Intangible assets—which is a polite, accounting-approved way of saying "things you can't drop on your foot"—now comprise roughly 90% of the market value of S&P 500 companies. IP-intensive industries generate 45% of EU GDP (€6.6 trillion, which is a number so large that writing it out would take up the rest of this paragraph) and over 38% of U.S. GDP. The most valuable things in the global economy are the things that don't, in any physical sense, exist.
And yet—and here is the thing that should keep you up at night if you are an in-house patent counsel or a startup founder, which, if you've read this far, you probably are—almost nobody knows how to figure out what these non-things are worth. This is like owning ninety percent of a house and not knowing whether it's a Tudor in Greenwich or a lean-to in a flood plain. It is, to use a technical term, a problem.
What follows is everything[^1] you need to know about patent valuation. It is intended for people who make decisions about patents—filing them, buying them, selling them, licensing them, waving them around in litigation like a very expensive flag—and who would prefer those decisions to rest on something more rigorous than vibes.
[^1]: "Everything" here being used in the same loose, hopeful, almost certainly inadequate sense that your college roommate used it when he said he'd "tried everything" to fix the Wi-Fi.
I. What Patent Valuation Is, and Why You Should Care (and Why the Fact That You Probably Don't Care Enough Is Itself Part of the Problem)
Patent valuation is the process of determining the monetary worth of a patent or patent portfolio. WIPO defines it as determining the monetary value of IP assets, noting that the value derives fundamentally from the owner's right to exclude competitors from using the invention. For an IP asset to have quantifiable value, it must generate measurable economic benefits and enhance the value of other assets it's associated with.
This sounds straightforward. It is not.
Here is a distinction that matters more than you think: price is what a buyer will pay; value is a calculated quantity based on tested methods. These are not the same thing, in the same way that what you'd pay for a bottle of water in the desert is not the same as what the water is "worth." Beyond this already-confusing bifurcation, there are four different standards of value that apply depending on why you're doing the valuation in the first place:
Fair market value is the IRS standard. It assumes a hypothetical willing buyer and seller, both with reasonable knowledge, neither under compulsion. It is, in other words, a fiction—but a useful fiction, like representative democracy or the efficient market hypothesis.
Fair value is the GAAP/IFRS standard under ASC 820. It's the exit price in an orderly transaction between market participants. Notably, it does not apply marketability or control discounts, which is one of those details that seems minor until it costs you several million dollars.
Investment value reflects what a patent is worth to a specific buyer, synergies and all.
Strategic value captures the full benefit of combining a patent with a buyer's complementary assets, which is why acquisition premiums routinely exceed fair market value, and which is why Google paid $12.5 billion for Motorola Mobility when the patents alone were arguably the prize. But we'll get to that. We'll get to everything.
The occasions that demand a patent valuation are numerous enough to make a list unavoidable, and I apologize for the list, because lists are the coward's way out of the paragraph, but here we are: fundraising (investors want to know what they're buying into), M&A (ASC 805 requires acquirers to allocate purchase price to identifiable intangible assets at fair value), licensing negotiations, litigation (where your royalty calculation has to survive Daubert scrutiny and Federal Circuit review, which is roughly the intellectual property equivalent of running a gauntlet), portfolio management, and various tax and accounting requirements including transfer pricing, charitable donation appraisals, and impairment testing. IP-backed lending is also growing rapidly—NatWest has accepted IP as sole collateral since January 2024, and China reported 26,000 businesses receiving IP pledge loans totaling 401.5 billion yuan in 2022.
The non-things have collateral value now.
II. The Six Ways Humans Have Figured Out to Price an Idea, Ranked Roughly from "Well, At Least It's Something" to "This Actually Kind of Works"
Method 1: The Cost-Based Approach, or, Measuring the Hole You Dug
The cost-based approach rests on what economists call the principle of substitution: a rational investor will pay no more for an asset than the cost to obtain one of equal utility. Three variants exist. Reproduction cost estimates what it would take to create an exact replica of the patent at current prices. Replacement cost estimates the cost of obtaining a functionally equivalent patent, adjusted for various forms of obsolescence—economic, functional, technological—which is a polite way of saying "adjusted for reality." Historical cost just tallies up what you actually spent: R&D, prosecution, filing fees, the psychic toll of dealing with the USPTO.
