When Does Insider Selling Support a 'Strong Inference' of Fraud?

Abstract:     
Plaintiffs in securities fraud class action lawsuits often allege that sales by insiders support a "strong inference" of fraudulent intent. This study examines insider transaction activity for a sample of 842 companies sued in lawsuits alleging a violation of Section 10 (b) of the Securities Exchange Act of 1934 to ascertain whether net insider selling during the litigation class period (the alleged period of misinformation) is unusual in amount and timing relative to net insider selling at other times and at other firms.

We find that class period net insider selling for the litigation sample significantly exceeds net insider selling before and after the class period and exceeds net insider selling around the same dates for a sample matched on industry. The more pronounced differences are for firms with a greater asymmetry of information between insiders and outsiders, namely, smaller firms and firms with lower analyst coverage. We also document a significant, positive association between net insider sales and short interest, particularly for smaller firms and for firms with lower analyst coverage and find that the level of insider transaction activity immediately prior to a corrective disclosure is not elevated to a statistically significant degree once we control for short interest and other information variables.

These results augment the financial literature by providing evidence of information-based selling by insiders around a disclosure event that leads to securities litigation. From a legal perspective, these results are consistent with the view that plaintiffs self select issuers whose patterns of class period insider selling diverge from historical or industry norms. The study is, however, unable to distinguish whether the elevated class period insider selling is the result of insiders' use of material nonpublic information or insiders' (and short traders') legal use of costly public information. The study nonetheless supports a statistically credible, non-fraud alternative explanation for observed elevated levels of insider selling. Accordingly, we suggest that courts should temper the adverse inferences that they might draw from an observed unusual level of insider selling when that level also reflects a heightened level of contemporaneous short interest activity, particularly if the issuer is a smaller company with a lower level of analyst coverage.

Working Paper Series