The fraction of quasar ’ s and gamma-ray burst ( GRB ) afterglows that vary due to microlensing by the stellar populations of intervening elliptical/S0 galaxies is computed by combining the joint distribution of effective microlensing convergence ( \kappa ) and shear ( \gamma ) with microlensing magnification patterns .
Microlensing is common in multiply imaged sources .
We find that 1 in 3 multiply imaged quasars should vary by more than 0.5 magnitudes per decade due to microlensing , while 10 % of macrolensed GRB afterglows should show a departure of more than 0.5 magnitudes from their intrinsic light-curve during the first 30 days .
However microlensing by stars is rare in general , with only 1 source in \sim 500 varying by more than 0.5 magnitudes during the same periods .
We find that most microlensing by stars will be observed in a regime where \gamma > 0.1 .
Thus point-mass lenses do not provide an adequate description for most microlensing events .
If dark matter halos contain a large fraction of mass in compact objects , the fraction of microlensed ( by 0.5 magnitudes ) images rises significantly to \sim 1 in 10 for quasars and \sim 1 in 5 for GRB afterglows .
Comparison of variability between macrolensed and normal quasar images , and a moderate number of well sampled GRB afterglow light-curves should therefore discover or refute the existence of stellar mass compact objects in galaxy halos .
While microlensing results in departures of the distribution of magnifications from that of a smooth model , the effect on the macrolensing magnification bias for the discovery of lenses in quasar surveys is small .
On the other hand , microlensing significantly broadens the distribution of macrolensed image flux ratios .