“Why is the Amihud (2002) Illiquidity Measure Priced?” by Tao Shu
The University of Georgia
University of Delaware
This paper empirically studies the widely used Amihud (2002) illiquidity measure using a large sample of NYSE and AMEX firms from 1964 to 2012. Although the Amihud illiquidity measure is intended to capture price impact using the ratio of return to trading volume, we find that the pricing of the Amihud measure is completely explained by its trading volume component. Specifically, an alternative measure using only the trading volume component of the Amihud measure (ratio of 1 to trading volume) has a correlation of 0.92 with the Amihud measure, and exhibits a similar return predictability. Additionally, the Amihud illiquidity measure is no longer priced after we remove its variation due to the trading volume component. These results are robust to multivariate regression analyses, alternative versions of the Amihud measure, and alternative sample selections such as NASDAQ stocks. These findings suggest that the pricing of the Amihud illiquidity measure is associated with total trading activities instead of price impact.