The inclusion of time-different covariates into survival Examination has led to higher predictions of time to default in behavioural credit history scoring styles. However, when these time-different covariates are endogenous, There's two significant troubles: estimation bias in the survival product and not enough a prediction framework for upcoming