违约风险下的信贷决策模型与机制

研究了违约风险下的信贷决策模型与机制,通过以银行个体合理性和激励相容性作为约束条件,建立了在考虑违约风险和项目成功概率条件下的信贷决策模型,分别给出了基于抵质押贷款和信用贷款策略下的信贷决策机制,探讨了信贷配给机制与无配给机制的设计方法,给出了在信贷出现配给时银行发放信用贷款和有抵质押贷款的条件.最后运用实例详细分析并讨论了不同违约概率条件下企业项目成功概率对银行期望收益的影响,得到了银行相应的贷款临界值和在不同项目成功概率条件下银行最大可接受的违约概率.     

基于蒙特卡洛模拟的贷款组合优化决策方法

通常项目的收益与风险成正比,银行贷款不仅要考虑总体收益最大,同时尽量降低贷款的总体风险,而目前的贷款组合方法不能做到收益和风险的同步优化.通过蒙特卡洛模拟,以贷款项目的财务内部收益率及其波动反映其收益和风险,建立组合收益最大、组合风险最小的贷款组合多目标规划决策模型.分步求解得到满足约束条件的若干贷款有效组合和有效前沿曲线,有效前沿曲线与银行的无差异曲线的交点是满足银行风险偏好的最优贷款组合.该方法直接应用组合收益和组合风险进行贷款组合优化,提高了决策分析精度,银行可以根据自身的需要灵活方便地调整贷款组合中各贷款的权重,降低风险,提高收益,并且通过最优贷款组合达到收益和风险的同步最优化.     

基于风险偏好的银行贷款组合优化研究

通过蒙特卡洛模拟，以贷款项目的财务内部收益率及其波动反映收益和风险，建立了组合收益最大、组合风险最小的贷款组合多目标规划决策模型。求解模型得到贷款组合有效前沿曲线，有效前沿曲线与无差异曲线的交点为满足银行特定风险偏好的最优贷款组合。该方法直接应用组合收益和组合风险进行贷款组合优化，实现了收益和风险共同最优化，提高了决策分析精度。     

基于利率非平行移动风险控制的资产负债组合优化模型

引入现金流离散度对收益率曲线非平行移动带来的利率风险进行免疫,以银行各项资产和负债的现金流离散度零缺口为约束条件,以银行各项贷款组合收益最大为目标函数,建立了基于利率非平行移动风险控制的资产负债组合优化模型.该模型通过现金流离散度的零缺口免疫匹配银行的资产与负债,控制了利率期限结构非平行移动带来的利率风险.该利率期限结构非平行移动的风险免疫更具有普遍性.通过不同时段的远期收益率来贴现资产和负债的现金流,使现金流离散度的计算更加准确.反映了不同时期收益率变化对各期现金流的影响,提高了现金流离散度的计算精度,改变了现有研究的久期免疫用恒定的名义利率贴现现金流的做法.     

银行贷款风险定价的"翘板效应"

商业银行贷款的风险定价机制是利率市场化的核心,实现贷款利率的自主定价将是一个必然的发展趋势.目前中国各商业银行已广泛开展了客户信用等级评定工作,在此基础上如何进行贷款利率的风险定价并确定市场竞争的策略是亟待解决的问题.基于巴塞尔新资本协议内部评级法所使用的4个假设条件建立贷款的损失分布模型,并根据不同信用等级的债务人具有相同风险调整后的资本收益率得出银行贷款的风险定价具有"翘板效应",即经营成本低的银行在对信用等级高的债务人放贷时有价格优势,可以收取相对(经营成本高的银行)较低的利率,而对于信用等级低的债务人放贷则要收取相对较高的利率,这对中国商业银行进行贷款市场细分并确定合理的市场竞争策略具有重要的指导意义.     

