The growing inflow of funds in the crypto market makes understanding the dynamics and drivers of market liquidity a significant concern to drive informed regulatory, trading, and risk management decisions. This paper investigates the nexus between liquidity and volatility in cryptocurrency markets. We adopt the Asymmetric MEM-X, an extension of the well-known MEM, to model the dynamics of liquidity including realised volatility as exogenous variable at high frequency. This unified framework overcomes the shortcomings of traditional volatility and liquidity models which fail to assess the impact of market volatility on liquidity. In particular, the information about market liquidity is derived from easy-to-access transaction data through easy-to-compute bid-ask spread estimators. We apply the model to real data on two major cryptocurrencies, bitcoin versus US dollar (BTCUSD) and ethereum against US dollar (ETHUSD), considering two different exchanges, Bitstamp and Bitfinex. Our findings reveal (i) a distinct U-shaped intraday pattern in both liquidity and realized volatility, and (ii) a strong negative nexus between volatility and liquidity in cryptocurrency markets.
Forino et al. (Wed,) studied this question.