【作者】 魏强；

【导师】 秦学志；

【作者基本信息】 大连理工大学 ， 技术经济及管理， 2012， 博士

【摘要】 随着我国利率市场化进程的日益加快以及资本充足率监管力度的不断加强,如何进行有效的资本配置,即为贷款业务确定合理的风险限额成为银行关注的重点问题。风险限额是在内部评级法量化信用风险的基础上,通过资产组合模型,并根据风险调整后的资本收益率最大化原则,确定的风险敞口上限；同时,巴塞尔委员会明确指出,信贷组合的集中度风险是造成银行危机的一个主要原因,是监管资本及经济资本的重要影响因素。因此,构建基于RAROC (即经风险调整后的资本收益率)和集中度约束的信贷组合优化配置模型,具有重要的现实意义和理论价值。虽有不少学者在相关方面取得了有价值的研究成果,但仍存在一些有待进一步完善之处,如：信贷组合相关性的估计,受制于数据不足、期限长度不够及代理变量取值难等现状；经济资本的计量,存在假设条件多、精确性较差等不足；组合集中度的计量,存在部分参数只能由模拟产生等不足；信贷限额的配置,存在视角过于微观、前瞻性不足等缺陷。在借鉴现有研究成果的基础上,本文主要取得了如下研究成果：(1)通过构建RAROC因子模型来计量信贷组合相关性,以体现宏观经济环境对银行资产收益的影响。基于RAROC涵义,将影响贷款利率的若干经济因素视为RAROC的系统风险因子,即根据凯恩斯和希克斯的经典理论,通过分析IS-LM模型,提炼出四个影响银行资产收益的系统风险因子：经济增长、物价水平、货币供给量和投资额度,据此构建了RAROC因子模型,并给出信贷组合协方差矩阵的计算公式。进一步,以S银行某分行数据进行实证分析,得到了信用评级、行业和企业规模三个维度下信贷组合的RAROC协方差矩阵及相关系数矩阵。(2)利用S银行某分行的1448笔对公贷款数据和X银行总行的4113笔对公贷款数据,分别测度了它们在信用评级、行业和企业规模三个维度下的集中度,并构建了相应维度下考虑集中度约束的信贷组合优化配置模型。使用VBA程序设计了一个组合集中度计量模块,测度了S银行某分行和X银行总行的信贷组合集中度及其变化状况,据此分别构建了信用评级、行业和企业规模三个维度下考虑集中度约束的信贷组合优化配置模型,并进行了压力测试。(3)将上述构建的三个维度下的模型与不考集中度约束的模型进行实证比较,从一个侧面检验与对比分析了上述模型的改进效果。主要结论为：①各维度下,无论是否考虑集中度约束,随着银行设定的收益目标的增加,组合集中度风险总是随之增加,意味着此时贷款资源的配置也趋于集中；同时,组合风险与收益目标呈正相关性。②各维度下,集中度约束会使银行收益适度变小。③无论是否考虑集中度约束,中间级别的信贷组合RAROC值最大,对银行最具吸引力；批发和零售、工业、房地产这三个行业类信贷组合的配置权重相对较大；值得一提的是：目前利率市场化时机尚未成熟,银行仍可享受息差保护政策的红利,导致大型企业贷款客户更受银行青睐,但从企业规模维度分析知：RAROC随企业规模的递减而递增,表明经营中小企业客户的资本效率远高于大企业,因此可以预计：伴随利率市场化和资本约束的加强以及政策的倾斜,银行的战略转型必将向中小企业倾斜。另外,④压力测试表明：信用评级为7的借款企业、房地产行业的借款企业或小型借款企业在遭遇极端危机时,给银行信贷组合带来的风险最大,应引起S银行特别关注,需结合风险偏好及差异化经营策略,对贷款资源进行优化配置。总之,本文的创新点主要包括：(1)为刻画信贷组合收益的相关系数矩阵并考察银行资本效率的影响因素,构建了考虑若干宏观经济指标的RAROC因子模型,相应指标有：经济增长、物价水平、货币供给量和投资额度等。(2)将RAROC和集中度风险作为约束条件引入信贷组合优化配置模型,并分别从信用评级、行业和企业规模三个维度对比分析了信贷组合最优配置方案,有力地保证了银行贷款不至于过度集中于某些信用级别的行业或某种规模的企业,同时保障银行贷款组合获得较好的收益。(3)考虑了银行贷款对各行业经济增长的贡献率及与区域经济结构的差异性、协调性,据此给出了兼顾银行自身发展和有利于促进区域经济增长的银行贷款优化配置方法。更多还原