Here is why this approach is the floor and not the ceiling: there is absolutely zero correlation between how much money you pour into developing something and how much that something turns out to be worth. You can spend $50 million and get a patent that nobody will ever license. You can spend $50,000 and get one that anchors a billion-dollar franchise. The cost approach is like valuing a painting by the price of the canvas and oils. The SEC has expressed skepticism about cost-based approaches for most intangible asset valuations, which is government-speak for "please stop doing this."
But—and this is important—it works as a floor. For early-stage technology where zero revenue data exists, it at least tells you: somebody thought this problem was worth this much money to solve. That is not nothing.
Method 2: The Market-Based Approach, or, What Did the Other Guy Pay?
The market approach values a patent by comparing it to similar patents that have changed hands in arm's-length transactions, with adjustments for differences in technology, claim scope, remaining life, geographic coverage, and deal terms. It is conceptually the simplest approach. It is also, in practice, among the hardest.
The problem—and it is a foundational, load-bearing, not-going-away problem—is that every patent is unique. Finding a "comparable" patent transaction is like finding a comparable snowflake. And most IP transactions are confidential, which means you're looking for comparable snowflakes in the dark.
Key data sources have emerged to address this gap, and they represent genuine progress. ktMINE (now Wolters Kluwer) houses 85,000+ royalty rates and 25,000+ licensing agreements. RoyaltyStat (via FactSet/Exactera) offers 15,500+ unredacted license agreements from SEC filings. RoyaltySource has tracked IP licenses for 35+ years. And Ocean Tomo's Bid-Ask Market posts sale prices for publicly known patent transactions—the only true global patent price-discovery platform, which is remarkable and also a little terrifying, like learning there's only one thermometer for the whole hospital.
Even with these resources, the market approach works best as a cross-check against other methods, not as a standalone valuation for individual patents.
Method 3: The Income-Based Approach, or, Show Me the Money (Discounted to Present Value)
And now we arrive at the workhorse. The income approach values a patent based on the present value of future economic benefits it will generate. It is the most commonly used method according to WIPO, courts, and accounting standards, and for good reason: it directly connects patent value to the thing that actually matters, which is money.
Three variants dominate.
The Discounted Cash Flow (DCF) method does exactly what it sounds like: forecast the cash flows attributable to the patent over its remaining economic life, select an appropriate discount rate, calculate present value, and—this is the hard part—try not to let your projections become an exercise in creative writing. It requires assumptions about revenue, market penetration, operating costs, tax rates, and terminal value, every one of which is, at best, an educated guess about the future. It works best for established technologies with predictable income streams, which is to say, the technologies that least need your guessing.
The Relief from Royalty method asks a beautiful hypothetical question: what royalty would you have to pay a third party if you didn't own this patent? You apply a market-derived royalty rate to projected revenue, adjust for taxes, discount to present value, and there you are. It elegantly marries the income approach (revenue forecasts) with the market approach (comparable royalty rates), which is why it's one of the most widely used methods for ASC 805 purchase price allocations.
The Multi-Period Excess Earnings Method (MPEEM) takes total operating cash flows and subtracts returns attributable to every other asset—tangible assets, working capital, workforce, other IP—leaving whatever's left as the patent's contribution. It's the valuation equivalent of determining who's actually carrying the team by benching everyone else.
Now here's where things get squirrely, and where a lot of valuations go wrong. Discount rate selection. Patent-specific discount rates should be substantially higher than a company's WACC because patents are, objectively, among the riskiest assets in the enterprise. Typical rates range from 20% to 40% or higher, versus corporate WACCs of 8–15%. Early-stage IP may warrant rates benchmarked against VC expected returns—25–50%+. Patent-specific risk premiums account for claim construction uncertainty, validity risk, technology obsolescence, and enforcement difficulty—which is a lot of ways for things to go wrong, when you think about it. And you should think about it. That's what the discount rate is for.
Method 4: Real Options, or, Your Patent Is a Call Option and That Is Not a Metaphor
This is the method that makes finance professors very excited and everyone else slightly nervous.
The real options approach treats a patent as a call option—the right, but not the obligation, to commercialize the underlying technology. Using Black-Scholes or binomial models, it maps five key inputs: the present value of commercialization cash flows (S), the investment required to commercialize (K), the remaining patent term (T), the volatility of the technology's value (σ), and the risk-free rate (r).
