Bitcoin On-Chain Transaction Volume Varies by Up to 6x Depending on Calculation Method

A new study shows Bitcoin (BTC) on-chain transaction volumes can differ by as much as six times depending on the calculation method used. The Bank for International Settlements (BIS) published a working paper titled 'What's Hidden in Complexity' on May 15, analyzing approximately 1.3 billion Bitcoin transactions and 3.6 billion outputs from 2009 to 2026.

The key factor is Bitcoin's Unspent Transaction Output (UTXO) structure. When sending 4 BTC with a 1.5 BTC transfer, the remaining 2.5 BTC returns as change to the sender. Blockchain records alone cannot fully distinguish between this change and actual transfers to recipients. Summing all outputs can overestimate economic activity, while excluding self-transactions (change transactions) can significantly reduce the figure.

BIS calculated transaction volume in three ways: an upper estimate summing all outputs, a revised estimate excluding self-transactions, and a conservative lower estimate excluding both self-transactions or the largest output. The revised estimate was up to six times lower than the upper estimate. Researchers noted this discrepancy stems from Bitcoin's protocol structure and how users and service providers conduct transactions, not simple data errors.

Self-transaction volumes have increased significantly since March 2016, likely due to rising address reuse and a focus on convenience by users and service providers. The lower estimate remains an approximation based on specific assumptions, as the study did not adjust for coinjoin, mixers, or spam transactions.

The analysis extends beyond Bitcoin to Ethereum (ETH) and Tron (TRX). On Ethereum, researchers classified 13 million active smart contracts, with about 1.4 million issuing tokens. Over 54 million contracts were excluded from technical classification. Token symbols like USDT are reused across thousands of contract addresses, making it difficult to determine economic significance or issuer based solely on symbol.

For example, USDT held by Ethereum smart contracts peaked above 20% at times, while on Tron it remained around 1%. BIS noted this reflects Ethereum's closer integration with DeFi activities like collateral and liquidity provision, versus Tron's use for payment or store-of-value purposes outside smart contracts.

The study emphasizes that on-chain metrics should be viewed as 'noisy approximations' rather than direct measures of economic activity, depending on methodology. It aligns with previous reports highlighting similar discrepancies in Bitcoin transaction volume calculations.