Credit Risk Measurement Of Chinese Commercial Bank Based On Industry Classification

In the modern society, finance is the great power of economy increase and banks are the most important make ups of financial system, which play a great role in the financial development and economy increase. According to empirical researches, finance is the core of the economic system and banks are playing great roles in the financial system. Bank is a high risk industry which facing different types of risks and among the risks, the most important is credit risk.Credit risk will damage the foundation of banks and the financial system in the long run if credit risk is out of control. It is the importance of credit risk management that Basel commission has issued many instructions on credit risk measurement, management and control measures to deal with credit risk and some banks are playing great roles in the development of models to measure credit risk and then decrease it. Nowadays credit risk which derives from default risks of firms is a serious problem for banks in China. Commercial banks and financial system in China are under reforms, and credit crisis management lags behind and credit system of China is not well established, so credit risk is a big problem for banks'performance and stability.So it is necessary and eager to predict the possibility of firm's default risk. Studies before do not pay much attention to the industry that firms belong to. But economic knowledge tells us that industry style is an important variant in financial distress prediction. This paper use entropy method to select financial ratios that play most important roles in the measurement of credit risk. Then different industry has different default risk is proved both theoretically and empirically. And for the first time, this paper develops a model to measure the credit risk with industry effects. Then the contribution of inputting industry variables is presented. We find that both the intercept and the slope coefficient are significant and input industry variables promote the model's forecasting ability.