住房按揭贷款违约损失率预测模型研究

本文用log-log回归法、logistic回归法、分类及回归树法分别构建住房按揭贷款违约损失率预测模型,并将这三类模型及历史数据平均法模型对样本内、外数据的预测效果进行比较分析。研究结果表明:针对住房按揭贷款违约损失率,logistic回归方法预测效果最好,分类及回归树方法其次,而log-log回归方法预测误差最大。模型回归结果还表明,贷款价值比越高,违约损失率越大;贷款年龄越大,违约损失率越小;信用按揭贷款违约损失率显著高于抵押贷款;华东、华南和西南地区的违约损失率比华北地区要小;面积大的房屋贷款违约损失率更大。     

个贷违约率与宏观经济指标相关性研究

本文将个人住房抵押贷款违约率作为因变量,选择可能对其有影响的主要宏观经济指标作为自变量,建立自变量与因变量之间的多元线性回归模型,运用1994～2009年第一季度间美国有关季度数据分阶段对该假设模型进行实证比较分析和检验,得出在不同宏观经济环境下各指标变量与个人住房抵押贷款违约率的影响关系,并尝试对与常理相矛盾的实证结果进行分析与解释。本研究对个人住房抵押贷款违约风险的管理及相关研究判断具有实证价值和现实意义。     

生存分析、宏观经济变量与违约概率

在研究个人贷款违约风险中,传统的研究往往只单纯地将宏观经济指标作为协变量,并未考虑宏观经济指标的时变交互特征(vandell,1993;Zandi,1998)。国外研究达成共识的是以Logistic回归的传统模型不能给出违约概率的动态预测值,且反映经济形势的宏观经济变量也不能纳入模型中。论文论述了生存分析与Logistic模型的理论机理,试图在借贷违约风险中加入系统性风险对违约的影响因素,克服了以Logistic回归模型为代表的传统模型在度量信贷违约概率时仅考虑个体非系统性风险的局限。研究结论说明宏观经济变量确实对违约风险有影响,对网贷违约风险来说,Cox模型更优于Logistic函数。     

我国农林牧渔业短期贷款企业违约判别实证分析

近几年,我国农林牧渔业短期贷款企业的违约严重程度一直居所有行业之首,从跨行业的角度评估该行业短期贷款企业的违约具有重要意义。本文基于全国跨银行的贷款企业海量数据库样本,针对农林牧渔业的短期贷款企业进行了分规模和分地区样本的多元判别分析模型、Logistic模型与神经网络模型等的构建与实证探索,进而找出了影响我国农林牧渔业企业违约的关键变量,构建了最佳违约判别模型。这些关键变量和判别模型对中国人民银行和各商业银行监测该行业企业的信用风险具有重要的参考价值。     

中小企业违约和破产研究文献综述

有效识别中小企业贷款的风险要素,准确预测和度量中小企业贷款的违约风险,是破解中小企业贷款难的关键。对于中小企业违约和破产的实证研究采用的方法有普通的多元回归（OLS）、简单一直观模型、传统的多元判别分析模型和逻辑回归模型等。影响中小企业信贷风险的要素除了传统的财务变量以外,还包括非财务因素,而且非财务的定性要素比定量的财务因素的作用更加重要。     

我国农林牧渔业短期贷款企业违约判别实证分析

近几年，我国农林牧渔业短期贷款企业的违约严重程度一直居所有行业之首，从跨行业的角度评估该行业短期贷款企业的违约具有重要意义。本文基于全国跨银行的贷款企业海量数据库样本，针对农林牧渔业的短期贷款企业进行了分规模和分地区样本的多元判别分析模型、Logistic模型与神经网络模型等的构建与实证探索，进而找出了影响我国农林牧渔业企业违约的关键变量，构建了最佳违约判别模型。这些关键变量和判别模型对中国人民银行和各商业银行监测该行业企业的信用风险具有重要的参考价值。     

浅析违约概率的应用

新巴塞尔资本协议的出台鼓励银行采用内部评级法评估所面临的信用风险,采用内部评级法,可以使银行资本和所承担风险的联系更加紧密、更加直接,可以大幅度提高资本对风险的敏感性。内部评级法计算资本充足率需要四个变量,违约概率是其中之一,同时只要采用内部评级法,不管应用高级法还是初级法,违约概率都需要由银行自己估算,国内某些商业银行已开始着手违约概率的计量。本文主要对违约概率的定义、作用及在现实中的应用作一些简要的介绍。     

基于Markov区制转换VAR模型的商业银行小微贷款压力测试研究

压力测试作为一般风险计量工具的重要补充,越来越受到金融监管部门和银行业的重视。通过引入MS-VAR模型,考察了宏观经济在不同区制下对小微贷款违约率的动态影响。基于MSIH(2)-VAR(1)模型、敏感性分析和情景分析的结果表明;在压力情景下,宏观经济变量对小微贷款违约率有冲击效应,但冲击效应并不显著;经济上升阶段银行信贷行为的顺周期性导致贷款违约率的增加。     

固定利率住房抵押贷款违约行为及其定价研究

固定利率住房抵押贷款的信用风险主要是违约风险,基于理性期权的定价模型往往会低估借款人的违约概率.通过分析违约成本及非理性违约因素,可以确定借款人违约时贷款机构收回的现金流,得到固定利率住房抵押贷款定价的期望值模型,并得出模型的求解方法.     

