Publications-Theses
Article View/Open
Publication Export
-
題名 擔保房貸憑證(CMOs)之評價：應用類神經網路預測提前還款率
Pricing Collateralized Mortgage Obligations: Using Neural Network to Forecast Prepayment Rate作者 張憲明
Chang, Hsien-Ming貢獻者 林士貴<br>陳亭甫
Lin, Shih-Kuei<br>Chen, Ting-Fu
張憲明
Chang, Hsien-Ming關鍵詞 房屋抵押貸款證券化
擔保房貸憑證(CMOs)
提前還款
類神經網路
對數常態遠期利率(Lognormal Forward LIBOR Model, LFM)市場模型
Mortgage securities
Collateralized mortgage obligations (CMOs)
Prepayment
Neural network
Lognormal forward LIBOR model日期 2018 上傳時間 3-Sep-2018 15:48:23 (UTC+8) 摘要 本研究主要透過使用類神經網路的方法來預測擔保房貸憑證(CMOs)之提前還款風險並加以評價，並且與另外兩種模型進行比較，PSA/CPR模型和美國官方(Office Thrift Supervision; OTS)30年期固定利率住宅抵押貸款動態提前清償模型，其中PSA/CPR模型為靜態模型，OTS模型為動態模型，另外由於實證個案現金流與LIBOR利率有關，因此採用的利率模型為對數常態遠期利率(Lognormal Forward LIBOR Model, LFM)市場模型，且因為擔保房貸憑證(CMOs)涉及提前還款風險以及眾多風險，並無法找出封閉解，因此採用蒙地卡羅模擬法來作為評價模型。透過蒐集Fannie Mae公開的房屋抵押貸款資料來做實證，實證結果有以下貢獻，第一，類神經網路預測提前還款率均方誤差(MSE)小於PSA/CPR模型的均方誤差，所以類神經網路在預測提前還款率的方面優於PSA/CPR模型。第二類神經網路模型得出個案的發行機構低估發行當下所設定之提前還款率，而動態模型的OTS模型也得到相同結果，綜合以上兩點，得到類神經網路模型預測提前還款率的評價結果比PSA/CPR模型更接近真實價格。
This paper used the Neural Network to forecast the prepayment rate of Collateralized Mortgage Obligations (CMOs), and then pricing it. Comparing the Neural Network Model with other prepayment model: PSA/CPR (static) Model and Office Thrift Supervision Model. Because the empirical analysis of this paper is related to LIBOR, we use LFM to simulate then LIBOR. CMOs involves a lot of risk like prepayment risk, so there is no close form of CMOs’ price, and we use Monte Carlo Method to pricing. The data for Neural Network is from Fannie Mae. There are some conclusions as following. First, the MSE of Neural Network is lower than MSE of the PSA or CPR model. Second, the prepayment rate of Neural Network is higher than the prepayment rate of PSA or CPR model, and the prepayment rate of OTS model is higher than the prepayment rate of PSA or CPR model. Finally, the price of Neural Network is closer to the real value than the price of the PSA or CPR model.參考文獻 中文文獻[1] 王琮生 (2003)，「房貸保險之費率結構分析-競爭風險模型之應用 」，朝陽科技大學財務金融系碩士班碩士論文。[2] 何澤蘭 (1999)，「台灣不動產抵押債券證券化之推行及評價」， 碩士論文，國立台灣大學財務金融研究所碩士論文。[3] 李俊民 (2006)，「不動產抵押貸款證券化之評價─以中國信託商業銀行特殊目的信託抵押貸款受益證券為例」，世新大學財務金融研究所碩士論文。[4] 林宗漢 (2003)，「應用存活分析於不動產抵押債權證券評價之研究」，朝陽科技大學財務金融系碩士班碩士論文。[5] 高心怡 (2000)，「結合Hull-White 利率模型與PHM 提前清償模型評價CMO 利率衍生性商品」，國立台灣大學財務金融研究所碩士論文。[6] 張繼文 (2010)，「擔保房貸憑證 (CMOs) 評價---以BGM利率模型為例」 政治大學金融研究所學位論文[7] 黃玉霜, 周清佳, & 林哲群 (2003)，「應用動態提前清償模型評價住宅抵押貸款證券」住宅學報, 中華民國住宅學會, 12(1), 43-56.[8] 黃世富 (2006)，「考慮違約損失下CMO商品的風險溢酬－應用One Factor Gaussian Copula 模型」，國立中正大學財務金融研究所碩士論文。英文文獻[1] Ambrose, B. W. and Michael, L. (2001), “Prepayment Risk in Adjustable Rate Mortgages Subject to Initial Year Discounts: Some New Evidence,” Real Estate Economics, Vol. 29, No. 2, pp.305-327.