Factor model can provide some predictive power to digital assets, especially cryptocurrency. The model might also indicate a new asset class that extends portfolio diversification. In short, large capital sizes, lower volatilities, and recent winners lead to better performance. Furthermore, the double sorted approach shows significant higher returns, which generated 5.40% and 9.92% monthly returns when we control Size-Volatility and Size-Momentum.
Long-Only factor model fits the price dynamics pretty well, and the outcome sheds some light for systematic trading. The Long-Short factor, on the other hand, provides impracticality since only BitCoin has futures for short positions. It will be engaging in seeing how the model behaves in 2019.
Toward a Factor Structure in Crypto Asset Returns, J. Li and G. Yi, The Journal of Alternative Investments, 21 (4)