進階搜尋


   電子論文尚未授權公開,紙本請查館藏目錄
(※如查詢不到或館藏狀況顯示「閉架不公開」,表示該本論文不在書庫,無法取用。)
系統識別號 U0026-1801201916204800
論文名稱(中文) 銀行授信品質、房地產變數與房貸逾放比率之探討-以L銀行某分行為例
論文名稱(英文) Bank Credit Quality,Real Estate Variables and Housing Loan Overdue Ratio-Evidence from L Bank
校院名稱 成功大學
系所名稱(中) 財務金融研究所碩士在職專班
系所名稱(英) Graduate Institute of Finance (on the job class)
學年度 107
學期 1
出版年 108
研究生(中文) 陳國清
研究生(英文) Kuo-Ching Chen
學號 R87061040
學位類別 碩士
語文別 中文
論文頁數 64頁
口試委員 口試委員-梁少懷
口試委員-謝喻婷
指導教授-林軒竹
中文關鍵字 個人房貸  貸放條件  逾期放款  羅吉斯迴歸 
英文關鍵字 personal mortgage  loan conditions  overdue loans  return to Rogers 
學科別分類
中文摘要 自1991 年政府開放新銀行設立後,國內銀行競爭過度激烈,不論是公營銀行或民營銀行,為爭取市場占有率,銀行授信原則有所鬆動,以致授信品質下降,造成金融機構逾期放款比率呈逐年增加的趨勢。復查近10年間本國銀行逾放比率從2008年1.54%逐年下降至2015年0.23%,2016~2017年間則緩步上升至0.27%~0.31%,截至2018年6月逾期放款比率為0.28%。面臨逾期放款似乎有逐漸惡化現象,金融機構如何健全銀行授信品質,減少逾期放款發生,儼然成為當前極需正視之課題。
2007 年美國發生次級房貸,引發全球金融風暴,次貸的核心問題,即是信用風險,意即借貸雙方資訊不對等,因而導致違約率的不斷攀升,進而造成全世界的金融風暴。為了探討違約放款之行為,本研究藉由探討個人房貸逾期放款所可能產生的影響變數,以某一不動產專業銀行南部某分行為範疇,就2013年至 2017年間曾發生應繳納貸款本息逾期一個月~三個月以上,及目前仍正常還本繳息之房貸戶抽樣案件,採用羅吉斯迴歸模式來進行實證研究。
本研究以銀行的貸放條件為自變數,應變數為個人房貸是否逾放違約。一些基本自變數如貸款利率、貸款成數(LTV)、借款金額及期限與擔保品座落區域,另外加入房地產景氣變化、貸款戶個人特徵之基本資料當作控制變數,如性別、負債負擔比率、學歷、婚姻狀況、職業、個人年收入、保證人之有無等,以控制其對應變數的影響,期望能找出與是否發生違約之間的關係。
實證結果發現,貸款金額大小、擔保品座落地點好壞、借款人教育程度等級及負債負擔比率高低與房貸逾放有顯著因素。貸款成數愈高,違約機率愈低,則與假設不符,個人職業愈穩定、收入愈高及負債負擔比率愈低,則貸款繳息能力佳,違約機率較低,銀行核貸成數相對較高。

實證結果,貸款金額愈高,違約機率愈低,與假設不符;銀行整體通盤考量借款人條件,輔以參酌金融聯合徵信中心個人票債信紀錄與信用卡繳款情形,如評估違約風險低,銀行債權可以確保,貸款金額當然會比較高。
由於個人房貸戶的不同特性,連帶影響銀行授信條件的品質,與是否產生逾期放款的違約。本研究藉由分析逾期放款的共同顯著變數,作為授信風險控管時的另一道防線,以增進銀行優質債權,減少逾期放款產生。
英文摘要 Abstract
Bank Credit Quality,Real Estate Variables and
Housing Loan Overdue Ratio-Evidence from L Bank
Chen- Kuo Ching
Hsuan-Chu Lin
Graduate Institute of Finance,National Cheng Kung University

