Bitcoin liquidity in a market microstructure

Abstract

This master thesis investigates Bitcoin liquidity in a market microstructure setting and has\nbeen divided into two parts. In the first part we examine whether there is any correlation\nbetween the results generated by three liquidity measurement techniques (The Rolls, ILLIQ\nand Coefficient of elasticity) and the actual market microstructure spread. We find that only\nILLIQ shows a moderate, statistically significant correlation with the spread.\nThe second part investigates whether selected primary and secondary variables affect the\nspread and whether their relationship is in line with existing financial research and our\nintuition. This part also examines whether there is any weekly cyclicality in Bitcoin liquidity,\nand if occurrence of events affects Bitcoin liquidity or not. We find that the primary and\nsecondary variables have a significant impact on the bid ask spread, and that the nature of the\nrelationship is in line with empirical research and our intuition. Interestingly, we find that\nvariables which are not directly related to the market microstructure have a relatively higher\nimpact on the spread in comparison to variables that are directly related. We also find that\nBitcoin liquidity does not have any cyclical, weekly trends and only a few of the events have\nany sort of significant impact on liquidity.