Coordinated charging schedules of plug-in electric vehicles (PEVs) by an aggregator agent may lead to increased system efficiency in allocating resources in generation, transmission and distribution. To achieve optimal charging schedules, many studies have assumed that the PEV aggregator can exercise full direct load control. This paper proposes a mathematical program with equilibrium constraints for the PEV aggregator's decision making in different electricity markets, using indirect load control by determining optimal retail prices for the PEV. This permits the final customers to decide on their charging schedule by decentralized profit optimization. These decisions respect a potential discomfort that may arise when PEV users have to deviate from their preferred charging schedule as well as include the option of using alternative sources of energy. In a small case study of 3 vehicle clusters and 6 time periods the model's functionality is highlighted. Results indicate that under reasonable competition on the retail market, the PEV aggregator's profitability depends on providing the right price signals to the final customers, such that the most efficient charging schedule response is achieved.