First of all, the sample period 1980-2007 is divided into seven-four year periods in order to study the time trends in idiocies and noncommon. Table 3 shows the result of a simple regression of noncommon on idiorisk. There is a significant positive relationship between noncommon and idiorisk in the periods 80-83, 88-91, 92-95 and 96-99. The relationship in the period 00-03 thoughhighly significant is negative and then turns nonsignificant in the 04-07 periods. These results give further credence to the findings of brandt et al. (2010) that the positive surge in idiosyncratic volatility up to the late 1990s underwent a reversal in the 2000s and by 2003 volatility had fallen back to the pre 1990 level. Our findings are therefore consistent with those of Brandt et al. in that we find that the positive relationship between noncommon and idiorisk reversed in the 2000-03 period. The R2 of the relationships are consistently high in all the periods, ranging from 0.43 to 0.68, and indicate that noncommon explains a good portion of idiorisk changes.