【Abstract】 As the marketization of interest rate in China is accelerating, banks put more and more concerns on capital adequacy ratio, i.e. how to conduct an effective allocation of capital, which means how to determine reasonable risk limits for its loans. Risk limit is the maximum exposure which is quantified in the Internal Ratings-Based Approach based on the principle of maximization of risk-adjusted return on capital according to portfolio model. Meanwhile, the Basel Committee declared clearly that the concentration risk of credit portfolio is one of the major causes of banking crisis and an important factor of regulations on capital and economic capital. Therefore, a risk limits optimal allocation model of credit portfolio based on RAROC (short for Risk Adjusted Return on Capital) and concentration constraints is of great practical and theoretical value.Although many scholars have obtained valuable research results in the relevant fields, there are still some aspects should be further improved. For example, the estimation of correlation in credit portfolio is subject to condition like the lack of data, not long enough of the period length and difficulties in valuing the proxy variables. Too more assumptions used in the measurement of economic capital result in a lack of accuracy of the model; some parameters in the measurement of portfolio concentration can only be generated by simulation; the viewpoint is too narrow and there is a lack of forward-looking in the study of allocation of exposure limits.Based on the results of existing research, this paper achieved the following findings:(1) A RAROC factor model to measure the relevance of the credit portfolio is constructed to reflect the impact of macroeconomic environment on bank assets.Based on the definition of RAROC, elements which can affect the credit rates are set as the RAROC systematic risk factors. It means that based on the classical theory of Keynes and Hicks, this paper build a RAROC factor model by using four systematic risk elements related to the asset return of banks, which are economic growth, price level, money supply and investment amount according to analyzing IS-LM model, and provide a calculating formula for covariance matrix of credit portfolio. Then this paper brings up an empirical research based on the data from a branch of Bank S and obtains the RAROC covariance matrix and correlation matrix under three dimensions which are credit rating, industry and firm scale of the credit portfolio. (2) According to1448corporate loans of the branch of Bank S and4113corporate loans of Bank X, this paper measures the concentration of the two banks respectively under dimensions of credit rating, industry and firm scale, and a credit portfolio optimization model considering concentration constraints is built under relevant dimensions.This paper designs a combination of concentration measurement module by using VBA program to measure the credit portfolio concentration and changes in certain condition of the branch of Bank S and Bank X. Accordingly, this paper builds credit portfolio optimization models considering concentration constraints respectively under credit rating, industry and firm scale, and conducts a stress test.(3) By conducting an empirical comparison between the models under the above three dimensions and models without concerning constraints of the degree of concentration, this paper is to find the effects of the improved model through comparative analysis from a certain perspective.The main conclusions are as follows:①Under each dimension, whether this paper takes the constraints of the degree of concentration into consideration or not, with the increase of objective earnings set for the bank, the portfolio concentration risk of the bank is always increased, meaning that the allocation of loans at this time tends to be concentrated; meanwhile, the portfolio risk and objective return are positively correlated.②Under each dimension, the concentration constraints will decrease the bank income moderately.③whether this paper takes the constraints of the degree of concentration into the consideration or not, the credit portfolio in the middle level rating has the maximum value of RAROC, which means it attracts the banks most; the allocation of credit portfolio weighs relatively large in the industries like wholesale and retail, industrial and real estate; it should be noticed that the current interest rate market is not mature enough, the banks can still enjoy the bonus from the interest rate differential protection policy leading banks to have a more positive tendency to allocate a loan to large enterprise customers. However, under the analysis of firm scale dimension, RAROC is negatively related to the firm scale showing that capital efficiency of small and medium business enterprise customers is much higher than the large ones. So we can expect that with the accelerating of the interest rate marketization, strengthening of capital constraints and a favorable policy, bank’s strategic transformation will tilt to SMEs.④Stress test indicates that when the extreme crisis occurs, the credit rating of7borrowers, real estate industry or small business enterprises may bring largest risk to the credit portfolio of banks, which needs particular concern for Bank S. It requires a combination of risk preferences and differences in management strategies to optimize the allocation of credit resources.In short, innovations of this paper include: (1) A RAROC factor model considering a number of macroeconomic indicators is built to characterize the correlation matrix of the loan portfolio returns and examine the impact factors related to the efficiency of bank capital. These macroeconomic indicators include economic growth, price level, money supply and the investment amount, etc.(2) This paper adds the concentration risk as a constraint to the credit portfolio allocation optimization model and conducts related comparative analysis about the optimal allocation of the loan portfolio under three dimensions, which are the credit rating, industry and firm scale, respectively. The results theoretically guarantee bank loans are not overly granted in certain sectors or companies of certain credit rating, and make sure to obtain a satisfactory return on credit portfolio.(3) This paper takes into consideration that bank loan’s influence on the contribution rate of economic growth in different industries, regional differences in economic structure and coordination, and it draws a conclusion that an optimized allocation of bank loans can help to promote the regional economic growth.更多还原