不完全信息下基于双目标的博弈并购决策研究

建立复合实物期权模型,研究并购企业在竞争环境下满足收益最大化和并购成功概率最大化的并购决策问题。基于并购企业对竞争对手企业价值补偿值具有不完全信息的现实,运用实物期权博弈的方法,以企业期望价值收益最大化和成功并购概率最大化为目标,建立并购决策模型并求解最优并购时机,在此基础上进行并购决策分析,通过数值比较,得到并购企业的若干策略选择。研究结果表明,不完全信息下存在竞争的并购企业的期望收益小于完全垄断下的期望收益,即竞争因素削弱了并购的实物期权价值;并购企业在与对手企业进行并购博弈时,并购企业给目标企业的价值补偿越高,其最优并购时机越迟;不完全信息下若使并购企业期望收益最大的溢价比例不在溢价比例概率分布的区间内,则并购企业应取区间内的最大值,并购企业可以获得最大的期望收益和最大的成功并购概率;反之,使并购企业期望收益最大的溢价比例在概率分布内,企业取该溢价比例时可获得最大的期望收益,但企业需要在获得最大收益和增大成功并购概率之间权衡,当溢价比例取值越高时,并购时机越迟,所获收益也就越大。     

基于VaR收益率约束的贷款组合优化决策模型

本文以贷款的收益率为金融资产的收益,以贷款收益率的波动为标准反映贷款风险,在VaR约束下,以拉格朗日乘子法为工具求二次规划,建立了在既定组合收益范围内,组合风险最小的贷款组合优化决策模型。该模型的特点一是以收益率最大损失的形式、而不是收益额的形式来反应VaR,使组合决策分析更为方便。二是考虑了风险之间的相关性,用组合VaR的收益率最大损失来控制贷款收益率的风险限额;使贷款的分配直接反映了商业银行的风险承受能力。三是在合理的目标收益范围内,给定任意一个决策者期望的收益率,总能找到对应的风险最小的贷款组合,由此模型求出的有效边界,为组合贷款的优化决策提供了科学的方法。     

CERs收益权质押融资业务的质押率模型

质押率的确定是进行质押融资的核心工作之一,合理的质押率是银行控制质押风险的重要途径。本文综合考虑了CERs收益权质押融资所面临的价格波动风险、借款企业的减排水平、违约风险和国家宏观经济风险等因素,在银行保持风险水平一致的基础上构建了CERs收益权质押融资业务的质押率模型。研究表明:最优质押率与银行愿意承受最大损失下的最大质押率之间的大小关系不是由企业期末CERs的完成率、CERs的初始价格及银行愿意承受的最大贷款损失度所决定的,而是由企业的违约概率、银行的贷款利率、资金成本、贷款时间及银行能够承受最大损失发生的概率所决定。最后,以欧洲气候交易所的CERs数据为基础进行算例分析。     

信息可信与贷款利率确定问题的供应链鲁棒模型分析

提供真实可信的企业信息是企业申请银行贷款时的必要条件,但银行可以选择相信这些信息的部分或者全部,从而提供不同的利率水平。本文研究可信信息与银行贷款利率制定之间的关系以及相应决策对借款企业所在供应链的影响问题。通过构建银行参与的由供应商和零售商组成二级供应链模型刻画了零售商存在资金货款需求的供应链金融情景。针对银行采用零售商提供的全部信息或部分信息情况,从供应链管理视角考察了鲁棒的报童方法和极大极小方法分别获得了银行的最优利率决策及其对供应链运营所造成的影响。研究表明,银行采用部分信息并不能有效增加其收益,但其利率决策对供应链整体运作效率有非常显著的影响。     