Here is why this matters: DCF systematically undervalues anything with high uncertainty and optionality. Damodaran's analysis of the Avonex pharmaceutical patent showed a real options value of $907 million versus $547 million NPV—a $360 million time premium reflecting the value of managerial flexibility. In another case, a TechFlow Corporation AI research program showed traditional NPV of negative $11.2 million but a real options value of positive $9.2 million. The NPV said "don't invest." The real options model said "this uncertainty is a feature, not a bug."
WIPO specifically recommends the binomial model for early-stage IP, because it handles staged development decisions and changing volatility at each node. Real options valuation is especially valuable for pharmaceutical patents with binary regulatory outcomes (FDA approval = jackpot; rejection = write-off), platform technologies that may spawn multiple products, and any early-stage IP where uncertainty is high but upside is substantial.
Which, if you're a startup founder, describes basically everything you own.
Method 5: Qualitative Scoring, or, Turning Vibes into a Spreadsheet
Scoring methods assess patent value through structured ratings rather than producing actual dollar figures. The EPO's IPScore® framework evaluates 40 questions across five categories (legal status, technology, market conditions, finance, strategy) on a 1-to-5 scale. Commercial platforms like AcclaimIP generate composite P-Scores (0–99) incorporating citation metrics, prosecution quality, and legal strength. PatSnap, Clarivate, and LexisNexis PatentSight each offer proprietary quality indices.
The academic literature has identified 985 distinct indicators used in patent quality assessment—forward citations, family size, and number of claims appearing most frequently. These methods are genuinely useful for triaging large portfolios before investing in full valuations of the top assets. They cannot, however, tell you what a patent is worth. They can tell you which patents are worth finding out about.
Method 6: AI-Driven Valuation, or, Teaching the Machine to Price Ideas
And so we arrive at the future, which is already here, just unevenly distributed.[^2]
[^2]: William Gibson said that first. Credit where due.
AI and machine learning are increasingly deployed for patent valuation. The AI patent analytics market was valued at $1.07 billion in 2024 and is projected to reach $5.67 billion by 2034. These tools reduce prior-art search time by 60–80%, automate claim-to-product mapping, predict litigation outcomes, and enable continuous portfolio scoring rather than periodic snapshots. As of 2025, 68% of patent practitioners had adopted AI-assisted search tools, up from 31% in 2022.
Funded startups in this space include Patlytics ($14 million Series A), DeepIP ($15 million Series A), and Perplexity Patents (launched October 2025 with 160 million searchable documents). We are watching, in real time, the automation of judgment about what ideas are worth.
Whether that makes you excited or terrified probably says something about you.
III. What Actually Makes a Patent Valuable, or, The Factors That Separate a $100 Patent from a $100 Million Patent, Which Turns Out to Be a Staggeringly Large Number of Factors
The ceiling is set by technology and markets
The single most influential external driver of patent value is market demand for the underlying technology. This should be obvious but frequently isn't. A patent covering technology with high, growing demand commands enormously more than one in a declining field. Total addressable market, industry growth rate, and technology lifecycle stage all compound this effect. Patents are most valuable during the growth and early maturity phases of the technology S-curve, when adoption is accelerating and the patent's exclusionary power matters most.
Standard-essential patents (SEPs) occupy a special position. They have guaranteed markets because any implementer of the standard must license the technology. However, FRAND commitments constrain royalty rates. Courts have determined FRAND rates ranging from approximately 0.02–0.5% of average selling price per SEP using top-down allocation methods. Guaranteed demand, constrained price. The universe loves symmetry.
Legal strength determines whether value is more than theoretical
Claim breadth—the range of products falling within the literal scope of the claims—is a primary legal value driver. Enforceability depends on the prior art landscape, prosecution history, and post-grant challenge survival. File wrapper estoppel from prosecution amendments can significantly narrow effective claim scope under the doctrine of equivalents (Festo Corp. v. Shoketsu Kinzoku, 2002). Every amendment you make during prosecution is a concession you'll live with forever.
Patent family size is one of the most validated quantitative indicators of value in the academic literature—larger families signal the holder's conviction that the invention is worth protecting across jurisdictions. Post-grant survival rates provide a critical enforceability signal: once IPR is instituted, approximately 80% of patents have at least one claim found unpatentable, though 58% of patents survive post-grant proceedings overall when including non-instituted petitions. These are not good odds. But they're also not zero.
Commercial factors bridge the gap between rights and money
Whether a patent reads on commercial products—evidenced through claim charts mapping claims element-by-element to specific products—transforms theoretical value into actionable monetization. Design-around difficulty determines true blocking power: when the cost of redesigning exceeds the cost of licensing, the patent holder has leverage. Remaining patent life directly impacts DCF calculations. Existing licenses provide the most concrete evidence of value, because they are, literally, evidence of value.
