Since the absence of portfolio model, banks may admit the credit card application without an effective evaluation standard in the extension of credit card business. This paper discusses a credit card customer portfolio selection problem in which the returns from customers are given by experts' estimations rather than historical data. The customer factors which directly affect the returns are regarded as uncertain variables. Based on customer segmentation, we proposed an optimal customer selection model with the objective of maximizing the expected return value within the constraints of capital limitation and risk diversification and the requirement on credit loss. To facilitate users to solve the model with practicable programming solvers, a crisp equivalent model is provided to show how credit card managers can make use of the model to select customers under the uncertainty theory.
展开▼
