A Bank for International Settlements working paper highlights that commonly used crypto metrics are much noisier and more uncertain than market participants often assume. Based on roughly 100 billion Mercurius processing records across Bitcoin, Ethereum and Tron, the researchers show that Bitcoin onchain transfer value estimates can differ by as much as a factor of six depending on how transaction outputs are treated. The study tests three approaches for Bitcoin: counting all outputs, excluding outputs returned to the sender as likely change, and a conservative lower bound that excludes identified self‑transfers or subtracts the largest output when none are identified. Applying those methods produced sharp divergence in monthly transfer estimates, especially during periods of heightened activity and price increases.
The paper also documents valuation effects: conventional market capitalization can at times reach up to four times realized capitalization during sharp price surges, while realized measures can temporarily exceed nominal market cap during declines because older outputs keep earlier valuations. The BIS data indicate about 3.5 million BTC remained dormant for more than ten years and roughly 1.8 million BTC for more than 15 years, yet thousands of coins moved after decade‑long inactivity, which undermines certainty in age‑based lost‑coin estimates.
Ethereum and stablecoins generate different measurement complications. The study examined roughly 67.5 million deployed and active Ethereum contracts and could not classify more than 54 million within the paper’s technical taxonomy; among classified contracts the authors identified close to 12 million proxies, about 1.4 million fungible token contracts and approximately 100,000 NFT contracts. Stablecoin flows vary substantially by chain: the share of USDT held in smart contracts on Ethereum climbed above 20% in 2022 and then settled roughly between 10% and 20% afterward, while on Tron smart‑contract holdings typically hovered near 1%. The authors warn that aggregating USDT activity across chains can conflate distinct economic roles and therefore recommend treating onchain indicators as ‘‘noisy approximations rather than direct measures of economic activity.’’
The paper calls for methodological clarity: protocol‑specific ranges, technical contract classification, and explicit assumptions when reporting onchain metrics. Some providers already apply filters; for example, Visa’s Onchain Analytics, powered by Allium Labs, reports both total and adjusted stablecoin volumes and uses filters to remove high‑frequency trading, bots, bridge routing and internal exchange activity. A September snapshot showed $6.4 trillion in total stablecoin transfers over 30 days versus $313.1 billion after Visa’s adjustments. The BIS working paper is not a regulation, but its findings press market analysts and data platforms to be transparent about measurement choices and the uncertainties those choices entail.
This summary is composed by the cFlash AI agent from multiple public sources, under human supervision. The content is for informational purposes only and does not constitute investment, financial, legal, or tax advice.