At the portfolio level, synergies between patents create compound protection harder to design around than any individual patent. Geographic coverage across key markets multiplies enforcement potential. Cross-licensing leverage—using a portfolio as currency in mutual licensing negotiations—is a dominant strategy in semiconductors, telecom, and computing, where everybody infringes everybody else's patents and the only sane response is mutually assured licensing.
IV. Patents and Startups: An Empirical Love Story
Here is something that should probably be printed on the wall of every startup incubator in the country, possibly in neon:
The PitchBook Q1 2023 study, analyzing 46 million+ patent documents, found that startups with patents raised capital at significantly higher valuations across every single stage: 93% higher at angel stage, 17% higher at seed, 30% higher at early stage, and 51% higher at late stage. Early-stage patent startups raised 73% more capital, and their exit values averaged 2.1x larger.
Let that settle in. Ninety-three percent higher at angel stage.
The quasi-experimental "patent lottery" study by Farre-Mensa et al. (2017, NBER) went further. By comparing startups that received patents to otherwise-identical startups whose applications were randomly assigned to tougher examiners, it found that obtaining a first patent caused startups to grow 55% faster in employment and 80% faster in sales over five years, increased VC funding likelihood by 47%, and more than doubled IPO probability. The EPO/EUIPO reported that startups with patent applications are 6.4x more likely to receive seed funding.
Patents help. A lot.
But—and there is always a but—context matters enormously
In biotech and pharma, patents are not a nice-to-have. They are the business model. Nearly 50% of biotech startups file patents, and without IP protection, you can't raise money, because without exclusivity, you can't recoup the $1 billion+ it costs to bring a drug to market. Investors conduct deep clinical-trial-aligned IP due diligence.
In software, patenting rates are lower (~19% of software startups) due to Alice eligibility uncertainty and the greater importance of network effects, execution speed, and the hope that you'll achieve market dominance before anyone bothers to copy you. This strategy works right up until it doesn't.
In hardware and deep tech, patenting rates mirror biotech because manufacturing processes are easily copied and the lead time afforded by first-mover advantage is, by itself, not enough.
For early-stage startups, the IPEG Startup IP Valuation Model identifies four value dimensions: growth/pivot value (covering more ideas than your current product), liquidation value (collateral for lending), defensive value (protection against suits), and offensive/licensing value (future monetization). Each additional patent can add roughly $1 million in valuation in subsequent funding rounds. Even a pending application often suffices to check the "IP box" at seed stage. The patent doesn't need to be granted to change the calculus. It needs to exist as a credible signal.
The mistakes that will cost you everything
I'm going to be direct here, which is unusual for this particular prose style, but some things are too important for footnotes.
The gravest startup IP mistake is failing to properly assign IP from founders and contractors. If founders develop IP before incorporation, or before becoming employees, without written assignment agreements, the IP belongs to them personally—not the company. This has torpedoed funding rounds. This has torpedoed exits. It is the equivalent of building a house on land you forgot to buy.
Other frequent errors: public disclosure before filing (most countries outside the U.S. have no grace period), filing too narrowly or in the wrong jurisdictions, skipping freedom-to-operate analysis (a major investor red flag), and using unqualified tools or counsel to draft patents because you're trying to save $15,000 on the thing that might be worth $15 million.
The right moments for formal valuation: pre-fundraising (Series A and beyond), before M&A exits, when launching licensing programs, for IP-backed financing, and for tax compliance including 409A valuations.
V. How Patents Get Valued When Enormous Amounts of Money Are Changing Hands
The deals that reveal how the sausage gets priced
Nortel's patent portfolio—roughly 6,000 patents covering wireless, 4G/LTE, data networking, and semiconductors—sold at auction in 2011 for $4.5 billion. The company's average 5-year book value was approximately $1.2 billion. The winning consortium—Apple, Microsoft, BlackBerry, Sony, Ericsson, EMC—paid roughly $750,000 per patent and 3x pre-auction estimates. The strategic value of not letting Google have those patents turned out to be worth quite a lot.
