Management buy shares for one reason, but how good historically have they been in accurately predicting future returns for their share price?