A number of proxies for illiquidity have been proposed in the literature that relates trading costs to asset prices. However, some of the illiquidity measures provide equivocal relations to returns. Other measures conceal important dynamics underlying highfrequency data because they are constructed from daily or lower frequency databases. In this study, we adopt a direct and intuitive approach to estimating illiquidity. Specifically, we estimate a set of price-impact parameters based on four different models using the intradaily order flows processed via the Lee and Ready (1991) algorithm from the tickby- tick databases for NYSE stocks over the past 23 years. Our empirical results provide strong evidence that illiquidity measured by the price-impact parameters is priced in the cross-section of stock returns, even after controlling for risk factors, firm characteristics, and other illiquidity proxies prevalent in the literature. Consistently high levels of statistical significance also suggest that the price-impact parameters estimated using the intradaily order flows are more reliable proxies for illiquidity.
목차
Abstract I. Estimation of Price-Impact Parameters A. Lee and Ready’s (1991) Algorithm B. Models for Estimating Price-Impact Parameters II. Methodology III. Data, Definitions, Descriptive Statistics, and Adjustments A. Order Flows and Price-Impact Parameters B. Other Definitions and Descriptive Statistics C. Gallant, Rossi, and Tauchen’s (1992) Adjustments IV. Empirical Results A. Features of the Portfolios Formed on Illiquidity and Firm Size B. Cross-Sectional Regressions C. Robustness Checks V. A Horse Race with Alternative Measures A. Selection of Alternative Measures and their Relations to the Price-impact Parameters B. Cross-Sectional Regressions with Alternative Illiquidity Measures VI. Conclusion References Table
저자
Sahn-Wook Huh [ Faculty of Business, Brock University, St. Catharines, Ontario, Canada ]