SUMMARY
The competition of domestic banks has been excessively fierce. In order to gain market share, the principle of bank credit has been loosened, resulting in a decline in the quality of credit, resulting in the overdue ratio of financial institutions overdue. Increased trend. How to improve the quality of bank credit and reduce the occurrence of overdue loans has become a topic that needs to be addressed at present. This research Using the Rogers regression model for empirical research.The bank’s loan conditions are self-variant, and the number of strains is whether the personal mortgage is over-represented. In addition, the basic data of the changes in the real estate boom and the personal characteristics of the loan households are added as control variables .Expectation can find out the relationship between whether or not a default occurs
The empirical results show that the size of the loan, the location of the collateral, the level of the education level of the borrower and the debt burden ratio and the over-release of the mortgage have significant factors. The higher the number of loans, the lower the chance of default, and the assumptions do not match. The higher the loan amount, the lower the probability of default, which is inconsistent with the assumption. This study analyzes the common significant variables of overdue loans as another line of defense when credit risk is controlled to enhance the bank's quality claims and reduce overdue loans.
Keywords: personal mortgage, loan conditions, overdue loans, return to Rogers

INTRODUCTION
The over-release ratio of domestic banks decreased from 1.54% in 2008 to 0.23% in 2015, and gradually increased to 0.27%~0.31% between 2016 and 2017. The overdue loan ratio was 0.28% as of June 2018. This study explores the variables that affect overdue lending and the probability of overshooting, which is used as a credit risk factor. Relevant literatures at home and abroad pointed out the loan interest rate, the number of loans, the length of the loan, the amount of the loan, the collateral area, the changes in the real estate boom, and the personal basic information and conditions of the borrower. Through the analysis of the chi-square test and the return of the Rogers model, it affects the overdue factors and probability of housing loan cases. The empirical results show that there is a significant impact on the overdue of home loan credit cases.

MATERIALS AND METHODS
Based on the collected data, this study organizes the data of the sample data first, and then uses the Stata software to perform statistical analysis according to the hypothesis and structure. The methods used are Narrative, Mean Analysis, Correlations Analysis, Chi-Square Test, Logistic Regression, etc.
This study intends to verify the following hypothesis for the overdue default risk of home loan households:
(1)The longer the customer loan period, the higher the default rate
(2)In terms of where the real estate collateral is located, different geographical locations may affect the risk of credit.
(3)In terms of changes in the real estate environment, different time and space backgrounds will affect the risk of credit
(4)Gender affects the risk of credit
(5)Single or marriage affects the risk of credit
(6)The higher the education level, the lower the default rate
(7)If the service agency is a public official or a larger one, it is inferred that the default rate is lower.
(8)There are guarantors who infer that their default rate is low
(9)The higher the customer loan interest rate, the higher the default rate
(10)The higher the customer loan is, the higher the default rate is.
(11)The higher the customer loan amount, the higher the default rate
(12)The higher the income and the lower the debt ratio, the lower the default rate

RESULTS AND DISCUSSION
Through Logistic Regression, the hypothesis verification results proposed by the Institute are compiled as shown in the table.

This research hypothes is Whether it is established
1 Conforms to the hypothesis, but not significant
2 Conforms to the hypothesis
3 Conforms to the hypothesis
4 Conforms to the hypothesis, but not significant
5 Conforms to the hypothesis, but not significant
6 Conforms to the hypothesis
7 Conforms to the hypothesis
8 Conforms to the hypothesis
9 Conforms to the hypothesis, but not significant
10 Does not match
11 Does not match
12 Debt ratio, Conforms to the hypothesis
Income, Conforms to the hypothesis, but not significant

