Asset Pricing with Extreme Liquidity Risk

Abstract

Defining extreme liquidity as the tail of the illiquidity for all stocks, I propose a direct measure of market-wide extreme liquidity risk and find that it is priced cross-sectionally in the U.S. Between 1973 and 2014, the stocks in the highest quintile of extreme liquidity risk loadings earned value-weighted average returns 5.6% per year higher than the stocks in the lowest quintile. The extreme liquidity risk premium is robust to common risk factors related to size, value, and momentum, and is different from that on aggregate liquidity risk documented in Pastor and Stambaugh (2003) as well as that based on the tail risk of Kelly and Jiang (2014). Extreme liquidity risk can provide an advanced warning about extreme liquidity events, and it reliably outperforms aggregate liquidity measures in predicting future market returns. I incorporate extreme liquidity risk into Acharya and Pedersen’s (2005) framework and find new supporting evidence for their liquidity-adjusted capital asset pricing model. I explore potential economic mechanisms through which the rare and large fluctuations in stock-level liquidity risks are priced.