Liquidity Dynamics and Cross-Autocorrelations

Abstract

This paper examines the mechanism by which the incorporation of information into prices leads to cross-autocorrelations in stock returns. We present a simple model where trading on private information occurs first in the large stocks and is transmitted to small stocks with a lag. Such trading impacts large stock liquidity, so that, in equilibrium, large stock illiquidity portends stronger cross-autocorrelations. Empirically, we find that the lead-lag relation between large and small stocks increases with lagged illiquidity indicators of large stocks. Further, order flows in large stocks significantly predict returns of small stocks when large stock spreads are high, at both the market and industry levels. In addition, the role of order flow and liquidity in predicting small stock returns is stronger prior to macro announcements (when information-based trading is more likely).