Arrow's Paradox

Amnetic Blog: May 16, 2026

Arrow's Information Paradox

Arrow's Information Paradox: a buyer can't evaluate information or technology for sale without seeing it. But after seeing it, they don't need to buy it.

A firm, PharmaTech, has developed a technique in drug synthesis that could be widely applied in my industry. Competitors or non-competing drug makers could use it to make their own production more efficient. They don't want to just give this away, but they would gladly sell it for what it is worth to non-competing firms.

Imagine a non-competing firm, AcmePharma, would love to buy this technology at a good price. It would improve their processes and lower drug synthesis costs. A trade of money for the technique would make PharmaTech, AcmePharma, and everyone who buys their drugs, better off.

Three structural traps make this simple trade expensive or impossible.

The Discovery Problem

How does AcmePharma find out that PharmaTech has something interesting to sell? PharmaTech can't list significant details about their technique anywhere. There is no bulletin board or search engine over this kind of information, because being able to search over it would be the same as disclosing it.

The Disclosure Trap

If they knew the technique in full, AcmePharma would gladly buy the technology for $5 million as it saves them much more. But, how can AcmePharma know this? If PharmaTech gives them the details of the technique and data to back it up, then AcmePharma has no incentive to pay.

The Free Look Trap

Even if AcmePharma is upstanding, if PharmaTech gives up the technique but AcmePharma decides the price isn't worth it for them, PharmaTech just gave away the details for free.

Each of these traps has the same root in Arrow's Paradox: information's value to the buyer can only be assessed by fully revealing it, but fully revealing it effectively gives it away for free.

The Costs of Arrow's Paradox

The problem PharmaTech and AcmePharma had isn't just a pharmaceuticals problem, it's a general problem across all industries, across all of the world economy.

Ultimately, Arrow's Paradox results in a tremendous economic loss: valuable information exists all around the world for which buyers exist who would like to buy it, and sellers exist who would sell it, but the transaction doesn't happen. This is a fundamental market failure, one that has been with us for as long as humans have had information to buy and sell.

Because of the tremendous value of these transactions, our society has invented workarounds, expensive workarounds. They are expensive for companies to navigate and expensive for society to maintain, and they're fundamentally inefficient even when they work. These workarounds are:

1. Patent Process
2. Lengthy procurement processes
3. Grants and Open Science

Most of how our economy handles information markets is a workaround to Arrow's Paradox.

Patents

Patents allow for a disclosure for a time-limited, government-enforced monopoly. The seller reveals everything in exchange for the right to sue anyone who uses it without licensing it. This works for protecting a narrow slice of information: novel processes that are describable in claims. It costs society the deadweight loss of monopoly pricing for twenty years, plus the legal infrastructure to litigate it. It also entirely excludes the largest category of valuable commercial information: trade secrets, know-how, processes, and data that either can't be patented or would be trivially designed around.

Lengthy Procurement Process

The bulletin board sale is impossible because of Arrow's Paradox, so firms need to establish slow, expensive, relationship-based transactions: NDAs with teeth, staged disclosures, escrow arrangements, trusted intermediaries, data rooms, on-site audits, sales teams and customer success orgs. These mechanisms partially work, especially when the information is extremely high value, but they're slow (months to years), expensive (legal fees, sales teams, diligence teams, executive time), and they still leak, because the buyer's diligence team is itself a disclosure surface. They also bias the market toward large, established players who can afford the friction, excluding smaller buyers and sellers entirely.

Grants and Open Science

These route around the market by having public or non-profit institutions like governments, foundations, and universities pay for the information up front and then give it away. This produces enormous value but costs taxpayers and donors directly, ties the research agenda to political and institutional priorities rather than market demand, and produces information that no one has a private incentive to commercialize, distribute, or maintain once the grant ends.

What if we solved Arrow's Paradox?

Imagine the marketplace that would exist if Arrow's Paradox were solved. AcmePharma is always on the lookout for new techniques, and searches for synthesis improvements the way they'd search for any other input. They would quickly find a full description of a new technology they could use, and how it works, and could fully evaluate that it would save them roughly $7M per year. They pay. They receive it. PharmaTech captures the value of work they otherwise couldn't sell. The whole transaction takes days, or even minutes, not years, and doesn't require either firm to staff a procurement team.

Now scale that across every industry. Research that isn't commercially viable today because the discoverer can't capture its value starts getting funded. Small firms with valuable know-how can sell to large firms without being acquired. Knowledge moves efficiently to where it creates the most value.

At Amnetic, we've solved Arrow's Paradox. At Amnetic, we're building the marketplace.