Same Blockchain, Different Numbers: The Case for Standards in Blockchain Analytics

What a central bank working paper reveals about the hidden assumptions behind on-chain data

Blockchain data is public. Anyone can inspect every transaction. Yet a new working paper from De Nederlandsche Bank (DNB), written with authors from the Bank for International Settlements (BIS), shows that two analysts can start from the same ledger and reach very different conclusions. For the blockchain analytics and intelligence community, this is more than an academic point. It is an argument for common standards.

The evidence: one ledger, many answers

Using 100 billion records from Bitcoin, Ethereum and Tron, the authors show how much headline indicators depend on method:

• Transfer values. Bitcoin transfer values vary by up to a factor of six, depending on how change outputs are treated. The highest figure is also the one most widely reported.
• Market capitalisation. The simple measure can be up to four times the realised capitalisation. Around 1.8 million bitcoins have not moved in over 15 years and are likely lost.
• Smart contracts. Of 67.5 million active contracts on Ethereum, over 54 million cannot be classified under any technical standard. The symbol “USDT” appears in roughly 7,000 token contracts, while only one is Tether’s.
• Stablecoins across chains. The same token plays different roles on different chains: smart contracts have held over 20% of USDT on Ethereum at times, against around 1% on Tron. Adding the two together hides the difference.
• DeFi metrics. The paper notes that total value locked (TVL), a widely cited indicator, is prone to inconsistent methodologies and even manipulation.

The authors also observe that the methodology behind many indicators is not always clear, and that many are built for trading rather than for measuring economic activity.

Why standards are necessary

Each of these problems has the same root: the numbers are produced without a shared way to describe how they were built. In practice, that creates three risks.

1. Results cannot be compared. Two reports on “Bitcoin activity” or “stablecoin volume” may use different definitions and still look equivalent.
2. Results cannot be verified. Without explicit assumptions, a reader cannot tell which part of a figure is observation and which is heuristic.
3. Decisions rest on noise. The authors conclude that on-chain indicators are noisy approximations rather than direct measures of economic activity, and that naive aggregation can misinform research and policy assessments.

The same applies to blockchain intelligence, which investigators, prosecutors, supervisors and researchers rely on in high-stakes settings such as criminal proceedings and supervisory enforcement. Attributing activity to an actor, labelling an address or clustering addresses all involve assumptions, and conclusions that inform such decisions should be traceable to them.

What the paper points to

Read together, the authors’ recommendations outline what a professional standard would need to cover:

• Bounded estimates with explicit assumptions, reported as ranges rather than single figures.
• Technical classification of contracts and assets, combined with expert judgement, because names and labels are easy to imitate.
• Chain-aware reporting, distinguishing infrastructure and asset identity instead of aggregating by token.
• Links between on-chain assets and legal issuers, which the paper says requires off-chain information, with registries or technical verification as possible tools. The authors also stress that this must be balanced against the privacy features of public ledgers.

From principles to practice: OBIS

These principles are already being turned into practice. Open Blockchain Intelligence Standards (OBIS), promoted by the Blockchain Intelligence Professionals Association (BIPA), is an open community effort developing vendor-neutral standards for blockchain intelligence. It starts from the same problem: data formats vary between platforms, the same concepts carry different names, and foundational computations such as address clustering follow divergent methods, so findings produced in one tool ecosystem are hard to verify or reuse in another.

OBIS works in public, with a process modelled on the IETF and W3C: open drafts, defined review windows and public responses to comments. Its first work items are a document lifecycle (OBIS-0001), a shared vocabulary for blockchain intelligence (OBIS-0002), and an attribution tag data model and exchange format (OBIS-0003). Three focus areas anchor the work: investigations, forensics and attribution; regulatory reporting and supervision; and open analytics and research data.

A key design principle is that openness applies to formats, vocabularies and methods, not to intelligence itself. A specification defines how an attribution record is structured or what a named clustering method computes. It does not require anyone to disclose which addresses belong to whom, or which cases are under investigation. That is also what makes the work defensible: a court, a defence expert or a peer reviewer can assess a finding only against a method whose definition is public.

The DNB paper itself follows up on Project Atlas, a collaboration between the BIS Innovation Hub Eurosystem Centre, De Nederlandsche Bank and the Deutsche Bundesbank. Published in March 2026, Atlas created a data platform tailored to the needs of central banks and financial regulators. It combines on-chain data from public blockchains with off-chain data from crypto exchanges, so that users can vet the data and assess these markets more accurately. Atlas showed the value of connecting data sources; OBIS addresses the challenges underlined by the paper, through the shared formats, vocabularies and methods that make such connections repeatable.

Why now

The timing is favourable. The stablecoin market has grown to over USD 300 billion, according to industry estimates cited in the DNB paper. Policymakers and central banks increasingly need reliable data for financial stability monitoring, payments oversight and monetary analysis. And public institutions are already building their own data infrastructure: the paper’s analysis relies on DNB’s Mercurius platform, developed with the BIS Innovation Hub and Deutsche Bundesbank.

Next step: a roundtable in Frankfurt on 5 November 2026

BIPA, as promoter of OBIS, is co-organising a roundtable in Frankfurt on 5 November 2026, together with the Frankfurt School of Finance and Management and the Complexity Science Hub. It will bring together experts from academia, industry and policy for an open discussion among financial authorities and regulators on interoperability in blockchain intelligence, with supervisory and regulatory needs as the guiding reference point.

The roundtable builds on a series of consultations: a workshop at the 9th Global Conference on Criminal Finances and Cryptoassets in Vienna, a breakout session at the 10th Global Conference in Luxembourg on 15 September, and a roundtable with the INTERPOL New Technologies Forum community in Vienna on 1 October. In Frankfurt, participants will:

• build a shared view of the interoperability challenges, grounded in concrete supervisory and investigative use cases;
• review the early outcomes of OBIS alongside related supervisory practice and regulatory developments;
• identify operational, legal and technical requirements, with particular attention to supervisory and regulatory needs.

Anonymised outcomes will be prepared, shared with participants and coordinated with parallel initiatives.

How to get involved

OBIS is open to anyone. Drafts are developed publicly on GitHub, where anyone can follow the discussions, comment and contribute, with no application or membership required. Institutions that cannot use public workflows, such as supervisors, financial intelligence units and government agencies, can send written commentary by email. Tool providers, authorities, researchers and reporting entities all have a role to play. More at obistandards.org.

The bottom line

Public data does not guarantee comparable conclusions. Standards for definitions, classification and the disclosure of assumptions would make blockchain analytics more reliable, more comparable and more useful for the decisions that depend on it. That is the work BIPA is committed to advancing together with its partners and the wider community.

Source: Aerts, Heijmans, Paulick and Vuletić, “Hidden by complexity? Measuring stablecoin, crypto and decentralised finance ecosystems”, DNB Working Paper No. 870, September 2026. The views expressed in the paper are those of its authors.