住宅抵押貸款違約損失率之實證研究--以臺北縣市為例

An Empirical Analysis of Loss Given Defult on Residential Mortgage Loans - Using Taipei City and Taipei County as Examples

梁原溢(Yuan-Yi Liang)

指導教授 ： 吳庭斌

國立臺北大學/商學院/統計學系/碩士(2010年)

摘要

由於1992年台灣政府味了加速金融自由化，以及準備邁入已開發國家之林，主管機關財政部陸續核發了15家新銀行的營業執照，開啟了國內金融業的戰國時代；但是國內金融業同質性過高，經營與創新能力不足，在激烈的市場競爭、利率自由化及資訊透明化之氛圍下，不得不走向價格競爭，甚至犧牲授信品質以換得獲利成長。近年來金融機構存放款利差逐漸縮小，為了追求獲利及生存空間，不斷擴增顧客群，因而放寬授信尺度進而承受更多的潛在性風險，加上近年來國內外經濟景期持續低迷不振，逾期放款金額將帶來可觀的損失。
根據中央銀行編纂之「金融統計月報」統計資料顯示，在消費者貸款業務當中，以購置房屋貸款加計房屋修繕貸款佔整體比重最大，已經超過80％，顯示房屋類貸款為銀行個人授信最主要也是最重要的曝險來源，亦隱含銀行放款過度集中於房屋類貸款。從實務上來看，房屋貸款因為有房屋作為擔保品，償還來源較明確，也是最近幾年歷經雙卡風暴後，銀行積極爭取的業務。
本研究中將利用金融機構實際發生之損失情形的歷史資料，並以符合新巴塞爾協定-內部評等法風險數量化估計之LGD（Loss Given Defult）以及其最低基本作業要求作一實證研究。透過此研究其能提供金融機構在資本計提上及使用內部評等法估計違約損失率時之參考。我國於2006年底實施Basel Ⅱ，而Basel Ⅱ協定內容中鼓勵銀行建立其內部模型，即採用IRB法，期望金融機構以健全的風險評等系統，落實售信業務之風險管理與公司治理與監督，以期能達到穩健經營之目的。
Basel Ⅱ鼓勵金融機構建立其內部模型，採行IRB法，期望金融機構以健全的風險評估系統，落實授信業務與公司治理與監督，以期能達到穩健經營的目的。本研究針對消金房貸借款者，分析損失率依其高低分類，並結合風險評估模式將借款客戶分為數個風險等級，以做為金融機構及銀行放款時評等借款客戶風險的依據。並據以作為資本計提計算中風險成分數量LGD之參數值。本研究初步以行政區域做分類，後續研究者若能取得足夠之資料，可再對不動產型態做分類。

關鍵字

新巴塞爾協定 ； 違約損失率 ； 住宅抵押貸款

並列摘要

In 1992, in order to accelerate financial liberalization and prepare to join one of the developed countries, the ministry of finance in Taiwan granted the business license for 15 new banks in succession. This sets the battlefield for domestic financial institutions. However, due to the homogeneity and lack of managerial and innovative abilities, fierce competition, pricing freedom and information transparency in the financial sector, price competition or even giving up credit quality are necessary for profit and growth. Interest margin is narrowing in recent years, and with the purpose of being profitable and going on, expanding customer base resulted from loose credit criteria will incur more potential risk. In addition continuous suppressed economy home and abroad will bring considerable loss to non performing loans.
According to the Financial Statistics Monthly published by Central bank, mortgage loans account for the majority of consumer banking with more than 80% of the pool. This shows that mortgage loans are the prime and most important exposure to personal credit; implying banks are over-exposed to mortgage loans. From the view point of practice, mortgage loans are backed by house as collateral, so the source of recovery is more certain. That is why banks are striving for it after recent credit card crisis.
The research will utilize actual historical loss experience happened in banks to meet LGD quantification and minimum requirement set in BASEL II IRB approach. With this research will serve financial institutions as a reference for capital requirement and estimate of LGD under IRB approach. BASEL II were implemented in 2006, the Standards encourages banks to set internal models, i.e. IRB approach, and expects financial institutions uses robust risk rating system to facilitate corporate governance and supervision when granting credits, with the aim of stabilizing business operations.
The research focuses on retail mortgage loan borrower, analyses loss based on severity, and combines risk assessing mode while segmenting customers into several grades. And use this as a reference when assessing customer risk in loan approval process. The parameter for LGD quantification in capital requirement also grounds on this paper. Administrative areas are the initial segmenting criterion in this research, and if more data are available in the future, realties will be the next segmentation.

並列關鍵字

Basel Ⅱ ； Loss Given Defult ； Residential Mortgage Loans

參考文獻

張金鶚（2003）房地產投資與市場分析-理論與實務，華泰文化事業股份有限公司。

吳顯輝（2007）運用結構法模型評估房貸的信用風險，台北大學國際財務金融碩士論文

陳盛如（2007）二十年房貸提前清償風險之研究所，清華大學統研所碩士論文

戴月華（2005）預測現金卡與持有現金卡的房貸戶違約風險之研究，元智大學管理研究所碩士論文

張大成（2003）「違約機率與信用評分模型」台灣金融財務季刊第四輯第一期