[2] Dunn, K. B. and McConnell, J. J. (1981), “A Compare of Alternative Models for Pricing GNMA Mortgage-Back Securities,” Journal of Finance Vol.36, pp.471-484.[3] Dunn, K. B. and McConnell, J. J. (1981), “Valuation of GNMA Mortgage-Backed Securities,” Journal of Finance, Vol.36, pp.599-616.[4] Deng, Y. H., Quigley, J. M. and Van Order, R. (1996), “Mortgage Default and Low Down Payment Loans: The Costs of Public Subsidy,” Regional Science and Urban Economics, Vol. 26, pp.263-285.[5] Deng, Y. H. (1997), “Mortgage Termination: An Empirical Hazard Model with Stochastic Term Structure,” Journal of Real Estate Finance and Economics, Vol.14, pp.309-331.[6] Green, J. and Shoven, J. (1986), “The Effects of Interest Rates on Mortgage Prepayments,” Journal of Money, Credit, and Banking, Vol.18, pp.41-59.[7] Kau, J. B., Keenan, D. C., Muller, W. J. Ⅲ and Epperson, J. F. (1993), “Option Theory and Floating-Rate Securities with a Comparison of Adjustable- and Fixed-Rate Mortgages,” Journal of Business, Vol. 66, No.4, pp.595-618.[8] Kau, J. B., Hilliard, J. E. and Slawson, V. C. (1998), “Valuing Prepayment and Default in a Fixed-Rate Mortgage: A Binomial Options Pricing Technology,” Real Estate Economics, Vol. 26, No.3, pp.431-468.[9] Riksen, R., Spreij, P. J. C. and den Iseger, P. W. (2017), “Using Artificial Neural Networks in the Calculation of Mortgage Prepayment Risk.”[10] Schwartz, E. S. and Torous, W. N. (1989), “Prepayment and the Valuation of Mortgage-Backed Securities,” Journal of Finance 44, 375 - 392.[11] Waller, B. and Aiken, M. (1998), “Predicting Prepayment of Residential Mortgages: a Neural Network Approach,” International Journal of Information and Management Sciences, 9, 37-44. 描述 碩士
國立政治大學
金融學系
105352032資料來源 http://thesis.lib.nccu.edu.tw/record/#G0105352032 資料類型 thesis dc.contributor.advisor 林士貴<br>陳亭甫 zh_TW dc.contributor.advisor Lin, Shih-Kuei<br>Chen, Ting-Fu en_US dc.contributor.author (Authors) 張憲明 zh_TW dc.contributor.author (Authors) Chang, Hsien-Ming en_US dc.creator (作者) 張憲明 zh_TW dc.creator (作者) Chang, Hsien-Ming en_US dc.date (日期) 2018 en_US dc.date.accessioned 3-Sep-2018 15:48:23 (UTC+8) - dc.date.available 3-Sep-2018 15:48:23 (UTC+8) - dc.date.issued (上傳時間) 3-Sep-2018 15:48:23 (UTC+8) - dc.identifier (Other Identifiers) G0105352032 en_US dc.identifier.uri (URI) http://nccur.lib.nccu.edu.tw/handle/140.119/119884 - dc.description (描述) 碩士 zh_TW dc.description (描述) 國立政治大學 zh_TW dc.description (描述) 金融學系 zh_TW dc.description (描述) 105352032 zh_TW dc.description.abstract (摘要) 本研究主要透過使用類神經網路的方法來預測擔保房貸憑證(CMOs)之提前還款風險並加以評價，並且與另外兩種模型進行比較，PSA/CPR模型和美國官方(Office Thrift Supervision; OTS)30年期固定利率住宅抵押貸款動態提前清償模型，其中PSA/CPR模型為靜態模型，OTS模型為動態模型，另外由於實證個案現金流與LIBOR利率有關，因此採用的利率模型為對數常態遠期利率(Lognormal Forward LIBOR Model, LFM)市場模型，且因為擔保房貸憑證(CMOs)涉及提前還款風險以及眾多風險，並無法找出封閉解，因此採用蒙地卡羅模擬法來作為評價模型。