基于bp人工神經(jīng)絡的商業(yè)銀行信用風險研究

1、1商業(yè)銀行信用風險評估模型研究——基于BP神經(jīng)網(wǎng)絡風險評估的改進天津財經(jīng)大學刑春玉、安菁菁、陳翔穎目錄摘要..................................................................................................................................................2一、商業(yè)銀行信用風險評估體系......

2、..............................................................................................4二、銀行信用風險評估模型發(fā)展....................................................................................................5（一）基于貸款客戶資

3、產價值數(shù)據(jù)的風險評估模型............................................................51.KMV模型.......................................................................................................................52.CreditMetrics模型...

4、..............................................................................................63.CreditPtfolioView模型...................................................................................64.CreditRiskPlus模型..

5、...........................................................................................7（二）基于貸款客戶財務數(shù)據(jù)的風險評估模型....................................................................7（三）商業(yè)銀行財務風險的評估模型有待改進...................

6、.................................................81.我國商業(yè)銀行現(xiàn)有評估財務風險的方法....................................................................82.國際商業(yè)銀行普遍使用的風險評估方法...............................................................

7、.....9三、信用風險評估的實證分析......................................................................................................10（一）YH銀行采用的貸款違約風險評估模型及評估結果...............................................101.計算資產價值和資產波動率.........

8、.............................................................................102.計算違約距離和期望違約率......................................................................................12（二）基于BP神經(jīng)網(wǎng)絡信用風險估計......................

9、.........................................................141.樣本選取......................................................................................................................142.實證分析.............................

10、.........................................................................................15四、對商業(yè)銀行信用風險評估提出的建議..................................................................................21（一）結合KMV法與BP神經(jīng)網(wǎng)絡模型評估財務風險...

11、................................................21（二）逐步推進內部評級法..................................................................................................22（三）建立使用預期損失模型.............................................

12、.................................................231.預期損失模型應關注謹慎性和中立性......................................................................232.預期損失模型提高相關性兼顧可靠性.............................................................

13、.........23（四）完善財務風險披露要求............................................................................................241.提供整體層面的風險信息.........................................................................................

14、.242.報告多種損失的分布結果..........................................................................................243.提高信息披露的全面性..............................................................................................25（五）通

15、過健全內部控制降低財務風險防范成本..............................................................25（六）建立非上市公司信用風險評估模型..........................................................................26（七）針對不同的貸款類型選擇適當?shù)娘L險評估方法......................

16、................................26五、總結..........................................................................................................................................26【參考文獻】....................................

17、..............................................................................................283WiththeeconomicdevelopmentofTianjinBinhaiNewAreathecommercialbankshaveoperationsintheBinhaiNewAreatoprovidefinancialservice

18、stocompaniesintheregion.Riskassessmentisanimptantpartofthecommercialbanktomakeinternalcontroltoestablishthevaluesofthecompanies.Thisresearchmethodmakesuseofthemeasurementstatisticsstructuredresearchmethodstoevaluatethecr

19、editrisk.InstructuringresearchweuseKMVmodeltogettheprobabilityofdefaultof11listedcompaniesinBinhaiNewArea.Inthemeasurementofstatisticalresearchthisarticleuse18companiesedassamplesthataresimilartotheabove11companiesusingB

20、Partificialneuralwkmodeltoassessthecreditriskofthe11listedcompanies.AssessmentresultsshowthatBPneuralwkmodelfittedbetterthantheKMVmodel.Sincethevariousmodelshaveshtcomingscommercialbanksshouldincreasethediversityofmodels

21、maketheuseoftheadvantagesofvariousmodels.ThispaperalsoproposedusingBPartificialneuralwkmodelcombinedwithKMVmodeltoevaluatethecreditriskofnonlistedcompaniestomakeupthatnonlinearmodel’sweakness.Keywds:CreditRiskBPartificia