What is the best proxy for liquidity in the presence of extreme illiquidity?

This paper aims to indicate the best daily proxy for unobserved liquidity in the presence of extreme movements on the market. We apply copulas to investigate the dependence between benchmarks based on intraday prices and proxies based on daily data. We focus on the tail dependence between both types of measures. Our results show that when the market experiences extreme illiquidity, the Closing Quoted Spread (CQS) based on daily closing bid and ask prices is superior to other percent-cost low-frequency proxies. We find the highest tail dependence coefficients for CQS and either the Percent Effective Spread or the Percent Quoted Spread.