BIS Study Reveals Sixfold Discrepancy in Bitcoin Transfer Estimates

According to a study conducted by the Bank for International Settlements (BIS), the estimated volume of Bitcoin transfers can vary by as much as six times based on the methodology used for calculation.

This analysis focuses on on-chain transfers rather than the trading volume of Bitcoin on exchanges. The discrepancies arise from the structural characteristics of the blockchain and the criteria for determining what constitutes an actual transfer of funds between participants.

The researchers examined approximately 100 billion records from Bitcoin, Ethereum, and TRON networks. They identified three primary challenges related to measurement:

• Calculating transfers in blockchains that utilize the UTXO model;
• Classifying smart contract activities;
• Comparing identical assets across different blockchains.

Change Returns Inflate Transfer Volumes

In the Bitcoin network, balances are comprised of unspent transaction outputs (UTXOs). When a transfer is made, the entire output is consumed, with any remaining amount typically returned to the sender as change.

For instance, if a user spends a UTXO worth 4 BTC to transfer 1.5 BTC, they receive 2.5 BTC back. However, the transaction data does not always clearly indicate which output was intended for the recipient and which represents change.

Simply adding all outputs can inadvertently inflate transfer volumes by including these returns. The researchers compared various counting methods: one without adjustments, one excluding transfers to the sender's address, and one with a more conservative estimate of change.

Depending on the chosen method, the final figures can differ by as much as six times.

Estimates of Bitcoin transfer volumes using different calculation methods. Source: BIS.

“Metrics such as transaction volume, market capitalization, and total value locked often create an impression of precision that is not supported by the underlying data,” the authors noted.

The methodology also impacts the estimation of Bitcoin's capitalization. The standard metric is calculated by multiplying the total supply by the current price. The researchers compared this with an estimate that excludes presumed lost coins and a realized capitalization that values each UTXO at the price during its last movement.

During periods of rapid price increases, the standard capitalization could reach four times that of the realized capitalization. The authors emphasized that the latter is neither a substitute for market capitalization nor a strict lower boundary; they reflect different assessment methods.

Classification of Most Smart Contracts Lacking

Another issue pertains to Ethereum. Of the 67.5 million active smart contracts, researchers were unable to categorize over 54 million into predefined technical categories. Approximately 11.8 million classified contracts turned out to be proxy contracts, while 1.4 million were ERC-20 standard interchangeable token contracts.

The free use of names and tickers adds additional "noise". For example, the designation USDT was found in about 6,900 token contracts. The genuine stablecoin is issued through a specific Tether contract, but any developer can assign the same ticker to another asset.

The authors clarified that a name match does not necessarily imply fraud or a lack of economic value for the token.

They also discovered significant differences in USDT usage across blockchains. In Ethereum, the share of "stablecoins" held on smart contract addresses exceeded 20% in 2022, while in TRON, it remained around 1% for most of the time.

The authors linked the higher share in Ethereum to USDT's use in DeFi, such as for liquidity provision and collateral. In contrast, TRON's activity aligns more with transfers and value storage. However, the researchers noted that the type of address only allows for a rough assessment of the economic purpose of the operation.

Therefore, simply aggregating USDT activity across different blockchains may combine fundamentally different types of economic activities.

Distribution of USDT between regular accounts and smart contracts in Ethereum and TRON. Data: BIS.

Visa Also Distinguishes "Noise"

A similar issue is addressed by Visa Onchain Analytics, which provides both total and adjusted volumes of stablecoin transactions.

For the latter calculation, Visa's methodology, in collaboration with Allium Labs, aims to exclude activities that could artificially inflate on-chain volumes, such as high-frequency trading, bot operations, transfers through bridges, internal exchange transactions, and various other technical operations.

“Some on-chain transactions do not resemble settlements in the traditional sense,” Visa explained.

The BIS researchers concluded that analyzing the cryptocurrency market requires more than relying on a single on-chain metric without considering the calculation methodology. They suggested using ranges of estimates, disclosing underlying assumptions, and segmenting data by the technical architecture of blockchains.

In August, BIS head Pablo Hernández de Cos stated that stablecoins do not yet appear to be a reliable payment instrument on a broader economic scale. The organization considers tokenized bank deposits as the foundation for a digital financial system.