MLiq a meta liquidity measure

The last crisis sheds light on the importance to consider liquidity risk in the financial industry. Indeed, liquidity had a predominant role in propagating the turmoil. In contrast, controlling for liquidity is a difficult task. The definition of liquidity links different dimensions that are impossible to fully capture together. As a consequence, there exist a lot of liquidity measures and we find in the literature some solutions to take into account more than one dimension of liquidity but also liquidity measures considering a long lasting liquidity problem. In this paper, we focus on drastic illiquidity events, i.e liquidity problems reported by several liquidity measures simultaneously. We propose a Meta-Measure of liquidity called MLiq and defined as the probability to be in a state of high liquidity risk. We use a multivariate model allowing to measure correlations between liquidity measures jointly with a state-space model that endogenously defines the illiquid periods.