On-Chain Factors and Cryptocurrency Asset Pricing: Evidence from Ethereum-Based Tokens

Abstract

This paper examines whether on-chain factors derived from Ethereum blockchain data contain pricing information beyond established cryptocurrency risk factors. We construct 27 on-chain factors across four dimensions (network activity, scale-adjusted activity, valuation ratios, and token distribution) for 122 Ethereum-based tokens from July 2020 to August 2025, and evaluate them against 26 benchmark factors spanning size, momentum, volume, and volatility using a double-selection LASSO framework, complemented by portfolio sorts and three-factor regressions. Eight on-chain factors are significant in the cross-sectional pricing test, with the strongest evidence concentrated in scale-adjusted activity and token distribution. Transaction count to network value is the only factor that remains significant in the cross-sectional pricing test, portfolio sorts, and three-factor regressions. In contrast, valuation ratios based on market-to-realized values do not survive as independent sources of abnormal return once momentum is taken into account, reflecting the mechanical overlap between recent price appreciation and slowly adjusting realized values. Token-distribution factors, particularly small-holder share and centralized exchange share, generate the most robust abnormal returns and remain economically meaningful under equal-weighted construction and conservative transaction-cost assumptions. Subperiod analysis further reveals a change in on-chain pricing power: scale-adjusted activity factors are stronger earlier in the sample, whereas distribution-based factors become more important over time. Overall, the results show that blockchain-native information, especially holder distribution, captures a distinct dimension of cryptocurrency asset pricing.