The interrelation of Liquidity Risk, Default Risk, and Equity Returns
by Maria Vassalou of Columbia University,
December 7, 2005
Abstract: As proxies for liquidity risk we consider the Pastor-Stambaugh measure, as well as the turnover and illiquidity measures. The default measure of choice is the one based on Merton's (1974) contingent claims approach. The alternative liquidity measures contain very different information about liquidity and share low correlations. However, they are all related to our default measure. Vector autoregressive tests reveal the existence of a two-way causal relation between default risk and stock market returns. Liquidity risk does not affect the future path of stock market returns. These relations hold, even when we take into account the correlation of the default and liquidity measures with aggregate stock market volatility. Low liquidity stocks earn higher returns than high liquidity stocks, only if these stocks also have high default risk, but in no other case. In contrast, high default risk stocks always earn higher returns than low default risk stocks, independently of their liquidity level. The inclusion of default and liquidity variables in popular asset pricing specifications improves the model's performance, but the improvement is larger in the case of the inclusion of the default variable. Finally, in the presence of the default variable, the inclusion of a liquidity proxy in the specification results in a marginal improvement of the model's performance, but the opposite is not true. Our findings regarding the interrelation of default and liquidity risk and their effects on equity returns are independent of the liquidity proxy considered.
Keywords: asset pricing, liquidity risk, expected returns.