Quantile liquidity connectedness in foreign exchange markets

This paper examines how liquidity connectedness in the foreign exchange (FX) market varies across the conditional distribution of liquidity. Using a quantile-based connectedness framework, we analyze nine major currency markets under both normal and extreme liquidity conditions. We find that liquidity connectedness is state-dependent and follows a U-shaped pattern across the distribution. Connectedness is stronger in both the lower and upper tails of the conditional distribution of liquidity, corresponding to highly liquid and severely illiquid states. These findings suggest that mean-based measures may not fully capture regime variations in liquidity transmission. Quotation- and trade-based liquidity measures exhibit heterogeneous dynamics across states, which are linked to market risk, funding conditions, and intermediary balance sheet constraints.