If you've gotten it to working conc, then presumably it's in the working buffer (PBS Tween maybe, or powdered milk?). You need to watch for other elements not going back into solution....
Many times we make extra working solution to save it from pipetting errors , but It is not a good idea to keep the working solution in -20. Instead you can keep the working solution of HRP Antibody in 4 degrees if you want to use it within 2-3 days and can add some extra Ab from stock at the time of use.
Grant got a point that some buffers can give you trouble. PBS Tween should be fine if you have added all your antibody to a buffer, but I would not recommend to do it as standard protocol. It is way easier to add 1:1000 or whatever than thaw a stock every time and I would guess it is also more precise. Normally you add buffer at 4 degree or RT - if you freeze your diluted antibody you should make sure temperature is same every time.
Manoj, you can´t just add some extra Ab before you use it - how would you ever repeat such an experiment or write this in a method section?
I concur with Judith - enzymes can be stored at -20 in 50% glycerol for significant periods of time. 50% glycerol solutions don't really freeze at -20. It's the crystallization of ice that denatures and inactivates many proteins.