Price Clustering on Cryptocurrency Order Books

Abstract

We investigate the price clustering effect, or the tendency of traders to place orders in round numbers, in cryptocurrency limit order books. Using 10-minute order book snapshots for four dollar-denominated cryptocurrencies over 35 weeks from January to August 2020, we find that price clustering is more pronounced at higher price levels. We also show that the larger the most recent cumulative dollar depth, the stronger the clustering effect. Finally, we examine the informativeness of the quote clustering effect in the order book by showing that the clustering effect of bid prices, but not of ask prices, can explain next-day returns.