Google, undeterred, then acquired Motorola Mobility for $12.5 billion, allocating $5.5 billion (45%) to patents and developed technology. Google later sold Motorola's hardware business to Lenovo for $2.91 billion while retaining the vast majority of the patents, implying an effective all-in cost of roughly $294,000–$303,000 per patent. Google paid a 63% premium over Motorola's market price, which is a sentence that should be tattooed on the forearm of anyone who thinks patents don't have strategic value.
The gap between what strategic buyers and financial buyers will pay for the same patent is vast. Strategic buyers pay premiums for defensive protection and cross-licensing leverage. Financial buyers—PE firms, patent assertion entities—focus on expected licensing and litigation revenue, apply higher discount rates, and arrive at lower figures because they lack the synergies that make patents part of an operating ecosystem. In distressed contexts, patent values can drop dramatically (Kodak's portfolio sold for far less than predicted), though competitive bidding among strategic buyers can produce outcomes that defy all pre-auction models.
How royalty rates actually get set
Royalty rates are established through comparable license analysis (preferred by courts and economists alike), income-based profit splits, or top-down/bottom-up allocation for SEPs. Common structures include percentage-of-revenue running royalties (typically 1–10%, with pharma higher and electronics lower), per-unit fees, lump sums, hybrids, and milestone-based payments.
For SEPs, courts have developed both top-down approaches (determine aggregate royalty burden for the entire standard, then apportion among patents) and bottom-up approaches. A persistent tension exists between market-negotiated rates and intrinsic technology value—one analysis found negotiated rates of approximately $20/unit versus intrinsic value of roughly $3/unit. The difference is lock-in premium. Once a standard is adopted, the switching costs make the patents worth more than the technology, considered in isolation, would suggest. This is either elegant market dynamics or a form of hold-up, depending on which side of the licensing table you're sitting on.
VI. How Courts Value Patents When People Are Suing Each Other, Which Is Surprisingly Often
The Georgia-Pacific Framework: Fifteen Factors Walk into a Courtroom
The foundation of reasonable royalty damages is the hypothetical negotiation framework from Georgia-Pacific Corp. v. U.S. Plywood Corp. (1970), which established 15 factors for determining what a willing licensor and licensee would have agreed upon. These factors cover established royalties, comparable licenses, license scope, the commercial relationship between the parties, the patent's commercial success, advantages over prior art, extent of infringing use, customary profit splits, apportionment between patented and unpatented features, and expert testimony.
The Federal Circuit has never described the factors as a "talisman," and instructing a jury on irrelevant factors is reversible error (Ericsson v. D-Link). The Federal Circuit Bar Association's Model Jury Instructions distilled the framework to just three factors. The Licensing Executives Society has documented the factors' diminishing role as courts increasingly favor direct comparable license analysis.
The factors are dying, in other words. But like many dying things, they refuse to leave the room entirely.
Apportionment: the actual battleground
If you retain one concept from this entire section, make it this: apportionment.
The smallest salable patent-practicing unit (SSPPU) doctrine requires that the royalty base for multi-component products be no larger than the smallest unit embodying the patented invention. VirnetX v. Cisco (2014) went further, requiring apportionment even within the SSPPU. The entire market value rule (EMVR) permits use of the full product as a royalty base only if the patented feature constitutes the basis for consumer demand—which is an extraordinarily high bar. And Uniloc v. Microsoft (2011) rendered the 25% rule of thumb inadmissible under Daubert, calling it a fundamentally flawed tool.
So: you can't just take 25% of the infringer's profits. You can't use the whole product's revenue as a base unless the patent drives consumer demand. You can't even use a component's revenue without further apportioning if the component has non-infringing features. What you can do is hire a very good expert and pray.
EcoFactor v. Google: the case that changed everything (2025)
The most consequential recent development in patent damages is EcoFactor v. Google (Federal Circuit, en banc, May 2025). The court, in an 8-2 decision, reversed a $20 million verdict and established the tightest standards yet for damages expert testimony. The holding: experts cannot rely solely on the patentee's unilateral assertions about what comparable licenses represent, particularly when deriving per-unit rates from prior lump-sum settlements. Combined with the 2023 amendments to Rule 702—which shifted the burden to the proponent of expert testimony—this signals a new era of scrutiny.
Daubert exclusion rates for patent damages experts reached approximately 50% in 2021, up from 40% in 2020. EcoFactor will likely accelerate this trend. Google has since petitioned for certiorari, arguing the Federal Circuit unconstitutionally displaced the jury's role.
