【作者】 洪忠诚；

【导师】 迟国泰；

【作者基本信息】 大连理工大学 ， 技术经济及管理， 2006， 博士

【摘要】 银行风险事关银行的生存和社会的稳定。贷款组合风险是商业银行的主要风险，由于贷款组合分配失误造成的新增不良贷款不断产生，是我国目前银行业面临的主要问题。科学合理的进行贷款组合优化决策，对于商业银行有效控制新增不良贷款，优化配置资源，提高银行的生存能力和竞争能力，具有重要的现实意义。 本论文共分八章。第一章绪论分析了论文的选题背景与意义、相关研究进展及缺陷、研究方法、技术路线和主要内容。第二章分析了贷款组合与优化决策的理论基础。第三章建立了基于有上界限制的组合贷款决策优化模型。第四章建立了基于行业组合的贷款总体风险优化决策模型。第五章建立了基于0-1规划的存量与增量全部组合贷款优化决策模型。第六章建立了基于VaR约束的多目标组合贷款优化决策模型。第七章建立了基于信用迁移下的CVaR限定组合贷款风险优化决策模型。第八章为结论与展望。 本文将Markowitz的投资组合理论、Sharpe的资本资产定价模型(CAPM)以及VaR和CVaR风险管理理论综合在一起，开展商业银行风险管理中的贷款组合分配模型的研究。论文的主要研究成果如下： (1)建立了基于有上界限制的组合贷款决策优化模型 在马科维兹的资产组合选择模型的基础上，运用了法律法规中的贷款比例限制为上界约束条件，以贷款组合风险最小为目标函数，建立了基于上界限制的银行贷款组合优化模型。本模型的特点一是这种上界限制在分散贷款信用风险的同时，又控制了贷款总量，使业银行有相应数量的资产与负债相互匹配。二是运用资产负债中的资产变化及其结构管理比率建立贷款组合比例上限，通过银行监管和设定贷款组合的预期收益率控制流动性风险，使贷款的分配决策满足银行监管要求和银行经营实际。解决了对贷款有上界约束条件下银行贷款组合优化的问题，为银行进行实际贷款组合提供了量化决策。 (2)建立了基于行业组合的贷款总体风险优化决策模型 应用资本资产定价模型(CAPM)中的单因子模型表达贷款收益和风险函数，以不同行业贷款组合后的总体风险最小化为目标，运用非线性规划方法，建立了基于组合贷款总体风险优化的行业贷款分配模型。本模型的特色与创新一是改变了现有研究仅仅优化非系统风险而与系统风险无关的现状，通过负相关行业的风险对冲，避免了选择单个或少数行业进行贷款所导致的当该行业不景气时的系统性风险对贷款质量的影响，降低了贷款组合的系统性风险。二是从组合贷款总风险中分离出系统风险和非系统风险，并通过实例验证行业组合可以降低系统性风险，揭示了通过行业组合可以部分地抵消由于行业自身所产生的系统性风险。更多还原

【Abstract】 The risk of banks is related to the steady of the society and the survival of the banks. Loan portfolio risk is the main risk of commercial banks. The nonperforming loan caused by the loan allocation mistakes, is the main problem of Chinese banks industry. The scientific optimization of loan portfolio is good for the control of new coming nonperforming loans, and is good for the banks branches to optimizing resources allocation and improving efficiency, is significant for the Chinese financial systems’ steady and developing.The thesis is divided into eight chapters. The first chapter is introduction, is about the background and the meaning of the thesis, the related researches reviews and the research method and technology route and main content. The second chapter analyzes the theoretical ground of the loan portfolio and its optimization. In the third chapter we build a bank loan portfolio optimization model on the ground of upper limits. In the forth chapter, we build a total risk optimization model of portfolio loan on the base of expect return. In the fifth chapter, we build a decision optimization model for accumulate and increment loan portfolio on base of 0-1 regulation. In the sixth chapter, we build a multi-objective decision optimization model for loan portfolio on constrain of VaR. In the seventh chapter, we build a loan’s portfolio optimization model of CVaR minimum based on credit risk transfer. The eighth chapter is the conclusion and expectation.(1) we build a bank loan portfolio optimization model on the ground of upper limitsOn the ground of portfolio selection model, we put forward a bank loan portfolio optimization model on the ground of upper limits, on constrains of bank loan portfolio’s upper limits. One of the characteristics of this model is that the paper put forward an upper bound control that can dispense the credit risk, and also control the loan sum as well, so the commercial can match the assets and the liability at the maturity day. Secondly, this model makes use of the assets changes and structural management ratio to build the loan upper limits, through the bank regulation and preset return of loan portfolio to control the liquidity risk, so the allocation of the loan will meet the demand of regulation and operation. The problem of loan investment’s optimization under the limitation of loan is solved since the model provides the quantity decision in the actual loan investment of bank. (2) we build a optimal model of loan portfolio total risk on the basis of industry portfolio更多还原