Optimal Research Of Loans' Portfolio Based On Simulated Annealing Algorithm

The business of modern commercial banks includes capital organization, credit business and other intermediate business, and major risks focus on credit loans. The domestic and international commercial banks put forth new ideas to control credit risk. And also, with economic globalization, development of science and technique, customer scale of bank enlarges rapidly, which makes thousands of banks compete face to face. Commercial banks confront incandescent competition. Commercial banks of our country control loans risk by managing the risk grade of single loan, which need adjusted in order to meet with the fierce competition.Firstly, this thesis starts with credit risk management of commercial banks, describes the necessary of applying loans' portfolio to reduce loan risk, and builds up a decision-making model of loans' portfolio optimization based on principle of maximum earning and minimum risk. Then, analyzed from calculation methods, this model is a type of Knapsack Problem with restrictions between lower-limit and upper-limit, which is a discontinuous and multi-maximum complicated problem. When problem scale is small, we can adopt many classical algorithms such as linear programming integer programming, branch and bound method to solve this problem. But, while scale increases, solving process become too long, so we need approximate algorithm to achieve the balance between the quality of result and time. By making use of simulated annealing algorithm with memory, and determining a set of effective cooling schedule, the thesis solves this complex and special knapsack problem successfully.