The full paper, “How Verifiable Trust Is Produced: From Verifiability to the Public Trust Root and a Theory of Trust Productivity,” is available on Chainless (chainless.hk). Given the space limit here, this article offers a plain-language introduction to its core ideas.
The paper makes a first attempt to bring questions that have long been studied separately—credit, auditing, monitoring, verification, blockchains, and productivity—into a single accounting framework. It asks a more fundamental question: how is trust actually produced?
Human societies relied on credit and trust long before modern economics existed. Banks, companies, governments, law, audit, and regulation all help answer the same practical question: can a critical fact stated by a person or institution be trusted enough for others to act on it?
If a bank says it owns assets worth RMB 10 billion, that statement does not automatically become trustworthy. The bank must keep records, maintain internal controls, accept managerial responsibility, undergo audit and regulatory review, and often face further scrutiny from rating agencies, counterparties, and investors. Only through these institutional processes does a claim that outsiders cannot directly verify become a fact on which society is willing to act.
The “trust” discussed here is not general emotional trust, reputation, or goodwill. It refers to a critical fact that has passed an appropriate verification process and has become reliable enough for outside parties to make decisions with economic, legal, or accountability consequences.
This points to something that economics has rarely treated as a separate production problem: trust itself consumes real resources to produce.
Not every cost borne by banks, auditors, or regulators should be counted as a trust-production cost. The paper asks a counterfactual question: if a critical fact could already be verified directly, continuously, and independently, which resources devoted to confirming that fact would materially fall? That avoidable portion is the cost boundary that matters here.
Once there are inputs, there are costs. Once reliability varies, there is quality. Once verified facts are actually used, there is output. And when the same independently reviewable proof is relied on by multiple parties in transactions, contracts, regulation, or accountability relationships, reuse and potential economies of scale appear. “Output” here does not mean how much data a system processes or how many API calls it receives; it means how much verified information is actually relied on in the real world.
That leads to the central question of the paper: if economists study labor productivity, why should we not study the productivity of trust production?
Where did this idea come from?
It did not arise in a vacuum. Our work on cryptocurrencies, transparent banking, Chainless systems, and verifiable finance gradually led to it. More than a decade of cryptocurrency practice, beginning with Bitcoin, revealed a new institutional possibility.
Traditionally, much credit comes from trusting a person, an institution, or a legal arrangement. Bitcoin showed that some critical facts can instead gain credibility through public rules, cryptographic proof, and independent verification that does not depend on a single central institution’s promise.
Satoshi Nakamoto set out to solve the problems of issuing and transferring electronic cash without double spending. In doing so, Bitcoin also demonstrated something more fundamental: for some classes of facts, trust can move from “believing an institution” toward “verifying the facts.” This may be one of the deepest and most underappreciated lessons of cryptocurrency.
Cryptography, hashing, proof of work, and digital cash were not all invented by Satoshi. The breakthrough came from combining existing ideas and technologies into an institutional result that had not existed before.
Theory often advances in the same way. We do not claim to have invented the ideas of trust, verification, auditing, or productivity. Each has a substantial literature. What was missing was a unified question: how many real social resources are spent turning facts into states that other people and institutions can confidently rely on?
The formal paper breaks this question into the cost, quality, effective output, reuse, and shared infrastructure of trust production, and puts them into one accounting framework. It contains mathematical expressions, but the mathematics is not the point. Its role is to make clear what should be counted, what should not, and under what conditions comparisons are meaningful.
This is not a law of gravity
The word “quantitative” can create a false expectation that the paper should immediately produce one precise and permanent number. That is not how economic measurement works. GDP, labor productivity, and capital stock all depend on definitions, classification, prices, quality adjustment, and continuing revision. They are not immutable physical constants, yet they remain indispensable tools for economic analysis.
Trust productivity should be approached in the same spirit. The first task is not to announce a final number, but to build a ruler. What counts as an input to trust production? What counts as genuine output? Can two systems be compared if they deliver different levels of reliability? What is the economic difference between one verified fact being used by ten institutions and ten institutions each re-verifying the same fact? If thousands of systems share the same public trust root, how should its common costs be allocated?
Sharing does not automatically mean lower cost. If security, governance, integration, capacity, and failure-risk costs rise faster than reuse, a public trust root may lose its productivity advantage. The paper treats scale economies as a condition to be tested with real-world data, not as a conclusion assumed in advance.
Once these questions are defined clearly, trust begins to move from a largely qualitative social relation into an economic object whose cost, quality, output, reuse, and productivity can be compared. The paper explicitly allows proxies, ranges, and iterative refinement as better data become available.
How might verifiable systems change trust productivity?
Traditional systems often confirm facts after the fact, through internal controls, sampling, audits, supervision, and legal responsibility. Verifiable finance points toward another possibility: embedding verification into the process by which the facts are created.
Consider stablecoin reserves. The real question is not whether an issuer can publish a snapshot saying, “we have this much in reserve.” A stronger structure would let us trace the process: did the customer’s money actually enter the bank? Were matching reserve assets purchased? How many tokens were issued? Did assets and liabilities remain continuously aligned? After redemption, did both reserves and token liabilities decline accordingly? Can the entire process later be replayed and independently checked?
If the underlying inputs are truthful and the scope of verification is comparable, continuous, comprehensive, replayable verification differs structurally from periodic sampling in its time coverage, process coverage, reviewability, and speed of anomaly detection.
But verification has clear limits. It cannot by itself guarantee that real-world inputs were truthful, cannot eliminate future risk merely because past facts are verifiable, and cannot replace law or ultimate human judgment. What verifiability changes is how facts are confirmed, preserved, reviewed, and reused—not whether technology can eliminate all uncertainty.
The paper therefore does not claim that verifiable systems must be cheaper. The size of any cost advantage has to be established with real implementation data. The USDC example in the paper is only a “partially identified account”: it shows where the accounting can begin, not the final answer. Public disclosures reveal only part of the traditional system’s external costs, while complete all-in cost data for both the traditional and verifiable approaches are still unavailable.
A research program that has only begun
This paper was developed over a short period, but it grew out of years of work on cryptocurrency, transparent banking, Chainless systems, and verifiable finance. It did not begin with a grand theory looking for confirming evidence. It emerged by repeatedly asking a deeper question.
Why do we want transparency? Because critical facts need to be verified. Why do we verify them? Because verification can create trust. Why should trust be studied separately? Because producing it consumes real social resources. Once there are inputs, quality, output, and reuse, a productivity question naturally follows.
That is how the idea of “trust productivity” emerged. The value of the paper is not that it has solved every measurement problem. No new economic accounting system becomes mature in a few days. Its contribution is to formulate a calculation task that has rarely been posed in complete form: how many real resources are required to produce one unit of sufficiently reliable trust that society actually uses?
If that question proves useful, the research can continue for decades. Different industries can build their own trust-production accounts. Data can improve. Variables that cannot yet be observed directly can be approximated with better proxies. Competing trust-production architectures can be compared over time at comparable quality levels.
That is how theories normally develop: later researchers stand on earlier work, recombine dispersed insights, and push the inquiry one layer deeper. No theory begins fully finished. Bitcoin did not. Economics did not. Verifiable finance will not.
So, can trust be measured? Our answer today is: we can begin by building a unified, comparable, reproducible, and continuously improvable accounting framework. More precise numbers will have to be earned gradually through data and practice.
