系统工程理论与实践
繫統工程理論與實踐
계통공정이론여실천
Systems Engineering—Theory & Practice
2012年
2期
257~267
,共null页
贷款组合 组合优化 收益率峰度 均值-方差-偏度-峰度模型 峰度控制
貸款組閤 組閤優化 收益率峰度 均值-方差-偏度-峰度模型 峰度控製
대관조합 조합우화 수익솔봉도 균치-방차-편도-봉도모형 봉도공제
bank loans portfolio; portfolio optimization; yield kurtosis; mean-variance-skewness-kurtosismodel; kurtosis control
以银行各项资产组合收益率最大化为目标函数,以VaR来控制贷款组合的风险价值,以偏度约束来控制贷款组合收益率的整体分布向大于均值的方向倾斜、以减少发生总体损失的单侧风险,以峰度来控制贷款组合收益率分布出现极端情况的双侧风险,建立了资产分配的收益率均值-方差-偏度-峰度模型.本模型的创新与特色是通过峰度约束控制了贷款组合收益率向极端损失偏离的程度.在马可维茨均值-方差模型的基础上,增加了偏度和峰度参数,建立了收益率均值.方差-偏度-峰度模型.模型通过方差约束,控制了组合收益率偏离均值的离散程度:通过偏度约束,控制了组合收益率总体分布向损失-侧偏离的程度:通过峰度约束,控制了组合收益率出现极端损失或收益的可能性.模型从多个角度控制了贷款组合的风险,拓展了经典的均值-方差优化组合思路.
以銀行各項資產組閤收益率最大化為目標函數,以VaR來控製貸款組閤的風險價值,以偏度約束來控製貸款組閤收益率的整體分佈嚮大于均值的方嚮傾斜、以減少髮生總體損失的單側風險,以峰度來控製貸款組閤收益率分佈齣現極耑情況的雙側風險,建立瞭資產分配的收益率均值-方差-偏度-峰度模型.本模型的創新與特色是通過峰度約束控製瞭貸款組閤收益率嚮極耑損失偏離的程度.在馬可維茨均值-方差模型的基礎上,增加瞭偏度和峰度參數,建立瞭收益率均值.方差-偏度-峰度模型.模型通過方差約束,控製瞭組閤收益率偏離均值的離散程度:通過偏度約束,控製瞭組閤收益率總體分佈嚮損失-側偏離的程度:通過峰度約束,控製瞭組閤收益率齣現極耑損失或收益的可能性.模型從多箇角度控製瞭貸款組閤的風險,拓展瞭經典的均值-方差優化組閤思路.
이은행각항자산조합수익솔최대화위목표함수,이VaR래공제대관조합적풍험개치,이편도약속래공제대관조합수익솔적정체분포향대우균치적방향경사、이감소발생총체손실적단측풍험,이봉도래공제대관조합수익솔분포출현겁단정황적쌍측풍험,건립료자산분배적수익솔균치-방차-편도-봉도모형.본모형적창신여특색시통과봉도약속공제료대관조합수익솔향겁단손실편리적정도.재마가유자균치-방차모형적기출상,증가료편도화봉도삼수,건립료수익솔균치.방차-편도-봉도모형.모형통과방차약속,공제료조합수익솔편리균치적리산정도:통과편도약속,공제료조합수익솔총체분포향손실-측편리적정도:통과봉도약속,공제료조합수익솔출현겁단손실혹수익적가능성.모형종다개각도공제료대관조합적풍험,탁전료경전적균치-방차우화조합사로.
By using VaR as risk control of the loans portfolio, using skewness constrain to avoid the distribution of loan portfolio yield toward left of mean to reduce left side risk of general risk, using kurtosis constrain as the control of the distribution's fat tail on both sides to reduce the extreme loss, the optimal model of loan portfolio which targets the maximum rate of return on bank loans portfolio based on the higher central-moment constraints is set up. The contribution of this article is we identified the importance of using higher central-moments, especially the kurtosis in bank loans portfolio optimization. Addition to the classic Markowitz model, we build a mean-variance-skewness-kurtosts model which introduced kurtosis constrain to reduce the extreme loss, skewness constrain to avoid general risk and VaR as risk control of the loans portfolio. The model we built controls the portfolio's risk from multi-angle and extends the classic mean-variance optimal theory.