Back to NewsWhy Did IPXI Fail? The Patent Exchange Post-MortemNovember 2026

Why Did IPXI Fail? The Patent Exchange Post-Mortem

Why did IPXI fail? The short version: the world's first exchange for patent licences shut its doors on 23 March 2015 because its flagship offering could not find enough buyers to even set a price. Its CEO blamed the buyers. The people who studied the wreckage blamed the product. Both are partly right, and the part they agree on is the lesson that still matters for anyone trying to make patents liquid.

This is the post-mortem with both sides on the table, scored, and translated into what it means for patent owners and investors now.

The short answer

On the day it closed, CEO Gerard Pannekoek wrote that IPXI "was not able to secure the cooperation of enough willing licensees to fully realize its concept," and that potential licensees "made it clear that the only way IPXI would really get their attention was through litigation".

The most thorough independent analysis reads the same facts differently. Law professor Jorge Contreras concluded that a lack of participation by key patent holders, an untested record of enforcing patents, and the constraints of its standardised licence led to IPXI's demise, and that its 802.11n offering drew too little interest even to be priced.

So: one side says the market refused to behave. The other says the exchange offered something the market had no reason to buy.

What IPXI was trying to build

IPXI was formed in 2008 with a big idea borrowed from commodities trading. Instead of negotiating every patent licence bilaterally, behind closed doors, for months, why not standardise the licence and let it trade?

Unit License Rights, explained

The instrument was the Unit License Right, or ULR: a standardised, non-exclusive licence sold up front and tradable on the exchange. In the flagship offering, one ULR cost USD 120 and let the buyer make and sell 1,000 Wi-Fi 802.11n devices. A manufacturer that bought too many could resell the rest. One that bought too few could buy more on the open market. Price discovery would, in theory, settle the argument about what a "reasonable" royalty is for standard essential patents.

Who backed it

This was not a fringe project. Ocean Tomo described itself as the original inspiration for the platform, and IPXI brought three patent estates to market. The model garnered support from Ford, JP Morgan Chase, Sony and research institutions, and the closing notice thanked more than 70 corporate, university, laboratory and associate members.

It still lasted only six years of trying, by IAM's count, and the closure came just six months after the 802.11n offering began marketing.

IPXI's own explanation: nobody pays without a lawsuit

The CEO's case is simple and it is not stupid. IPXI said it had done the hard work for the buyer in advance: evidence of use, quality, pricing, transparent terms. In an interview Pannekoek said they "could not get a significant number of entities to be good corporate citizens and respond to legitimate licensing offers where we had done 90% of the work in advance for them." He went further and blamed a US patent system that offers no incentive for good-faith licensing.

There is real economics under that complaint. A patent is worth what its owner can credibly enforce. If nobody behind the offering has ever sued anyone, an implementer acting rationally waits. Contreras concedes the point: IPXI had no history of enforcement and no announced plan to litigate, so a "wait and see" stance was the rational one for manufacturers.

A postscript put it bluntly: after more than five years of fits and starts, the transparent marketplace gave way to the litigation-driven status quo.

The other explanation: a product the buyers did not need

The academic post-mortems accept the enforcement point and then list everything else. Read together, it is a long list.

A partial portfolio from sponsors with no enforcement history

The 802.11n ULR covered 194 patents, 142 of them deemed essential, from eight sponsors. That sounds substantial until you set it against the thousands of patents estimated to read on the 802.11 family. A manufacturer still had to negotiate with everyone else, so the ULR solved a fraction of the problem. And none of the eight sponsors had an active record of licensing or enforcing in 802.11.

No room for cross-licensing or non-SEPs

In real technology markets, cross-licensing is how big players pay: they trade their own patents as currency, often netting to zero royalty. A fixed-price ULR made that impossible. It also covered only standard-essential patents, while the licence a manufacturer actually wants usually bundles essential and non-essential patents from the same holder.

Artificial scarcity in a non-depletable asset

Wheat runs out. Licences do not: a patent owner can grant one more at zero marginal cost. To make secondary trading work, IPXI had to cap supply. The first tranche was 50,000 ULRs, enough for 50 million chipsets against a lifetime projection of 8 billion. A buyer could reasonably fear that a scarce, tradable licence would get more expensive the more the industry needed it, the opposite of what FRAND licensing is supposed to deliver.

"Patent lawyers aren't commodities traders"

That line, from one of Contreras's interviews, may be the most important sentence in the whole record. In-house licensing teams were asked to guess how many units to buy, accept loss if they guessed wrong, and adopt a mechanism with no precedent in their field. Commodity exchanges grew out of how buyers and sellers already traded. IPXI was imposed from the outside.

Bad timing: Alice and the PTAB

The 802.11n offering went to market a few months after the US Supreme Court's 2014 Alice decision cast doubt on many software patents, and while the new Patent Trial and Appeal Board was invalidating patents at pace. Paying up front for a basket of 194 patents of uncertain durability was a hard sell in that mood.

