Forcasting Default Applications: Based On Incomplete Information Model

Credit risk is the oldest risk which influences most in the finance market. It can be considered as the risk that the counterparty have no capacity or willing to perform the contract. As the case in our country, credit risk plays more important part, which is determined by the condition of China. Partly of the short history and unsound function of our finance market, the phenomenon of"information asymmetry"tends to be more serious. Moreover, market segment is a big problem in current finance market, which means that bank loan is often used by enterprise to fund than security market. Credit risk concentrates in banking system, so we call credit risk a"more important risk"in China.Chapter I is Literature Review, we elaborate on the state of research both aboard and domestic. Chapter two is about modern model of credit risk measurement. Structural model, based on option pricing theory, combines default process with micro factors like capital structure, operation risk, and macro factors like interest rate. We value risk by the changes of these factors(Altman & Anthony,1998,Guangming Hou & Ran Zhang,2005). Structural model aims at default rate. It's an model of first passage which means that default occours only when the firm value falls below the default barrier. Merton creats the first structural model: Merdon Model, which is very influential. Afterwards, structual model plays an important part in finance marker and risk pricing. We present Merdon Model and Black-Cox Model in the following context.The hypothesis of Merdon Model consists that firm value follows Geometric Brownian motion, and there's a company takes a one-year risky bond and equity as capital structure. When the firm value falls below the debt value at expire date, default occours. So we can treat debt as a put on firm value. Suppose the market is effective and the claimer obey the bankruptcy claim order. Default probability can be conducted by Black-Scholes option pricing formula. Merdon Model is the first structural model which creates the view of option. It measures the default risk of listed enterprise by data from equity market. It's also a good financial warning system. But the limitation is also apparent, which comes from the hypothesis such as single capital structure,etc. The hypothesis behind is that default is predictable, which means that the model forecasts zero short-term credit spreads. First passage model is another type of structural model which holds on the same default defination. But the default not only occur on the expire date, as long as the firm value falls below the debt value. So this is what the"First passage"mean. We describe the Black-Cox model which conduct the default probability by mathematical way. The revised form is also included.Structural model holds on the economic appeal. It's the first time to lead option pricing into credit risk measurement, but it underestimates the credit spread. In structural model, default is predictable, but it's not hold in reality for information asymmetry.Reduced-form model consideres default unpredictable. In this model, it decomposes the default-influencing factor, and introduces intensity to express the default process, here the intensity follows Poisson distribution. Reduced-form model emphasizes the use of math, and is calibrated to market data. But it lacks economic support, and the default maybe not external.In the second chapter, we continue to introduce incomplete information model. It combines the economic reasonability and empiricial plausibility. Suppose default occurs when firm value falls to a barrier, and a pricing trend is introduced to express default process. Default is unpredictable, investors make decisions by incomplete information. Here we have 3 different types of incomplete information, one is incomplete information of default barrier, the other is about firm value,and the third one is both are unknowm. We discuss the first condition.In the actual process , based on the Kay Giesecke (2004) Research results ,company default boundary is considered to be follow scaled scaled beta distribution , varies as leverage and short-term liabilities changes. When the company value falls to default boundary, the company defaults .Then ,we compares Morton model,first passage model and incomplete information model. Trying to explain their differences in default definition, model sensitivity and default point.The fourth chapter is empirical analysis. This paper uses the data of the capital market, and the validity and applicability of the model is verified by the sample: ST SamSung(000068) and other 26 companies that were"ST"or"*ST"in 2008,Through calculation, incomplete information model generate the higher default probability than Merton Model does,holds on stronger explanatory result to reality.