Expected Returns and Markov-Switching Illiquidity
2006 Institute for Quantitative Research in Finance Meetings (Q-Group)
26 Pages Posted: 23 Aug 2006
Date Written: August 19, 2006
This paper employs an intuitively appealing Markov-switching regime model to investigate the time-series nature of stock market illiquidity, and the intertemporal relation between illiquidity risk and expected stock returns. We introduce a two-state Markov-switching regime model for illiquidity, stock returns and stock volatility. We find evidence of a significant illiquidity risk premium; the expected stock return is positively related to the conditional probability of the illiquid regime. By combining the Markov-switching model with a log-linear model for stock returns, we derive a tractable expression for the illiquidity feedback effect. The illiquidity feedback effect is critical to understanding the relation between realized stock returns, expected returns and Markov-switching illiquidity. We develop a flexible Bayesian Markov chain Monte Carlo (MCMC) approach for estimating the model.
Suggested Citation: Suggested Citation