不同决策目标下受资金约束零售商的最优采购策略研究

在提前订购和延时采购两种情景下,分别考虑最大化期望收益和最小化违约概率两种决策目标,对受资金约束零售商的最优采购量和相匹配的融资策略进行分析,结果表明,相比于以最大化期望收益为决策目标,以最小化违约概率为决策目标的零售商在提前订购中将奉行"保守"的采购量和融资策略（仅耗尽自有资金量采购而不融资）,而在延时采购中采取"激进"的采购量和融资策略（与最大化期望收益下一致）;在此基础上,对不同决策目标下受资金约束零售商的最优采购时机问题进行研究发现,在最大化期望收益下,零售商的最优采购时机由产品采购成本、市场容量均值、市场容量方差、自有资金量、银行贷款利率等多个参数共同决定,而在最小化违约概率下,零售商将始终选择延时采购。     

The lion's share: The impact of credit expectations and credit allocations on commitment to leaders

In this paper we examine how the relationship between leaders' credit allocation behavior and subordinates' commitment to their leader is influenced by the fulfillment of subordinates' expectations (i.e., expecting one type of behavior and having that behavior occur). We predicted that subordinates would display less commitment to their leader when their leader took credit for the subordinates' work. However, based on expectancy violations and psychological contracts research, we also predicted that expectation fulfillment would moderate this relationship. In two experimental studies we found that the negative effects of leader credit taking on commitment to the leader were mitigated when the subordinate expected the leader to take credit. However, when subordinates expected to receive credit and did not, the negative impact of leader credit taking was enhanced. We discuss the implications of these results for both theory and practice.     

代理成本与“信用价差之谜”

作为一个内生因素,企业中的代理问题可能会恶化企业的信用风险.通过将代理人与股东之间的最优合约模型嵌入到Leland-Toft内生违约框架中,研究了道德风险这一具体的代理问题对信用风险及信用价差的影响机制.结论表明:企业中的道德风险问题对企业信用风险产生明显影响,并且显著增加信用价差;对于描述道德风险的关键参数,论文建立的模型明确了它们在影响信用价差中起到的作用.由此,论文从代理问题的角度对“信用价差之谜”给出了合理解释.此外,文中也给出了债券价值、股权价值和内生违约边界的显性表达式.     

Conflict analysis of an export credit trade dispute

The improved metagame analysis algorithm is employed to study an international trade dispute involving government subsidised export credits. In particular, the methodology is applied to a conflict over the awarding of a large contract to supply subway cars by New York City. Logically feasible resolutions to the dispute are highlighted and the possible role of strategic deception in the awarding of the contract is investigated.     

筛选不同竞争类型零售商的贸易信用合同设计研究

考虑了信息不对称环境下资金约束供应链中贸易信用合同的设计问题。其中,零售商面临资金短缺困境,对自身的竞争类型具有信息优势;供应商愿意为资金短缺的零售商提供贸易信用。本文基于高低不同类型零售商的保留收益信息,将零售商的竞争环境分为三种类型,即高度竞争、适度竞争和低度竞争。据此,供应商应用筛选方法设计合同甄别不同竞争环境的零售商类型,提高自身和供应链的整体收益。研究还表明,在信息不对称的情况下,适度竞争环境中,贸易信用合同能实现供应链整体收益最大化。     

Banking credit worthiness: Evaluating the complex relationships

In developing economies agriculture and farming play crucial roles for economic sustainable development. Farmer credit risk evaluation is an important issue when determining financial support to farmers, improving agricultural supply chain performance, and ensuring profitability of financial institutions. Credit risk evaluation, or creditworthiness, is not a trivial exercise due to various complexities. Honoring complexity is necessary to effectively evaluate and predict farmer creditworthiness. A methodology using fuzzy rough-set theory and fuzzy C-means clustering is used to evaluate and investigate the complex relationships between farmer characteristics, competitive environmental factors, and farmer credit level. The methodology is detailed using actual bank data from 2044 farmers within China. This empirical methodology generates decision rules that provide insight to more complex relationships than can be found through standard econometric multivariate approaches. A rule-based methodological outcome can be used to predict the creditworthiness of farmers and to aid in agricultural loan decision making. Prediction accuracy of the rule-base was 81.16%. A central finding is that education and skills related characteristics are important for determining farmer credit-worthiness. Other implications are presented along with study limitations and future research directions.     

