Reversal in Small Cryptos

Value Premium, Network Adoption, and Factor Pricing of Crypto Assets

AuthorsLin William Cong; George Andrew Karolyi; Ke Tang; Weiyi Zhao

Paper linkhttps://papers.ssrn.com/sol3/papers.cfm?abstract_id=3985631

The investment universe consists of cryptocurrencies from CoinMarketCap.com. Stablecoins, and coins with zero price, market capitalization, or trading volume across all periods are excluded. Non-return variables are winsorized at the 1st and 99th percentiles. Market capitalization and prices are sourced from CoinMarketCap.

First, cryptocurrencies are sorted into five portfolios based on their week-end market capitalization. Within each size quintile, coins are then sorted into five portfolios according to their past two-week returns (momentum). Among the three smallest size portfolios, the strategy goes long the bottom momentum portfolios. This variant is used and averaged to ensure the strategy is executable and avoids relying on the very smallest cryptos. The momentum effect is strong in smaller to mid-sized cryptos but reverses in the largest size quintile.

The three lowest-momentum portfolios in the small size quintiles are equally weighted, while each portfolio is value-weighted individually. The strategy is rebalanced weekly.

The underlying economic reason for the strategy is not explicitly defined. The effect was identified using a portfolio-sorting approach and is highly statistically significant, but the paper does not provide a precise behavioral explanation. Possible drivers include mean-reversion or prior underappreciation of certain cryptocurrencies before the signal formation.

Robustness tests across different lookback periods—one, three, or four weeks—show qualitatively similar results, confirming that the effect is not sensitive to the exact two-week window and remains statistically significant.