I worked for a while in a quantitative equity shop. They had useful models of how executive stock purchases predicted share price change. However, as hard as they tried, they could never find a formula that made executive stock sales a predictive variable. Too many exogenous variables, from family needs to upcoming additional stock awards.