透過蒐集Fannie Mae公開的房屋抵押貸款資料來做實證，實證結果有以下貢獻，第一，類神經網路預測提前還款率均方誤差(MSE)小於PSA/CPR模型的均方誤差，所以類神經網路在預測提前還款率的方面優於PSA/CPR模型。第二類神經網路模型得出個案的發行機構低估發行當下所設定之提前還款率，而動態模型的OTS模型也得到相同結果，綜合以上兩點，得到類神經網路模型預測提前還款率的評價結果比PSA/CPR模型更接近真實價格。 zh_TW dc.description.abstract (摘要) This paper used the Neural Network to forecast the prepayment rate of Collateralized Mortgage Obligations (CMOs), and then pricing it. Comparing the Neural Network Model with other prepayment model: PSA/CPR (static) Model and Office Thrift Supervision Model. Because the empirical analysis of this paper is related to LIBOR, we use LFM to simulate then LIBOR. CMOs involves a lot of risk like prepayment risk, so there is no close form of CMOs’ price, and we use Monte Carlo Method to pricing. The data for Neural Network is from Fannie Mae. There are some conclusions as following. First, the MSE of Neural Network is lower than MSE of the PSA or CPR model. Second, the prepayment rate of Neural Network is higher than the prepayment rate of PSA or CPR model, and the prepayment rate of OTS model is higher than the prepayment rate of PSA or CPR model. Finally, the price of Neural Network is closer to the real value than the price of the PSA or CPR model. en_US dc.description.tableofcontents 表次....1圖次....2第一章 緒論......31.1 研究背景與動機.....31.2 研究目的..........5第二章 擔保房貸憑證(CMOs)介紹......62.1 轉手證券(Pass-Through).......62.2 轉支付證券(Pay-Through)......72.3 擔保房貸憑證(CMOs).........8第三章 文獻回顧............133.1 利率模型..........133.2 提前還款模型.......14第四章 模型介紹........164.1 利率模型..........164.1.1 LFM 模型........164.2 提前還款模型.......194.2.1 CPR 模型........194.2.2 PSA 模型........204.2.3 OTS 模型.........204.2.4 類神經網路模型....224.3 評價模型...........23第五章 評價及實證分析.....275.1 個案商品介紹..........275.1.1 Group 3 現金流......285.1.2 Group 1、2、4、5 現金流.....295.2 評價結果.......305.3 敏感性分析.....34第六章 結論及建議.....36參考文獻......37 zh_TW dc.format.extent 2410189 bytes - dc.format.mimetype application/pdf - dc.source.uri (資料來源) http://thesis.lib.nccu.edu.tw/record/#G0105352032 en_US dc.subject (關鍵詞) 房屋抵押貸款證券化 zh_TW dc.subject (關鍵詞) 擔保房貸憑證(CMOs) zh_TW dc.subject (關鍵詞) 提前還款 zh_TW dc.subject (關鍵詞) 類神經網路 zh_TW dc.subject (關鍵詞) 對數常態遠期利率(Lognormal Forward LIBOR Model, LFM)市場模型 zh_TW dc.subject (關鍵詞) Mortgage securities en_US dc.subject (關鍵詞) Collateralized mortgage obligations (CMOs) en_US dc.subject (關鍵詞) Prepayment en_US dc.subject (關鍵詞) Neural network en_US dc.subject (關鍵詞) Lognormal forward LIBOR model en_US dc.title (題名) 擔保房貸憑證(CMOs)之評價：應用類神經網路預測提前還款率 zh_TW dc.title (題名) Pricing Collateralized Mortgage Obligations: Using Neural Network to Forecast Prepayment Rate en_US dc.type (資料類型) thesis en_US dc.relation.reference (參考文獻) 中文文獻[1] 王琮生 (2003)，「房貸保險之費率結構分析-競爭風險模型之應用 」，朝陽科技大學財務金融系碩士班碩士論文。