The Information Content of the Amihud Ratio for Intraday Liquidity Forecasting

Abstract

This study aims at forecasting periods of intraday low execution price uncertainty. We compute an adjusted Amihud illiquidity ratio to detect those periods in the stock market. Our methodology is built around a fixed volume event chart and presents a threefold advantage. First, we solve the non linearity issue of the widely used Amihud ratio. Second, we consider liquidity as a trade size specific concept and bypass optimal sampling frequency issues. Third, we turn liquidity into a volatility proxy, which allows us to transpose the extensive volatility forecasting literature into the liquidity framework. We outline some predictable patterns that statistically outperform a random walk model. We also provide an illustration of execution improvement strategy that outperforms a naive strategy. Overall, we found that the best time to trade for patient traders is towards the end of the continuous trading session.