CONCLUSION
The empirical results show that the size of the loan, the location of the collateral, the level of the education level of the borrower and the debt burden ratio and the over-release of the mortgage have significant factors. The higher the number of loans and the lower the probability of default, the more inconsistent with the assumptions, the more stable the individual occupation, the higher the income and the lower the debt burden ratio, the better the interest rate of the loan, the lower the probability of default, and the higher the number of bank loans.
The empirical result shows that the higher the loan amount, the lower the probability of default, which is inconsistent with the assumptions; the bank considers the borrower's conditions as a whole, supplemented by the personal credit information record and credit card payment of the financial joint credit center.If the risk of default is assessed, bank claims can ensure that the loan amount will of course be higher.
The limitations of this study, the insufficient number of housing loan data, may affect the true prediction accuracy of the model. There is no analysis of the management style and personality traits of the operators, whether it indirectly affects the outcome of the default of the mortgage; the personal data filled by the borrowers cannot actually know the accuracy, so the adverse selection and moral hazard problems cannot be avoided.
The research proposal is based on personal mortgage as the research subject, and the follow-up researchers can conduct research and analysis on other credit granting services such as personal credit loans, credit card auditing and issuing cards. Comparing and analyzing different types of banks, if more samples can be obtained from private banks and public banks and foreign banks, the correctness of the empirical results of this study can be verified.
論文目次 第一章 緒論.............................................1
第一節 研究背景與動機..................... .............. 1
第二節 研究目的.. .......... ..... ..................... 3
第三節 研究範圍.. ...................................... 4
第二章 文獻探討......................................... 5
第一節 銀行授信評估原則.................................. 5
第二節 逾期放款之定義.................................... 9
第三節 房地產變數對銀行逾放比的影響................. .... . 11
第四節 其他影響房貸逾放變數...... ........................ 13
第三章 研究方法................. ........... ........ ...19
第一節 研究架構及研究假設... ..............................19
第二節 研究變數及操作性定義.. ... .................. .......23
第三節 研究樣本...... .......... ..... ...................26
第四節 資料分析及統計分析方法 ................... ..........28
第四章 實證分析與結果. ...................................32
第一節 樣本結構分析.............................. ........32
第二節 平均數分析.......... ...... .......... ... ...... .42
第三節 各變數與房貸逾放違約差異性之檢定................ ......45
第四節 相關分析....... ..... .... ................ .......47
第五節 羅吉斯迴歸分析.......... . ................ ........50
第六節 研究假說與實證結果分析............... ...............54
第五章 結論與建議................ ......... ..... ........58
第一節 研究結論.............. ... ........... ..... ......58
第二節 研究限制及建議... ... ..................... ..... ..60
參考文獻..................... ... ......... ....... .....61
國內部分........... ......... ... ........... ..... .....61
國外部分........... ......... ... ......... ....... .....63

表目錄
表3.1 本研究變數之操作性定義... ... .............. .........24
表3.2 樣本變數分析彙整表............ .......... ...........27
表 4.1 貸款利率在正常戶、違約戶之間的交叉分析表...... .........33
表 4.2 貸款成數在正常戶、違約戶之間的交叉分析表.. .............33
表 4.3 貸款年限在正常戶、違約戶之間的交叉分析表.. .............34
表 4.4 貸款金額在正常戶、違約戶之間的交叉分析表.. .............35
表 4.5 擔保品區域在正常戶、違約戶之間的交叉分析表.. ...........35
表 4.6 房地產景氣變化在正常戶、違約戶之間的交叉分析表.. ........36
表 4.7 性別在正常戶、違約戶之間的交叉分析表.............. ....37
表 4.8 婚姻狀況在正常戶、違約戶之間的交叉分析表......... . ....37
表 4.9 教育程度在正常戶、違約戶之間的交叉分析表........ .......38
表 4.10職業別在正常戶、違約戶之間的交叉分析表............ ....39
表 4.11 所得水準在正常戶、違約戶之間的交叉分析表.......... . ..39
表 4.12 負債負擔比率在正常戶、違約戶之間的交叉分析表........ ...40
表 4.13 有無保證人在正常戶、違約戶之間的交叉分析表......... ....41
表 4.14 貸款利率平均數分析.......... .. ..... ...............42
表 4.15 貸款成數平均數分析... ...... .. ..... ...............43
表 4.16 貸款金額平均數分析... . .... .. ... ..... ...........43
表 4.17 借款人所得平均數分析... . .. .. .. ....... ..........44
表 4.18 負債負擔比率平均數分析... . ... .. .. ....... .......44
表 4.19 各變數對房貸是否違約差異性卡方檢定彙整表. ..............45
表 4.20 各變數之相關性分析(Pearson) ... . .. .. .......... ..47
表 4.21 各變數之相關性分析(Pearson) ... .... ........... ....47
表 4.22 羅吉斯迴歸分析彙整表(1) ... ....... .. ......... ....50
表 4.23 羅吉斯迴歸分析彙整表(2) ... ....... .. ......... ....52
表 4.24 本研究假說檢定結果彙整... . .... .. .... ......... ..56

