A large CVaR-based portfolio selection model with weight constraints is proposed.
It includes the standard CVaR-based portfolio selection model as a special case.
We solve the new model using penalized quantile regression techniques.
The real data application illustrates the efficacy of our proposed model.