This study investigates the stochastic optimal electricity procurement problem of an electricity retailer in a smart grid environment. Various energy provision means, including distributed generations (DGs), pool market (PM), forward contracts (FC), photovoltaic (PV) system, wind turbine (WT), and Battery energy storage system (BESS) are considered. A novel plug-in electric vehicle (PEV) based demand response program (DRP) is integrated into the problem considering economic incentives for PEV owners, which is computationally fast, and it is based on the uncertainties derived from PEVs, such as arrival/departure time, daily traveled miles and vehicle type. Moreover, the uncertainty of PM price, solar irradiation, wind speed, and conventional loads are taken into account by an artificial neural network (ANN)-based scenario generation method. Furthermore, an analytical linear BESS degradation cost model is integrated into the objective function to ensure fast global optimization. To confirm the functionality of the proposed method, different operation schemes, such as with and without DRP, with and without battery degradation cost, are thoroughly inspected. Additionally, the sensitivity of the problem for the number of PEVs is analyzed. The proposed two-stage stochastic mixed-integer linear problem (MILP) is solved by commercially available optimization software.