A Research Of Robust Optimization In Financial System

The financial globalization, financial liberation and the rapid development of network finance have raised new challenges to both theories and techniques of financial management. The uncertainty and fluctuation of financial market have increased, more financial innovations are created, and competition environment of financial market becomes more complicated. Especially since the frequent eruption of financial crisis in recent years, as a result, the financial decision problems that consider the uncertain factors are extremely critical. Robustness and robust optimization in financial system have become challenging problems in financial engineering, and are playing an important role in both theoretical research and practical application.This paper put the research emphases on the problems in financial decision by using robust optimization. On the basis of the literature review on robust optimization in financial system, the robust optimization problems studied in this paper are divided into three aspects:The first one, the decision of portfolio selection; the second one, the decision of fund operation under network environment; The last one, the decision of dynamic financial assets allocation in the view of time periods.The main work of this paper can be illustrated as follows:Firstly, the robust optimization problem of portfolio selection based on the decision of portfolio selection, including the robust optimization model of loan portfolio selection of commercial banks, robust optimization model of tracking error of securities portfolio selection and the robust optimization model of tracking error with VaR constrains.The robust optimization of linear matrix inequity is applied to tackle the drawbacks that mean-variance model is sensitive to parameters and requires normal distribution so as to solute the optimal solution under the worst case.The results of empirical analysis and numerical simulation indicate that the robust optimization based on linear matrix inequity is effective and feasible.Secondly, the robust optimization problem of fund operation of bankcard under network environment. On one hand, the full-time operation of bankcard network is considered. On the other hand, minimization the system cost is taken into account. On the foundation of these two considerations above,the dissertation establishes the robust optimization model of fund operation in the bankcard network under the background of China UnionPay.The results indicates that the model solutions are relatively conservative, which guarantees the robustness of bankcard network.Thirdly, the robust operation problem of guaranteed cost control in the dynamic financial assets allocation. For the uncertainty in the financial assets allocation problem, the dissertation establishes the dynamic assets allocation model with uncertain assets price and interest rates. The principal of optimal quadratic guaranteed cost with deterministic upon bounds is obtained by applying the guaranteed cost control of linear matrix inequity. The outcome suggests that the strategy of guaranteed cost control can restrained the uncertain disturbance in the system of dynamic assets allocation.The last part of this dissertation summarizes the main conclusions and contributions of the research and points out the further research aspects.