2021

Grobys, Klaus; Junttila, Juha

This is the first paper that explores lottery-like demand in cryptocurrency markets. Since recent research provides evidence that cryptocurrency returns appear to be short-memory processes, we modify Bali, Cakici and Whitelaw’s (2011) and Bali, Brown, Murray, and Tang’s (2017) MAX measure and employ a weekly forecast horizon and daily log-returns from the previous week to calculate the metric for our portfolio sorts. From an econometric point of view, this study proposes statistical tests that are robust to unknown dynamic dependency structures in the cryptocurrency data. Our results show that average raw and risk-adjusted return differences between cryptocurrencies in the lowest and highest MAX quintiles exceed 1.50% per week. These results are robust to control for Bitcoin risk or potential microstructure effects. Our findings are important also from a theoretical point of view because they suggest that parallel to stock markets, similar behavioral mechanisms of underlying investor behavior are present also in new virtual currency markets.