We offer novel indicators of market-wide liquidity. Previous literature uses averages of individual liquidity indicators to track the evolution of market-wide liquidity. Instead, we focus on the tails of the market liquidity distribution. First, we construct aggregate liquidity indicators using low and high quantiles of six liquidity measures (total volume, number of trades, effective spread, realized spread, price impact and lambda). Our results show that market conditions have an asymmetric impact on the tails of the liquidity distribution. In the second part of the study, we test for nonlinearity of the effects of market determinants on market liquidity.