项目来源

台湾省政府科研基金(GRB)

项目主持人

杨晓文

项目受资助机构

台湾大学财务金融学系

财政年度

2018,2017,2016

立项时间

未公开

项目编号

MOST105-2410-H004-207-MY3

研究期限

未知 / 未知

项目级别

省级

受资助金额

2579.00千元台币

学科

财政(含金融，保险)

学科代码

未公开

基金类别

应用研究/学术补助

关键词

未公开

参与者

未公开

参与机构

未公开

项目标书摘要：高龄化社会和寿命改善已是全世界关注的议题，如何透过创新金融商品来增加退休财源也变得愈来愈重要。近年来反向抵押贷款商品在英国及美国的市场渐渐受到重视，了解其定价方法以及相关风险是开发反向抵押贷款商品的首要考量。房价、利率及死亡率不确定是其风险主要来源，而房价风险对於反向抵押贷款商品的影响最大，因此，本研究主要目的是提出可捕捉房价长期结构的模型，并进而考量跨区系统性风险架构下反向抵押贷款定价及风险评估之探讨。第一年的研究主题是探讨反向抵押贷款中房价模型的建构，由於反向抵押贷款商品可供老人住到身故，该商品的期间通常较传统金融商品来得长，在寿命改善的影响下，房价长期的变化趋势对该商品的定价和风险评估有显着的影响，因此第一年度的研究试图考量总体指标如GDP 或CPI 来建构房价模型，以实际资料来进行模型的配适度分析，并和文献中所提出的房价模型进行比较以及分析其对反向抵押贷款商品定价及风险评估之影响。第二年的研究主题是探讨跨区反向抵押贷款风险聚集(Risk Pooling)的效益，由於跨区房价间可能有相关性，并需考量其间的系统风险，因此本研究主要探讨跨区房价的相依性(dependence),并采用两种方法来捕捉其相依性，一是以copula functions,另一是以co-movement 的方式来建构。前者将考量动态及静态copula 的架构下并分析相依性的对称及不对称模式，而後者是考量房价及总体因子共同移动的模式，为检视相依性的模式及模型的配适度，本研究将进行实证分析。第三年的研究主题是建构考量系统风险下跨区反向抵押贷款商品的定价及风险评估模型，并以美国HECM program 为架构。现有文献在分析HECM loans 时皆忽略跨区的房价系统风险，本研究因此透过本研究所提出的长期总体经济因子跨区房价模型来推导反向抵押贷款商品之定价以及风险模型，其中定价模型是建构在风险中立的架构下，而风险评估是推导反向抵押贷款商品的风险值以及尾端期望值，数值分析将比较考量房价相依性、不同相依结构以及风险聚集对定价及风险衡量的影响。

Application Abstract: Aging populations and increases in longevity on a global scale have increased the need for retirement financing.Recently,the development of innovative financial products capable of increasing retirement income has received great attention.According the equity release markets in the U.S.and U.K.,reverse mortgages products have become more and more popular in funding the retirement cost.Understanding the pricing and risk evaluation for such products are important.This research project focuses on reverse mortgage products.The primary aim of this study is to examine the risk factors involved in the pricing and risk analysis for reverse mortgage products.Particularly,we consider the systematic risk across regions and develop a pricing and risk analysis framework for reverse mortgage products on a portfolio basis.In year one,we first deal with the house price risk.Reverse mortgage products often cover a long duration,so macroeconomic factors such as GDP or CPI could induce changes in house price return dynamics.To develop a house price return model for pricing long-duration financial products,we can’t ignore the structure change in the house prices.We propose the house price return dynamics incorporating the macroeconomic factors.We will investigate the economic factors that could affect the house price return modeling and compare the goodness fit of the existing house price return model relating to pricing reverse mortgage products.We provide a general valuation model to study the house price model on pricing and risk analysis of HECM program ignoring systematic risk in the first part.In year two,based on the proposed house price return model with economic factors,we then attempt to deal with systematic risk of house price return dynamics in pricing and risk analysis framework.We model and investigate the dependence structures across different housing regions.To model dependence structure,we consider two possible ways.The first method is to use copula functions and the second method is to model the co-movement structure of the house price series and economic factors.Accordingly,we consider both static and dynamic copula models with symmetric and asymmetric structure.Empirical study is conduct to examine the dependence structure.In year three,according to Markowitz diversification principles,if the underlying housing prices do not exhibit strong correlation,a diversified portfolio of reverse mortgages reduces their risk.We attempt to study the impact of dependence risk on HECM program by taking into account the dependence structure of housing returns across different regions.Existing literature has not addressed the dependence of housing prices across different regions to investigate risk for the HECM program yet.In response,we account for the dependence risk of housing price returns when pricing the no-recourse provision(NRP)of reverse mortgages and mortgage insurance to provide a further assessment of the risk in the HECM program.With a risk analysis,we measure risk by comparing benchmark models of an independent structure,including the value at risk(VaR)and conditional tail expectation(CTE)approaches,against the estimated copula models or the co-movement structure with respect to their quantification of the risk.

项目受资助省

台湾省