Measuring The Credit Default Risk Based On Asymmetric Volatility KMV Model

Credit risk is a risk indicator generally concerned by investors,issuers,regulators,rating agencies and accounting standard setting agencies,especially in the context of the dramatic changes in the external economic environment and frequent defaults in the domestic bond market.The paper derived asymmetric volatility KMV model based on asymmetric volatility compound option pricing model,which can both separate short term and long term default risk and be used to analyze the effect of asymmetric volatility on default probability.The model overcomes the drawbacks of the traditional KMV,which treats long-term debt as short-term one and can be only used to estimate short-term default probability.It generalizes the symmetric volatility KMV model.As a useful supplement to the existing credit risk analysis model,it can deeply study the impact of volatility asymmetry on the company's credit risk.In view of asymmetric volatility KMV model,symmetric volatility KMV model and classical KMV model,empirical analysis estimates parameters using different methods and shows how the short term default risk and joint default risk of the A-share listed company ZHONGJIN GOLD stock evolve during 2007-2019.Specifically,iterative method is put forward and compared with simultaneous equations method and maximum likelihood method.The empirical results elucidate that iterative method is more stable and more efficient,which can be used as a benchmark.Three conclusions from empirical analysis are given as follows.Firstly,during the financial crisis of 2008-2010 and during 2015 A-share crash down period,the overall probability of default was high,and there was a long-term default risk in 2011,2013 and 2018.Secondly,there are significant differences in the analysis results of various models based on short-term volatility and long-term volatility.During turbulent times,the estimated default probabilities using short-term daily data are almost similar with the results using long-term monthly data qualitatively but much different quantitatively.On the contrary,during relatively stable times,some patterns are identified by short-term daily data but not by long-term monthly one.Thirdly,credit risk is mainly driven by persistent volatility.The theoretical model and estimation method in this paper are expected to be ben-eficial to the regulators,the credit agencies,the institutional investors,the individual investors and the issuers.