Excess earnings | Earnings above what the business would make without it |
| Incremental cash flow | The difference the asset makes |
| Premium pricing | What buyers pay extra for the branded or protected version |
| Cost savings | Where the asset reduces expense rather than raising revenue |
Relief from royalty is the most portable. It works for patents, trademarks and copyright, and it converts the question into a royalty rate and a revenue base.
Which means the same two inputs decide most valuations — and both are judgements.
Why the right type matters
| Patents | Trademarks | Copyright | Trade secrets | |
|---|---|---|---|---|
| Duration | Fixed expiry | Indefinite with renewal | Long defined term | While secret |
| Value over time | Decays | Can appreciate | Decays slowly | Stable, then zero |
| Ongoing cost | Maintenance fees | Renewal fees | Minimal | Security measures |
| Main risk | Invalidity | Non-use, genericide | Ownership, termination | Disclosure |
| Registration | Required | Usually | Optional | None |
| Best approach | Income, market | Income | Income | Income, discounted |
Trademarks appreciating is the structural difference. A patent's value falls mechanically as term runs down; a mark's can rise as recognition grows.
Remaining term drives patent value
| Remaining term | Effect on value |
|---|---|
| 15+ years | Full income horizon |
| 10 years | Strong |
| 5–7 years | Meaningfully reduced |
| Under 3 years | Difficult to sell at all |
| Expired | Zero |
The decay is not linear in practice. Buyers apply their own thresholds, and below roughly three years many will not transact at all regardless of the claims.
Which makes timing a value driver in its own right, separate from quality. See patent portfolio.
Maintenance cost is part of the calculation
| Fee | Due after grant | Large | Small | Micro |
|---|---|---|---|---|
| First | 3.5 years | $2,150 | $860 | $430 |
| Second | 7.5 years | $4,040 | $1,616 | $808 |
| Third | 11.5 years | $8,280 | $3,312 | $1,656 |
| Total | $14,470 | $5,788 | $2,894 |
A buyer inherits every remaining fee, plus foreign annuities, and prices accordingly.
| Ipiry Patent Survival Curve v1.0 | Rate |
|---|---|
| Survive the 3.5-year fee (2022 cohort) | 85.8% |
| Survive the 7.5-year fee (2018 cohort) | 64.6% |
| Reach full term (2014 cohort) | 41.4% |
| Abandoned before full term | 58.6% |
Computed from 27,273,654 USPTO maintenance fee records covering 8,262,336 US utility patents — see the patent survival curve.
Nearly three in five utility patents are released by their own owners, which is the market's own assessment that the ongoing cost exceeded the value.
What actually drives patent value
| Driver | Weight |
|---|---|
| Claims read on products in the market | Highest |
| Documented evidence of that | Very high |
| Remaining term | High |
| Clean recorded ownership | Blocking if absent |
| Family depth and jurisdictions | High |
| Validity robustness | High |
| Prosecution spend | Low |
| Patent count | Low |
A documented claim chart against a shipping product is the single largest factor. A patent with one is a different asset from a patent without one.
Ownership defects are blocking rather than discounting. A gap in the recorded chain stops transactions rather than reducing the price. See assignment search.
Worked example: one portfolio, four asset types
A company being acquired holds a mixed IP estate.
| Asset | Approach | Key input |
|---|---|---|
| 8 patents, 6–13 yrs remaining | Relief from royalty | Rate × revenue on covered products |
| Brand and 3 marks | Relief from royalty | Brand-attributable revenue, no expiry |
| Software and documentation | Cost to replace, cross-checked on income | Development cost |
| Manufacturing process | Income, discounted for loss risk | Reverse-engineering probability |
Why one number would mislead
| Finding | Implication |
|---|---|
| 2 patents expire within 4 years | Their contribution is short-lived |
| Trademark value grows with revenue | Opposite direction from the patents |
| Trade secret carries loss risk | Not comparable to registered rights |
| 3 patents have no evidence of use | Speculative contribution |
The patents and the marks move in opposite directions over time. Aggregating them into one figure hides that entirely.
The three patents with no documented use are the soft part of the number, and a buyer will find that in diligence whether or not the valuation flagged it.
Where the variance comes from
| Input | Effect of a small change |
|---|---|
| Royalty rate | Proportional |
| Revenue base | Proportional — and contested hardest |
| Discount rate | Compounds over the horizon |
| Useful life | Truncates or extends everything |
| Probability adjustments | Multiplies through |
Base and rate together decide most of it. The same rate applied to a component or a finished product can differ by more than an order of magnitude. See patent royalty rates.
Two defensible valuations can differ enormously, which is why stated assumptions matter more than the headline figure.
What data exists, and what does not
| Available | Not available |
|---|---|
| Patent value estimates from academic datasets | Comprehensive transaction prices |
| Some disclosed licence terms | Most licence terms — private |
| Litigation damages awards | Settlement amounts — usually confidential |
| Maintenance fee behaviour | Reasons for individual decisions |
Academic datasets such as KPSS provide patent value records, estimated from stock-market reactions at grant. They are value estimates, not sales or transaction records, and treating them as prices is a common misreading.
No public dataset contains comprehensive patent sale prices, which is why the market approach is harder for patents than for most asset classes.
Purpose shapes method
| Purpose | Emphasis |
|---|---|
| Sale or purchase | Market and income |
| Licensing negotiation | Relief from royalty |
| Financing | Realisable value, conservative |
| Litigation damages | Statutory framework, not general valuation |
| Tax and accounting | Standard-driven |
| Insolvency | Liquidation basis |
Damages are not a valuation exercise. They follow the statutory framework — reasonable royalty floor, lost profits above it, apportionment — which is a different analysis from what an asset is worth. See patent infringement damages.
Intellectual property valuation: the checklist
- Choose the approach by right type. One method does not span a mixed portfolio.
- Prefer income methods for commercial questions; cost tells you least.
- Value patents against remaining term, which drives decay directly.
- Value trademarks knowing they can appreciate, unlike patents.
- Discount trade secrets for loss risk. No expiry, but no protection either.
- Subtract maintenance and annuity costs from any patent income figure.
- Weight documented evidence of use above every other patent factor.
- Treat ownership gaps as blocking, not discounting.
- State every assumption. Base, rate, discount rate and useful life.
- Report components, not just a total, so the reader can see where value sits.