And yet. Courts awarded a record $4.3 billion in patent damages in 2024, with 67 jury verdicts including 12 exceeding $100 million. Major verdicts included General Access Solutions v. Verizon ($847 million, later set aside), Personalized Media v. Amazon ($525 million), and Netlist v. Micron ($445 million). However, approximately 40% of damages appeals result in reversal at the Federal Circuit.
Record damages. Record reversals. The courts giveth, and the Federal Circuit taketh away.
VII. What In-House Counsel Should Actually Be Doing About All of This
From counting patents to measuring value
The era of measuring portfolio value by filing counts is over. Executives expect IP teams to demonstrate which patents protect revenue, which justify rising maintenance costs, and where freedom-to-operate risk lurks. The shift is from "how many patents do we have" to "what are they doing for us," which is the kind of question that sounds obvious until you try to answer it.
Best practice now involves a decision-first framework: consolidate and normalize the portfolio, segment by technology and product, score for value and risk, map to products and competitors, triage FTO and invalidity concerns, identify monetization candidates, and document decisions with audit trails. Frame IP in business terms for the C-suite: which products are protected, where is the company exposed, and what happens if a competitor files first. Interactive dashboards—showing technology coverage, competitive benchmarking, cost trends—transform IP from what the CFO sees as a cost center into a visible strategic function.
When you need outside help
External appraisers are necessary for M&A purchase price allocations, financial reporting and impairment testing, tax compliance and transfer pricing, board and investor presentations requiring independent opinions, and expert testimony. Internal assessment suffices for routine pruning, annual reviews, competitive monitoring, and preliminary budgeting. Professional valuations typically cost $10,000–$25,000 for boutique engagements to $100,000+ for complex Big 4 valuations and take 2–6 weeks.
Portfolio optimization in practice
AI tools can automatically identify abandonment candidates and calculate instant savings projections. Culture, not data, is typically the obstacle—nobody wants to be the person who recommended dropping the patent that turned out to be valuable. Strategic filing should be surgical in budget-constrained environments, guided by gap analysis revealing where the portfolio lacks coverage in areas where competitors are active.
Track outside counsel performance through scorecards covering 30+ prosecution metrics: allowance rates, cycle times, office action counts, costs. Push for fixed-fee arrangements. The days of paying $800/hour for someone to respond to an office action using the same template they used last time are ending, however slowly.
VIII. The Toolbox: Who Makes What and What It Costs
Analytics platforms
PatSnap dominates enterprise analytics with 2+ billion structured data points, AI-driven semantic search, and strong Asian market coverage.
Clarivate's Derwent Innovation builds on the Derwent World Patents Index (127 million+ patents, 900+ editors).
LexisNexis PatentSight is widely respected for its Patent Asset Index and IP valuation work. IPlytics leads in SEP analysis.
Orbit Intelligence (Questel) offers strong European capabilities with modular pricing. Free options—Google Patents, Espacenet, WIPO PATENTSCOPE—serve preliminary research.
Royalty rate databases
ktMINE (85,000+ cataloged rates, 25,000+ agreements) is the industry standard. RoyaltyStat specializes in transfer pricing analytics with 15,500+ unredacted agreements. RoyaltySource has 35+ years of tracking data.
IP management platforms
Anaqua provides enterprise lifecycle management with integrated AcclaimIP analytics covering 150 million+ documents. For smaller portfolios, cloud-based tools like IPfolio and Alt Legal offer simpler interfaces at lower price points.
Valuation service providers
Ocean Tomo (now part of J.S. Held) pioneered IP financial services—1,000+ engagements, IP worth $10 billion+, serving 80%+ of AmLaw 100 firms. Ocean Tomo also operates the Bid-Ask Market and IP transaction advisory services. Kroll provides independent analyses for financial reporting, M&A, and litigation. The Big 4 each maintain substantial IP practices. Boutique specialists include Black Stone IP, Stout Risius Ross, and EverEdge.
Patent marketplaces
Allied Security Trust (not-for-profit cooperative) has offered 20,000+ portfolios and spent $500 million+ acquiring patent rights.
RPX Corporation serves 320+ members with 40,000+ acquired assets. The secondary market shows sustained activity: Q1 2025 tracked 466 transactions encompassing 3,597 assets.