圖目錄
圖3-1 研究架構........................ ... ......... ..... 20
圖3-2研究模型................... ....... ......... ..... ..20
圖3-3 Logistic迴歸函數曲線圖....... ...... ......... ... ...30
參考文獻 一、 國內部分
1、江百信及張金鶚,購屋貸款核貸條件,住宅學報第3期,P1-20,1995。
2、李馨萍,住宅抵押貸款違約風險之實證分析,國立臺灣工業技術管理技術研究所,博士論文,1997。
3、87年台上字第1346號,最高法院民事裁判書彙編第 32 期,P811-816 頁,1998。
4、李桐豪、呂美慧,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 ; 1卷1期 , P1 - 20 ,2000。
5、江惠櫻,商業銀行對企業授信決策考量因素與授信品質之關係,管理評論 ; 第22卷第2期 , P1 – 23,2001。
6、林左裕,台灣住宅抵押貸款終止行為之研究 ,農業經濟半年刊 ; 76期 , P169 – 195,2004。
7、林耀宗,房屋抵押貸款之資訊不對稱問題-以台北市和新北市為例,國立政治大學地政研究所,碩士論文,2013。
8、張鼎煥,陳璦玲,銀行住宅貸款授信品質研究-預期違約機率之應用,明道學術論壇;10卷3期,2017/7。
9、彭淮南,立法院財委會進行「健全房市租稅措施、不動產授信風險控管之成效及精進作為」專題報告,2013。
10、邱于修,周美伶及張金鶚,購屋者投資機率預測模型之探討,臺大管理論叢第23卷第2期,2013/6。
11、楊顯爵,林左裕,陳宗豪,住宅抵押貸款違約之影響因素之顯著性分析,台灣土地研究,第11卷第2期,P1-P36,2008/11。
12、楊謹鍹,影響房屋貸款逾期因素之實證分析,國立臺中科技大學企業管理系事業經營研究所,碩士論文,2013。
13、郭姿伶,住宅貸款之提前清償與逾期還款,國立中正大學財務金融研究所,碩士論文,2000。
14、陳宗豪,住宅抵押貸款逾期率之廣義線性模型研究,國立高雄第一科技大學管理研究所,博士論文,2007。
15、陳麗珍,不動產抵押貸款授信品質之研究,朝陽科技大學財務金融系,碩士論文,2011。
16、陳鈞鎮,房貸違約風險因子分析,清雲科技大學國際企業管理研究所,碩士論文,2011。
17、顏貽謙,台灣地區房價與銀行逾放比關係之研究,國立中正大學國際經濟研究所,碩士論文,2006。
18、蕭浩彬,抵押品對銀行風險之影響,國立中正大學經濟系國際經濟研究所,碩士論文,2012。
19、蘇家郁,房屋貸款違約風險與銀行獲利分析,國立臺灣大學經濟學研究所,碩士論文,2015。
20、金融監督管理委員會銀行局,金融統計指標-金融機構國內總、分行放款餘額;逾期放款餘額及逾期放款比率,2018/6。
二、國外部分
1、Bester, H. Screening vs.Rationing in Credit Markets with Imperfect
Information.American Economic Review 75. (1985):850-855.
2、Deng, Y. H., J. M. Quigley, and R. Van Order, (1996), “Mortgage defaultand low
downpayment loans: The Costs of public subsidy,” RegionalScience and Urban
Economics, Vol.26, pp.263-285.
3、Jung, A. F. (1962). Terms of conventional mortgage loans on existing
houses.Journal of Finance, 17(3), 432-443.
4、Kelly, A.,1995 "Racial and Ethnic Disparities in Mortgage Prepayment", Journal of Housing Economics,4, 350-372
5、 Kau, J. B., & Keenan, D. C. (1999). Patterns of rational default. Regional Science and Urban Economics, 29(6), 765-785
6、Lawrence, E. C., L. D. Smith, and M. Rhoades, (1992), “An Analysis of Default Riskin Mobile Home Credit,” Journal of Banking and Finance,Vol.16, pp.299-312.
7、Page, A. N. (1964). The variation of mortgage interest rates. Journal of Business,37(3), 280-294.
8、Von Furstenberg, G. M. (1969). Default risk on FHA-insured home mortgages as afunction of the terms of financing: A quantitative analysis. Journal of Finance,24,459-477.
9、Vandell, K. D. (1978). Default risk under alternative mortgage instruments. Journalof Finance, 33(5), 1279-1296.
10、Vandell, K. D., & Thibodeau, T. (1985). Estimation of mortgage defaults usingdisaggregate loan history data. AREUEA Journal, 15(3), 292-317.
論文全文使用權限
  • 同意授權校內瀏覽/列印電子全文服務,於2024-07-07起公開。
  • 同意授權校外瀏覽/列印電子全文服務,於2024-12-20起公開。


  • 如您有疑問,請聯絡圖書館
    聯絡電話:(06)2757575#65773
    聯絡E-mail:etds@email.ncku.edu.tw