有序多分类logistic模型在违约概率测算中的应用

初始违约概率的测算是商业银行实施经济资本管理的必要环节。针对我国商业银行的现状,结合贷款五级分类,通过对银行的公司类客户的财务指标作时间加权化处理、因子分析、ROC检验以及使用有序多分类logistic模型对初始违约概率的测算作了有价值的探索,并通过算例分析论证了其可行性。     

Mortgage Default with Asymmetric Information

This article analyzes mortgage-market equilibrium when borrower default costs are private information. By applying the approach of Rothschild and Stiglitz (1976), it is shown that asymmetric information regarding default costs distorts the contract choices available in the mortgage market, preventing safe borrowers (those with high default costs) from fully satisfying their demand for mortgage debt. Large loans are available for a substantial interest-rate premium, but only risky borrowers find this premium worth paying. The article builds on an empirical literature designed to test the ruthless-default principle from option-based models of mortgage pricing. That literature provides evidence against ruthless behavior, suggesting that default costs play an important role in borrower decisions. The article takes a further step by arguing that such costs are private information, which has important implications for market equilibrium.     

The Default Premium and Corporate Bond Experience

The development of organized markets for speculative-grade corporate debt has provided financial researchers with an opportunity to examine the pricing of default risk. By incorporating previous work on the default experience of low-rated corporate debt, this paper presents an introduction to risk-neutral models of risky-bond pricing and uses these to examine the relationship between the default premium embodied in bond yields and actual default rates. The contribution of macroeconomic information to the default premium is also examined. The author finds that holders of low-grade bonds have, on average, been compensated for losses due to default.     

Default risk and equity returns: Australian evidence

We test whether default risk is related to equity returns using the Fama and MacBeth [Fama, E.F., MacBeth, J., 1973. Risk, return, and equilibrium: empirical tests. Journal of Political Economy 81, 607–636.] regression framework. The proxy we use for default risk is the default probability obtained from option-based models. Our findings show that default probability is negatively related to returns. While we find that size and book-to-market are related to default risk, the ability of these variables to explain cross-sectional variation in returns is not because they are proxying default risk. Further, our evidence suggests that the negative relationship between default probability and returns is not due to a leverage, volatility or momentum effect.     

Default Risk in Futures Markets: The Customer-Broker Relationship

The traditional view of the futures clearinghouse as an insurer that eliminates the need for customers to evaluate default risk is inaccurate. A clearinghouse member default in 1985 confirms that the clearinghouse only guarantees payment from member to member, not from customer to customer or member to customer. Thus, non-defaulting customers are subject to losses as a result of the action of individuals with whom thay have no contractual obligations. This study models the behavior of customers choosing a futures commission merchant (FCM) given the current legal position of the clearinghouse. In a single-period model with symmetric information, customers can eliminate their exposure to defaults of other customers or of their FCM only by choosing to trade through “boutique” (undiversified) FCMs. In practice, monitoring and rebalancing costs may impede the attainment of zero default risk. However, FCM diversification remains an important factor in customer choice of an FCM. When setting capital requirements, clearinghouses and government regulators need to consider the implications of diversification for both customer and market protection.     

A Proportional Hazards Model of Commercial Mortgage Default with Originator Bias

A proportional hazards model with competing risks is specified and is extended to correct for the possibility of originator bias. The model is used to examine the ability of option-theoretic models of mortgage pricing to forecast commercial mortgage defaults. Among the findings, those especially of interest include the influence of contemporaneous loan-to-value and debt-service-coverage ratios on commercial mortgage default probabilities. The paper also finds that option-theoretic models of mortgage pricing are quite capable of producing default estimates that fit the actual default rates well, especially when the model is corrected for originator bias.     

Default prediction with dynamic sectoral and macroeconomic frailties

This paper extends the macroeconomic frailty model to include sectoral frailty factors that capture default correlations among firms in a similar business. We estimate sectoral and macroeconomic frailty factors and their effects on default intensity using the data for Japanese firms from 1992 to 2010. We find strong evidence for the presence of sectoral frailty factors even after accounting for the effects of observable covariates and macroeconomic frailty on default intensity. The model with sectoral frailties performs better than that without. Results show that accounting for the sources of unobserved sectoral default risk covariations improves the accuracy of default probability estimation.     

信用违约风险传染模型的比较研究

本文试图对几种有代表性的模型进行比较,来分析由于建模方式的不同,而导致的对信用期权定价和对冲的结果的不同.如果将违约风险传染考虑进去,类似德隆帝国崩溃的事件,或许就能避免.     

Bond Yields,Taxes, and the Dimensions of Default Risk

This work develops and empirically estimates models of bond yields subject to default risk. Parameters for the probability of survival and the recovery rate subsequent to default are included in a model of corporate bond yields that allow a study of interactions among them. The municipal version of the model includes the tax rate as an additional factor. Empirical evidence is found that supports the notion that both dimensions of default, considered jointly, are related to bond quality. In addition, statistically significant differences in tax rates suggest that higher tax rates are associated with higher grade municipal bonds.     

Pricing the Risks of Default: A Note on Madan and Unal

In their well-known article, Madan and Unal (1998) presented one of the first intensity-based credit risk models. In this approach the default intensity is directly linked to the market value of the firm's equity. In order to derive the probability of default Madan and Unal have to solve a partial differential equation (PDE). Here, we show that one of the transformations in the derivation of the solution of this PDE is not correct and analyze the difference between the correct solution of the PDE and the solution based on the incorrect transformation. As a consequence of the transformation error the credit risk of a debtor is systematically underestimated.