IX. The Forces That Are Reshaping Everything, or, Why 2025–2027 Might Be the Most Consequential Period for Patent Value in a Generation
Three bills that could rewrite the rules
The Patent Eligibility Restoration Act (PERA), reintroduced May 2025 with bipartisan support, would eliminate all judicial exceptions to patent eligibility—the entire Alice/Mayo framework—and replace them with specific statutory exclusions. If enacted, software, AI, diagnostic, and biotech patents would gain substantial value overnight. Overnight. Think about what that means for portfolio valuations.
The PREVAIL Act, which passed the Senate Judiciary Committee 11-10 in November 2024, would reform the PTAB by requiring standing for IPR petitioners and raising the evidence standard to "clear and convincing." This would make patents significantly harder to invalidate—and therefore more valuable.
The RESTORE Act would restore the presumption of injunctive relief for infringement, reversing the eBay v. MercExchange regime that made permanent injunctions rare for non-practicing entities.
None have passed. But the combined legislative momentum is the strongest in years.
The Unified Patent Court is making European patents worth more
The UPC filed 883 cases in its first two years (June 2023 – May 2025), including 320 infringement actions, with a roughly 60% patentee win rate and average disposition times of 13.9 months. Filings surged 120% in the second half of 2024 versus the first half. The February 2025 CJEU decision in BSH v. Electrolux confirmed UPC authority to issue cross-border injunctions. Unitary Patent uptake reached 38.9% of eligible applications in 2025, and Romania became the 18th participating member state.
Cheaper, faster, pan-European enforcement means European patent portfolios are measurably more valuable than they were three years ago.
Industry-specific dynamics are diverging sharply
In pharma, the Inflation Reduction Act is reshaping patent economics. CMS selected 15 Part D drugs for Round 2 price negotiations in January 2025 ($40.7 billion annual spending). The IRA's "pill penalty"—small molecules eligible for negotiation after 9 years versus 13 for biologics—has driven small-molecule funding down 70% since September 2021. Patent thickets remain pervasive: U.S. litigation involves 11–65 asserted patents per product.
In semiconductors, the CHIPS Act is driving a patent surge. Global semiconductor patent applications exceeded 81,000 in 2023 (22% increase year-over-year). The USPTO's Semiconductor Technology Pilot Program, launched December 2023, had received 251 petitions by December 2024.
In connected vehicles, the Avanci 5G patent pool now covers 75 licensors and 225 million licensed vehicles at $32/vehicle.
NPE activity surged 21.6% in 2024, with patent assertion entities now driving over 40% of SEP litigation and responsible for the vast majority of high-tech patent cases. Patent transaction volume is increasing even as prices face headwinds.
AI inventorship is settled. For now.
Every major jurisdiction has rejected AI as a named inventor following the DABUS/Thaler cases. The Federal Circuit affirmed in Thaler v. Vidal (2022), and the Supreme Court declined certiorari. The UK Supreme Court, EPO, German Federal Court of Justice, Japan IP High Court, and Canadian Patent Appeal Board all reached the same conclusion.
AI is not an inventor. Humans are inventors. The machine is a tool. This is the consensus.
For now.
X. What It All Means
Three things define this moment in patent valuation.
First, record litigation damages ($4.3 billion in 2024) paired with unprecedented appellate scrutiny (EcoFactor en banc) mean that the rigor of your valuation methodology determines whether your damages survive appeal. Sloppy work gets reversed. Good work holds.
Second, AI tools are collapsing the gap between portfolio-level screening and granular patent-by-patent assessment, making continuous valuation economically feasible for the first time. You no longer need to choose between knowing roughly what your whole portfolio is worth and knowing precisely what three patents are worth. You can start to know both.
Third, pending legislation (PERA, PREVAIL, RESTORE) and the maturing UPC could materially increase patent values within the next two years. Counsel and founders who understand their portfolios' worth under both current and reformed regimes will be the ones positioned to capture that upside.
For startup founders, the empirical evidence is unambiguous: patents increase valuations by 17–93% depending on stage, raise capital outcomes by 46–73%, and more than double IPO probability. But only well-executed IP strategies deliver these benefits. Bad assignment practices, narrow claims, missing FTO analysis—these destroy more value than patents create.
For in-house counsel, the shift is from cost-center management to strategic portfolio intelligence. The executives above you want answers to three questions: What's protected? Where are we exposed? What's it worth? If you can answer those, you are no longer a cost center. You are a competitive advantage.
The organizations that treat patent valuation as an ongoing capability—rather than a thing they do the week before a deal closes—will consistently capture more value from their most important intangible assets.[^3]
[^3]: Those non-things, it turns out, are everything.