Patterns of Insider Trading: It Is Not All Black and White

Abstract

The regular pattern of quarterly earnings announcements sets up a predictable pattern of information asymmetry in the market. Both regulatory restrictions and voluntary corporate restrictions direct trading to low information asymmetry periods. To understand the effect of these restrictions, this study examines insider trading in three different windows: white windows (3–12 trading days after the earnings announcement, periods with low information asymmetry), black windows (all the other days in the quarter, periods with higher information asymmetry), and the blackest windows (the last 10 trading days of the black window, periods with the highest information asymmetry). First, our results show that a large proportion of insider trading in the United States takes place in the black window. Second, we document that trading in the white period exhibits a strong self-selection bias. We also show that the excess returns earned by black period trades vanish if postponed to the next white period following the earnings announcement. Finally, we show that a relatively large proportion of pre-specified trading under SEC-sponsored 10b5-1 plans are filed for black window periods, but the difference across black and white window plans is a matter of frequency of trade rather than the magnitude of profits. Overall, these results suggest that insiders balance the risk and profitability of their trading in white and black windows and that insider trading restriction in high-information asymmetry periods is not effective in practice.