[2] 何澤蘭 (1999)，「台灣不動產抵押債券證券化之推行及評價」， 碩士論文，國立台灣大學財務金融研究所碩士論文。[3] 李俊民 (2006)，「不動產抵押貸款證券化之評價─以中國信託商業銀行特殊目的信託抵押貸款受益證券為例」，世新大學財務金融研究所碩士論文。[4] 林宗漢 (2003)，「應用存活分析於不動產抵押債權證券評價之研究」，朝陽科技大學財務金融系碩士班碩士論文。[5] 高心怡 (2000)，「結合Hull-White 利率模型與PHM 提前清償模型評價CMO 利率衍生性商品」，國立台灣大學財務金融研究所碩士論文。[6] 張繼文 (2010)，「擔保房貸憑證 (CMOs) 評價---以BGM利率模型為例」 政治大學金融研究所學位論文[7] 黃玉霜, 周清佳, & 林哲群 (2003)，「應用動態提前清償模型評價住宅抵押貸款證券」住宅學報, 中華民國住宅學會, 12(1), 43-56.[8] 黃世富 (2006)，「考慮違約損失下CMO商品的風險溢酬－應用One Factor Gaussian Copula 模型」，國立中正大學財務金融研究所碩士論文。英文文獻[1] Ambrose, B. W. and Michael, L. (2001), “Prepayment Risk in Adjustable Rate Mortgages Subject to Initial Year Discounts: Some New Evidence,” Real Estate Economics, Vol. 29, No. 2, pp.305-327.[2] Dunn, K. B. and McConnell, J. J. (1981), “A Compare of Alternative Models for Pricing GNMA Mortgage-Back Securities,” Journal of Finance Vol.36, pp.471-484.[3] Dunn, K. B. and McConnell, J. J. (1981), “Valuation of GNMA Mortgage-Backed Securities,” Journal of Finance, Vol.36, pp.599-616.[4] Deng, Y. H., Quigley, J. M. and Van Order, R. (1996), “Mortgage Default and Low Down Payment Loans: The Costs of Public Subsidy,” Regional Science and Urban Economics, Vol. 26, pp.263-285.[5] Deng, Y. H. (1997), “Mortgage Termination: An Empirical Hazard Model with Stochastic Term Structure,” Journal of Real Estate Finance and Economics, Vol.14, pp.309-331.[6] Green, J. and Shoven, J. (1986), “The Effects of Interest Rates on Mortgage Prepayments,” Journal of Money, Credit, and Banking, Vol.18, pp.41-59.[7] Kau, J. B., Keenan, D. C., Muller, W. J. Ⅲ and Epperson, J. F. (1993), “Option Theory and Floating-Rate Securities with a Comparison of Adjustable- and Fixed-Rate Mortgages,” Journal of Business, Vol. 66, No.4, pp.595-618.[8] Kau, J. B., Hilliard, J. E. and Slawson, V. C. (1998), “Valuing Prepayment and Default in a Fixed-Rate Mortgage: A Binomial Options Pricing Technology,” Real Estate Economics, Vol. 26, No.3, pp.431-468.[9] Riksen, R., Spreij, P. J. C. and den Iseger, P. W. (2017), “Using Artificial Neural Networks in the Calculation of Mortgage Prepayment Risk.”[10] Schwartz, E. S. and Torous, W. N. (1989), “Prepayment and the Valuation of Mortgage-Backed Securities,” Journal of Finance 44, 375 - 392.[11] Waller, B. and Aiken, M. (1998), “Predicting Prepayment of Residential Mortgages: a Neural Network Approach,” International Journal of Information and Management Sciences, 9, 37-44. zh_TW dc.identifier.doi (DOI) 10.6814/THE.NCCU.MB.029.2018.F06 -
