Probit Model Question

#1
I often work with time-series models and this is my first attempt at a probit model. I am working on a probit model for leveraged buy-out (LBO) transactions. LBOs occur when a public company becomes a private company (i.e., buy-out). The firm that purchases the public company typically uses a high amount of debt (i.e., leverage).

I assigned the dependent variable a "1" if a LBO occured and a "0" if no LBO occured. I have tested about 20 independent variables in this model. Two of the independent variable are significant, however the McFadden R-squared is around 0.14. When I remove the insignificant variables, the McFadden R-squared drops to 0.03.

My questions are: Does the low McFadden R-squared prohibit me from using the probit model to predict future LBOs? If so, is the fact that I found two significant independent variables useful (assuming the model as a whole is not).

Also any books or reference recommendations are appreciated.

Thank you for your time.