Pricing and momentum of syndicated credit in Europe

We investigate the determinants of syndicated loan prices for European borrowers, spanning the entire period of credit expansion and crisis. We construct a large data base consisting of 2102 rated syndicated loan deals from 1990 to 2008 in twenty three countries and ten economic sectors. We investigate the effects on the spread paid over the risk free rate of three major groups of factors: loan characteristics, borrower characteristics and characteristics of country of the borrower, using both qualitative and quantitative variables in a hedonic regression. The results show that all three groups are significant joint determinants of prices of syndicated credit. Our results confirm the existing literature but also provide for first time results stemming form the risk characteristics of the borrower and the country of the borrower. It is shown that distance-to-default as well as aggregate risk associated with country of the borrower are of great concern to the lenders and hence significantly affect the pricing of syndicated loans. Furthermore, we report that financial institutions and public utilities are able to negotiate for cheaper loans, but this result is reversed when financing is for acquisition purposes. Overall, risk, liquidity, solvency and sustainable performance by both the borrower and its domicile country are key determinants of syndicated loan prices.     

金融危机下带传染效应的违约预报

考虑多阶段状态变量的动态信息对违约风险的影响,同时考虑宏观因素和公司个体因素来构建违约预报模型,并且通过在状态变量中包含的行业因素来刻画行业间可能存在的信用传染效应;建立违约风险强度中参数的极大似然估计和渐近性质,进而建立条件违约概率期限结构的极大似然估计.利用极大似然估计及其渐近性质考虑传染效应的显著性检验问题,最后...     

基于多元随机波动模型的信用风险衍生定价

本文将随机波动期权定价封闭解模型扩展到多资产分析框架下.在考虑Cox-Rose模型多资产模拟以及随机波动矩阵的Wishart动态过程的基础上,将风险升水引入收益率方程,并将Merton模型下的公司违约风险扩展到随机分析框架下.最后利用数值模拟技术,对多资产随机分析模型的适用性及解的稳定性进行了模拟分析.其结果表明,多元随机波动模型对违约风险随机发生条件下的公司信用过程具有较之单元确定性模型更强的解释力.利用多资产模型,有助于金融机构更深入地把握企业信用体系中资产价值和负债的动态联动关系,对金融机构的信用风险管理具有十分重要的意义.     

Evaluation of corporate governance systems by credit rating agencies

In periods of high market volatility, and in order to minimize their risks, some investors prefer to invest their funds in well-governed companies. This paper aims to describe the methodology used by rating agencies to assess corporate governance systems (CGS) and to compare agencies’ practices. More and more the shareholders and creditors incorporate mechanisms related to CGS in the assessment of risks. Using a sample of five rating agencies, we distill 51 governance criteria to two governance factors using principal components analysis. The first factor represents the “shareholders rights and board of directors”. The second one is related to “remuneration policy and convergence of interests for shareholders and managers”. We identify three different business models for the corporate governance rating process.     

限贷和限购政策对一般均衡中房价的影响

住房价格的调控是当前中国社会经济中的热点问题,但目前国内外尚少见文献对其进行全面探讨． 该文试图涉及这个领域,按照住房限购条件是否有效建立了两个关于住宅市场的一般均衡模型． 研究发现,当“限购令”未生效时,在其他条件相同的情况下,只要复合商品部门为规模经济递减或不变时,严厉的房贷调控政策倾向于降低市场的均衡房价(分情况不同,还需要满足一些其他条件) ． 而当“限购令”发挥效力时,首付比例和限购数量都将影响市场中的均衡房价,且其关系将会是非常复杂的． 仅当首付比例不太大且复合商品的生产复合规模效应递增的前提下,当首付比例落入特定区间时,房地产调控政策就能有效地降低均衡房价．