A new study by the Bank for International Settlements (BIS) found that Bitcoin on-chain transfer estimates can vary by up to sixfold, depending on the measurement method.
The BIS Working Paper No. 1377, Hidden by complexity? Measuring stablecoin, crypto and decentralised finance ecosystems, published on 15 September 2026, examined how the structure and complexity of blockchain data can affect estimates of economic activity. The researchers found that monthly estimates of Bitcoin transfer values could differ by as much as sixfold depending on how transaction outputs were accounted for.
“The resulting estimates of transfer volume can differ by a factor of six, depending on the methodology used,” the researchers said.
The study examined how blockchain data can be affected by the scale, fragmentation and technical complexity of crypto networks. While on-chain data is publicly available and transparent, the BIS said its economic meaning is not always straightforward.
The issue stems partly from Bitcoin’s unspent transaction output (UTXO) model. A Bitcoin transaction can send funds to a recipient while returning unused funds to the sender as change, making it difficult to determine how much value represents an economic transfer.
Three ways to measure transfers
The researchers compared three approaches to estimating Bitcoin transfer values.
The broadest measure counts the value of all transaction outputs. An adjusted measure removes outputs sent back to the sending address, treating them as likely change. A more conservative estimate attempts to remove self-transfers and other outputs that may not represent an economic transaction.
These approaches produced substantially different results. The researchers noted that the conservative estimate is itself a heuristic rather than a definitive measure, as CoinJoin transactions, mixers, spam, and intermediaries can make the underlying economic activity difficult to identify.
The study examined Bitcoin data covering 2009 through 2026, including about 1.3 billion transactions and 3.6 billion transaction outputs.
Some analytics providers already adjust blockchain data to reflect actual economic activity better. Visa’s Onchain Analytics dashboard, powered by Allium Labs, shows both total and adjusted stablecoin volumes. Its adjusted data removes activity linked to bots, high-frequency trading, bridge routing, and internal exchange transfers.
Over the past 30 days, the dashboard shows $6.1 trillion in total stablecoin volume compared with $306.1 billion in adjusted volume.
Bitcoin market cap faces similar issues
The BIS also highlighted challenges in measuring Bitcoin’s market capitalization.
Traditional market capitalization applies the current Bitcoin price to the entire outstanding supply. However, some coins have remained inactive for many years, and inactivity alone does not establish whether their associated private keys have been permanently lost.
The researchers found that more than 1.8 million BTC had remained untouched for over 15 years, although some long-dormant coins have moved after more than a decade.
The study also compared conventional market capitalization with realized capitalization, which values Bitcoin outputs based on the price when they last moved. Conventional market capitalization reached up to four times realized capitalization during periods of rapid price increases, according to the research.
As per CoinGecko data, Bitcoin was trading at around $75,795, with a market capitalization of approximately $1.52 trillion. Its circulating supply stood at about 20.09 million BTC, while 24-hour trading volume was around $39.3 billion.
Ethereum and Stablecoins add complexity
The measurement challenges extend beyond Bitcoin.
On Ethereum, the BIS researchers identified 67.5 million active deployed contracts, including 54.2 million that remained uncategorised. The study also found widespread duplication and symbol reuse among tokens, creating additional difficulties when measuring activity.
Stablecoin activity also varied significantly between blockchains. USDT held in Ethereum smart contracts accounted for more than 20% of supply at its peak and generally remained between 15% and 20% before falling to roughly 10%–15%. On Tron, smart-contract holdings remained around 1% for much of the period studied.
The researchers said the difference suggests that USDT can serve different economic roles depending on the blockchain. Ethereum activity appeared more closely associated with DeFi collateral and liquidity, while Tron activity appeared more transactional and store-of-value oriented.
BIS calls for more cautious metrics
The findings underline a broader issue with blockchain analytics: publicly available data does not necessarily provide a straightforward measure of economic activity.
“On-chain indicators should be treated as noisy approximations of economic activity, rather than precise measures,” the BIS researchers said.
The researchers recommend using ranges rather than relying solely on single-point estimates, while also making the assumptions and technical classifications behind those estimates explicit.
The paper does not introduce new regulation or reporting requirements. Instead, it focuses on the methodological challenges involved in interpreting blockchain data.
The research, titled Hidden by complexity? Measuring stablecoin, crypto and decentralised finance ecosystems, was authored by Timothy Aerts, Ronald Heijmans, Jan Paulick and Violeta Vuletic. The BIS notes that the views expressed in its working papers are those of the authors and do not necessarily represent the views of the institution or its member central banks.
The study highlights an ongoing challenge for blockchain researchers and analytics providers: developing more consistent methods to distinguish economic activity from technical transactions, internal movements and other on-chain noise.
Also Read: Bitcoin Mining Sector Faces Tightest Margins Since the 2024 Halving