Scoring both explanations

CauseArgued byEvidenceVerdict
Licensees only pay under threat of litigationIPXI's CEONo enforcement history, wait-and-see was rationalTrue, but it describes every licensing market, and bilateral deals still close every day
Portfolio too partial, sponsors not key playersContreras194 of thousands of 802.11 patents, 8 sponsors with no licensing recordStrong: the ULR did not remove the buyer's real workload
Standard terms blocked cross-licensing and bundlingContrerasFixed price, SEP-only packageStrong: it priced out how large players actually pay
Capped supply in a non-scarce assetContreras50,000 ULRs vs 8 billion projected chipsetsMedium: a structural risk buyers could see
Cultural mismatchContreras, Steele"Patent lawyers aren't commodities traders"Strong: no buyer pull, only seller push
TimingContrerasAlice (2014), PTAB invalidationsContributing, not decisive

Strip the rows down and they share a root. IPXI built an efficient venue for sellers and assumed buyers would arrive because the venue was efficient. They had no reason to. The CEO was describing the symptom; the law reviews were describing why the cure did not take.

Same failure, different technology: IPwe

If the diagnosis is right, swapping the technology should not change the outcome. It did not. In 2021 one commentator argued that IPXI's big problem was a lack of predictability on both sides of the transaction, and that blockchain would have solved it.

Then the experiment was run. IPwe, the blockchain patent platform backed by IBM, picked up where IPXI left off and, like its predecessor, found implementers largely unwilling to buy or license through the platform, leaving the marketplace one-sided with sellers but not enough willing buyers. We traced that story in detail in what is actually real about tokenized patents.

Two attempts, nine years apart, one with a commodities rulebook and one with a ledger. Same missing side of the market.

What IPXI teaches about patent liquidity today

For patent owners. A listing is not a sale. Putting rights in front of the market, however transparently, does not create a buyer. Patents move when someone does the work of matching a specific asset to a specific buyer's specific gap, and can show the buyer why it reads on what they make. That is what a patent broker is for, and it is why the M&A model has outlived every exchange.

For investors. Be careful with any pitch that treats patents as a commodity. They are not fungible, demand for each one is narrow, and liquidity is something an operator earns transaction by transaction, not something a venue confers. Patent economics can be real and still be illiquid, and both facts belong in the same sentence.

What we took from it. We run a patent monetization platform, so we have an interest in this question and say so. We built the buyer side first. Before an asset reaches a buyer it goes through a validation gate of physics-grade claim reconstruction, detectability analysis and validity probability, benchmarked at 0.76 percent mean absolute percentage error against real-world outcomes. Detectability is the lesson from IPXI in one word: a claim nobody can show is being practised is a claim nobody prices, and nobody fears. Our brokerage is success-based, with commission paid by the seller on completion, minimum EUR 5,000, which keeps us paid for closed deals rather than for listings.

We also state our own stage plainly. Our published traction shows 24 patent families filed or in progress, around 100 identified and validated in the pipeline, and tokenization tested end to end only inside a private sandbox. That is not a public exchange, and after IPXI and IPwe nobody should let it sound like one.

Where investors get exposure to patent family economics through us, it is via profit-participating loan notes issued by EX Financial Solutions Ltd under English law, prospectus-exempt, not reviewed or approved by any regulator, directed exclusively at persons outside the United Kingdom and not available to U.S. persons or retail investors resident in Indonesia. Returns are not guaranteed and total loss is possible. This article is general information about patent market history, not investment advice and not an offer.

IPXI identified a real inefficiency. It failed because it fixed the seller's problem and assumed the buyer's would fix itself. Any patent liquidity venue built since has to answer the question IPXI could not: why will the buyer show up?

FAQ

When did IPXI shut down? IPXI ceased operations on 23 March 2015, effective the same day, when it also withdrew its 802.11n licence offering. The announcement came from its president and CEO, Gerard Pannekoek.

What was a Unit License Right (ULR)? A standardised, non-exclusive patent licence designed to trade on an exchange. In IPXI's 802.11n offering one ULR cost USD 120 and covered the manufacture and sale of 1,000 compliant devices, and unused ULRs could be resold.

Who was behind IPXI? Ocean Tomo described itself as the original inspiration, and former Ocean Tomo colleagues were on the IPXI team. Supporters included Ford, JP Morgan Chase, Sony and research institutions, plus more than 70 members at closing.

Would blockchain have saved IPXI? It was argued in 2021 that it would have fixed the information problem. IPwe then tried a blockchain patent marketplace and hit the same wall: plenty of sellers, too few willing buyers.

We use essential cookies to operate this website and, with your consent, optional cookies to understand site usage. You can accept or decline non-essential cookies at any time. See our Impressum for our contact and legal details.