Cross-Sectional Return Predictors of Utility Tokens

Abstract

This paper takes into account the differences between crypto-coins and crypto-tokens, and investigate the performances of cross-sectional return predictors based on a large sample solely consisting of utility tokens (over 1,000 ERC-20 tokens). Besides the most famous and longstanding predictors such as size and momentum, we thoroughly examine the fundamental-related predictors formed by using on-chain variables including dollar-value of transactions, transfer counts and unique active addresses, which reflect real economic activity on the blockchain and proxy intrinsic values of the tokens. We further construct a pricing-factor model including a quasi value factor, which is a counterpart of the value factor HML in equity market. By following Fama and French (1996), we found that the pricing model could to some extent explain the excess returns of 25 double-soring portfolios.