Abstract
In the spirit of Merton (1973), we assert that temporary aggregate market illiquidity is compensated for in the form of higher conditional market returns. In order to test this hypothesis, we use two available liquidity proxies, namely versions of the Amihud illiquidity measure and a measure based on exchange traded fund prices. Our investigation is based on vector autoregressive models for the German stock market between July 2006 and June 2010. The fund-based illiquidity proxy dominates in capturing consistent results for the determination of time-varying market returns. Temporary illiquidity is indeed compensated for by higher market returns. We confirm a bidirectional relation between illiquidity and market returns, i.e. current returns depend on lagged illiquidity and current illiquidity can be determined by a combination of past returns as well as past illiquidity. The relation shows that illiquidity is persistent and driven by market declines.
1. Introduction
In their seminal paper, Amihud and Mendelson (1986) point out that liquidity has a significant impact on asset prices. Periods of market stress repeatedly exhibit that liquidity dries up and therefore stock markets overall do no longer provide the characteristics of stable turnover, balanced spreads, and smooth adjustments of price.1 Any form of temporary illiquidity increase in stock markets is therefore an important signal to market participants. It is of particular interest to study the characteristics and determinants of aggregate market liquidity as well as its impact on aggregate asset prices.
3. Conclusion
In the spirit of the pricing of non-diversifiable risks on the market level, this paper finds significant evidence for temporary illiquidity premia in market returns and hence the pricing of aggregate market illiquidity. Using the German stock market as an example, the fund-based illiquidity measure EILLIQ is found to deliver significant and consistent results on the dynamic bivariate relation between market illiquidity and market returns. Future research on alternative measures and alternative versions of existing illiquidity measures, as well as approaches that split illiquidity into expected and unexpected components, may derive even more detailed results on this important relation.
