Research On The Validity Of Merton Distance-to-default Model On Financial Distress Of Chinese Listed Companies

Since the pioneering research of Beaver and Altman, company’s financial distress has been paid widely and continuous attention. Many researchers devote themselves to develop the best model or approach for forecasting financial distress. Now, there are three dominant models:(1) traditional models predominantly based on accounting information,(2) structural models that view equity as a call option on assets, and(3) hazard models that consider time element of financial distress. The last model has been the dominant approach of current international research for viewing company’s financial distress as a dynamic process.There are many papers about the first model in domestic, but the research on the structural models is still in the validation and applicability of the model itself. Many practitioners and scholars do not know very much about its statistical properties. Recognizing this point, this thesis will analysis the validity of the Merton distance to default model on forecasting company’s financial distress based on the empirical evidence of Chinese companies. We ask the following three question:(1) how about the predictive ability of the model on financial distress? Whether the model is a good choice for Chinese listed companies based on its predictive ability,(2) where the power of the model’s forecasting ability comes from? Whether the functional form of the distance to default used by the model is too metaphysical and general for Chinese company,(3) whether the key indicator of the model, asset volatility, has capture some crucial information about company’s financial distress. The three question embody the three statistical properties of the model. First, whether the distance to default given by the Merton DD model is a sufficient statistic for forecasting financial distress. Second, whether the functional form used by the Merton DD model is an important construct for predicting financial distress. Three, whether the key indicator of the Merton DD model is an necessary index for forecasting financial distress.Because the domestic research takes on the following state: many static studies but less dynamic research, using too much sectional data but less time-series data and focusing on paired sample but less sample data, this thesis depends on the discrete time hazard model to analysis. Given that most scholars are only concerned about the discriminate power of the model, this paper focus on two aspects of model’s performance, one is the discriminate power, and another is model’s validation. The specific method contains relative information test, ROC curves, deciles forcasts and out-of-sample test. We define companies that are special treated by the CSRC are financial distress, and obtain data for the period 1998-2013 from the database of CSMAR. In all, we obtain a total of 334 ST firms covering the period 1980–2003, and there are 1372 non-ST firms. Our final combined sample of ST and non-ST firms consists of 16000 firm-year observations.By the empirical analysis and test, the main conclusions are as follows:First, the distance to default is an significant and stable but not sufficient indicator for financial distress prediction, it is not a better choice for Chinese companies. The information it contains is significant less than the classical Z-Score mode, the Area under ROC curve is not more than 0.76, the performance of out-of-sample predictive power is not satisfacting. Moreover, the model behaves obvious procyclicality.Second, the special functional form of DD is not vital relative to its component. DD is too metaphysical and general that lose information of company’s market value. The current paper doesn’t pay attention to this point. In fact, DD only measures two power of asset volatility, one is the direct hindering effect, and another is the indirect promoting role by means of the leverage. This is just the value of the model in Chinese stock market.Three, the asset volatility contains very important information of company’s financial distress, and has some significant incremental information for financial ratios. However, it doesn’t obviously improve the predictive ability of financial ratios model. Accounting information is so powerful that covers the force of market information on financial distress. For Chinese companies, ratios analysis is still an effective method for discriminating financial distress. The earnings before interest and tax to total assets and total liabilities to total assets are sufficient to explain company’s financial distress.