Factoring Information into Returns
Journal of Financial and Quantitative Analysis
We examine the potential profits of trading on a measure of private information (pin) in a stock. A zero-investment portfolio which is size neutral, but long in high pin stocks and short in low pin stocks earns a significant abnormal return. The Fama-French, momentum and liquidity factors do not explain this return. However, significant covariation in returns exists among high pin stocks and among low pin stocks, suggesting that pin might proxy for an underlying factor. We create a pin factor
... eate a pin factor as the monthly return on the zero-investment portfolio above and show that it is successful in explaining returns to independent pin-size portfolios. We also show that it is robust to inclusion of the Pastor-Stambaugh liquidity factor and the Amihud illiquidity factor. We argue that information remains an important determinant of asset returns even in the presence of these additional factors. Factoring Information into Returns In this research, we develop a pin factor and investigate its performance for explaining cross-sectional asset returns. Using the time-series regression approach of Fama and French (1993), we find intriguing evidence of success in explaining returns to independent pin-size portfolios. In particular, inclusion of a pin factor substantially reduces the intercepts relative to those of the Fama-French factors, momentum factor, plus liquidity factors model. Indeed, for seven of ten size decile portfolios these intercepts become statistically insignificant, although they remain at least marginally significant for the three smallest portfolios. We interpret these results as supporting the role of pin or information risk as a statistical factor in affecting asset returns. These results also demonstrate that liquidity and information risk may both play distinctive and important roles in explaining asset returns. Some research (see Duarte and Young (2007) ) argues that liquidity effects unrelated to information risk explain the relation between pin and the cross-section of returns. However, we find that the pin factor remains significant even when both the Pastor-Stambaugh factor and Amihud illiquidity factor are included. Therefore, pin does not appear to merely be proxying for liquidity effects. Such a finding should not, however, be unexpected. Liquidity effects can arise for a variety of reasons that are unrelated to information, and measuring such liquidity effects is at best inexact. 1 Nonetheless, our results here suggest that factoring 1 The Amihud measure reflects total price impacts and does not sort them into information-based and inventory-based components. So finding that the Amihud illiquidity factor is priced is not informative about information versus inventory price impacts. Whether this factor is actually priced or not, or more precisely whether it is reliably and stably priced, is still in debate. Charoenrook and Conrad (2004) find that an ILLIQ factor based on Amihud's illiquidity measure is priced over the 1963-2003 sample period. Using returns on the Amihud factor provided by Charoenrook and Conrad we find significant average returns in the period 1963-1983, but not in the 1984-2002 sample period of Acharya, V., and L. H. Pedersen. "Asset Pricing with Liquidity Risk." Journal of Financial Economics, 77 (2005), 375-410. Amihud, Y. "Illiquidity and Stock Returns: Cross-section and Time-series Effects."