The Study On Loss-Given-Default Of Banking

The global financial crisis triggered by the subprime loans is motivating the interests of scholars, practitioners and regulators. The basic technology of credit risk research is under developed in China. Loss Given Defaul(tLGD), or Recovery Rate(RR) plays a pivotal role in credit risk(LGD+RR=1).The importance has been punctuated by Baselâ…¡and CBRC's continued emphasis on both expected and unexpected loss of credit assets. Most of the published researches pertain to recoveries on corporate bonds rather than bank loans because of loans'private character, few data is publicly available to researchers. Although there is an increasing studies on LGD(RR) in domestic, most of them are limited by small data sample which can't give out convincing results . Moreover, the analysis approaches in these researches are just simple descriptive statistics or discriminatory analysis.This paper reviews the theoretical and empirical literature concerning LGD(RR), one of the essential variables in IRB, looks at its drivers and estimation., introduces the LGD(RR) management requirement by Baselâ…¡and CBRC. There is an increasing need for LGD drivers evidence supports and estimation approaches in advanced and developing countries. Based on the empirical and classical financial theory with Sichuan wholesale non-perform loans data, this paper dedicates to answer: what is LGD(RR), why is important, what drives it, how to measure or estimate, how to construct database from the domestic perspective.With the help of empirical analyses and models, this paper describes and measures LGD(RR) from different dimensions, including the distribution, macro, meso and micro factors (term of loan, corporate asset-liability ratio,risk free rate, loan size, collateral,etc.) which influence LGD(RR),and the association with PD ( Probability of Default.Under the risk neutral framework of the structure credit risk model, this paper develops LGD analytic expressions, in which corporate value drift rate is replaced by risk free rate, the results are empirically tested on aggregate level. The cash flow is uncertainty when a corporate is faced with financial distress. Based on continuous-time asset pricing theory, this paper develops a RR model under the assumption of stochastic cash flows, where RR is determined endogenously by the corporate finance structure. Under our assumptions, recovery rate is positively correlated with the company's cash flow pressure, cash flow drift rate and corporate default interval relative length, negatively correlated with cash flow volatility. The results are also empirically tested. We utilized parameters estimate method to analyze the systematic risk in Recovery Rates. Our empirical results indicate that systematic risk exerts influences over Recovery Rates, and the influence sensitivity is stable. We also find there exists a statistically significant negative correlation between Default Rates and Recovery Rates on aggregate. Loan provision and Lorenz curve method are built or introduced to measure LGD(RR), At the last part, his paper suggests that CBRC,PBC and some banks'existing information systems could be utilized to construct LGD database as the data source in China. Practical suggestions are provided. Some theoretical results of the current paper are applied to design the variables of the LGD database.