透過您的圖書館登入
IP:3.91.19.28
  • 學位論文

運用結構法模型評估房貸的信用風險

CREDIT RISK MEASUREMENT USING STRUCTURAL MODEL APPRAISE THE REAL ESTATE MORTGAGE DEBT

指導教授 : 梁世安
若您是本文的作者,可授權文章由華藝線上圖書館中協助推廣。

摘要


傳統的銀行實務,對房貸信用風險的評估,其衡量時點只侷限在申貸時,對於核貸後信用風險的變化則難以顧及,這使得銀行難以隨時針對現存的房貸債權評估其信用風險,也很難隨時針對情況,進行有效的信用風險管理,因此本研究嘗試探討運用結構法模型,來為銀行所承做的房貸債權,評估其信用風險,以期改善。 本研究之目的主要為探討結構法模型在理論層面上適用房貸債權的信用風險評估,並希望研究模型所判別的違約機率、件數及違約金額比率,能與實際情況吻合。此外,亦期待研究模型可以靈敏而快速地補捉出微小變化,以提出預警訊號,俾作為銀行信用風險管理之參考。在方法上本研究在不動產價格變動為幾何布朗運動並服從對數常態分配的假設下,針對長期分償債權的評估及其違約點的設定,提出方法,俾使模型能更適用於評估長期分償的房貸債權。 本研究資料為某銀行截至96年6月30日止尚未結清的全部不動產抵押資料,共有20,627筆,其最遠的一筆,初貸日期為75年6月14日。為與都市地價指數配合,爰將75年6月14日起至81年6月30日止共164筆資料刪除,最後總筆數為20,463筆。 經以本研究模型評估上開研究資料,所得到的違約情況與實際情形相當吻合,因此本研究結論認為結構法模型在理論層面上相當適合用於房貸債權的信用風險評估。這代表兩個重要的意涵:第一、為結構法模型開拓了新的貸款信用評估領域;第二、為銀行實務開發了更即時且具前瞻性的房貸信用風險評估方法。基於研究結果,本研究認為銀行應更加重視授信續後管理、要重視房貸利差偏低的情況,並建議銀行間必需建立不動產估價資訊的交換平台,以利不動產價格更加透明。

並列摘要


The bank appraised the credit risk measurement of real estate mortgage debt , its point in time is when application loan. After change then takes into consideration with difficulty , this causes the bank as necessary to aim at the extant loan creditor's rights to appraise its credit risk with difficulty and very difficult to aim at the situation as necessary, carries on the effective credit risk management. Therefore this research attempt discussion utilization structural model appraisal the credit risk measurement of real estate mortgage debt for improvement. This research goal is suitable for the confirmation structural model for the loan credit risk appraisal, breaks a contract the situation by the time accurate distinction to propose the early warning. For appropriate appraisal, in method this research regarding: The real estate price proposed the change process explanation, in view of divides the creditor's rights for a long time which recompenses and its breaks a contract a hypothesis, proposed the theoretically view, to the time reaches the cost research the goal. This research data on 2007/6/30 all the real estate mortgage for some bank which yet settles account, the data quantity has 20,627, the earliest, initially loans the date is 1986/6/14. For with the Index of Urban Land Price coordination, whence on 1986/6/14 to 1992/6/30 stops the data quantity has 164 to delete, finally the total pen number is 20,463. After by this research model appraised on opens the research material, obtained arrives breaks a contract the situation and the actual situation quite tallies, therefore this research conclusion reveals structural model true suitable real estate mortgage debt of credit risk assessment. This represents two important significance: First, has developed the new loan credit appraisal domain for the structural model; Second, has developed immediate also foresightedness real estate mortgage debt of credit risk assessment method for the bank. Based on the findings, after this research thought the bank should even more take to a loan and continue manages, must take interest rate is somewhat low situation, and suggested between bank essential establishment real estate estimate information exchange platform, is more transparent in order to help the real estate price.

參考文獻


李桐豪、呂美慧(2000),「金融機構房貸客戶授信評量模式分析-Logistic 迴歸之應用」,台灣金融財務季刊,pp.1-20
李彥仕(2006),「影響現金卡違約風險之因素分析」,國立中央大學資訊管理研究所,碩士論文
鄭歆蕊(2007),「兩階段預警模型之研究-以台南市房貸為例」,國立成功大學,碩士論文
鄭婷月(2003),「汽車貸款客戶之風險研究」,國立成功大學統研所,碩士論文
劉展宏、張金鶚(1999),「一般購屋貸款與首次貸款提前清償之比較研究」,中華民國住宅學會第八屆論文集, pp.177-195

被引用紀錄


陳璦玲(2014)。從不動產預期違約機率探討銀行授信風險管理品質〔碩士論文,淡江大學〕。華藝線上圖書館。https://doi.org/10.6846%2fTKU.2014.01061
梁原溢(2010)。住宅抵押貸款違約損失率之實證研究--以臺北縣市為例〔碩士論文,國立臺北大學〕。華藝線上圖書館。https://www.airitilibrary.com/Article/Detail?DocID=U0023-2801201001331500
許章億(2011)。房屋貸款違約因素之探討—以M銀行為例〔碩士論文,元智大學〕。華藝線上圖書館。https://www.airitilibrary.com/Article/Detail?DocID=U0009-2801201414584073

延伸閱讀