This study examines whether mutual funds herd in industries and the extent to which such herding impacts industry valuations. Using two herding measures proposed by Lakonishok et al. (1992) and Sias (2004) we document that mutual funds herd in industries. We show that industry herding is not driven by fund flows and that it is not a manifestation of individual stock herding. We also find evidence indicating that herding in industries by mutual funds is related to the industry momentum phenomenon first documented by Moskowitz and Grinblatt (1999), but it does not drive industry valuations away from their fundamentals.

Empirical studies on whether investments are sensitive to cash flows in imperfect markets often report conflicting results and have been criticized on conceptual and methodological grounds. Our study mitigates some of these problems using a research design that relates changes in investment–cash flow sensitivity to changes in the bid-ask spread measure of information asymmetry surrounding (i) implementation of the Sarbanes-Oxley (SOX) Act and (ii) deregulation of firms in the Transportation, Telecommunication, and Petroleum and Natural Gas industries. Consistent with our hypotheses, we find that information asymmetry decreases following SOX and that there is a corresponding decrease in the investment–cash flow sensitivity, pre- to post-SOX. Further, greater decreases in information asymmetry following SOX are associated with greater decreases in investment–cash flow sensitivity. The results for the deregulation sample are also consistent with our hypothesis, wherein we observe an increase in information asymmetry and corresponding increase in the investment–cash flow sensitivity following deregulation.