A Comparative Study Of Accounting Models And Market Models On The Ability Of Corporate Financial Distress Prediction

The number of listed companies in China is increasing year by year.The market competition is increasingly intensified.Once the company does not pay attention to guard against risks,strengthen business management,the company will be easier to make the financial conditions worse and in financial difficulty.The financially troubled company will cause serious losses to market participants.Therefore,we can through some financial distress prediction model to predict in advance enterprises' financial condition,providing certain warning role for investors,creditors,making them not suffer from a heavy loss.Since the 1930 s,foreign scholars began to study of enterprise bankruptcy.Based on foreign studies,Chinese scholars began to study financial distress forecast from 1987.Abundant research results have been formed at home and abroad on the prediction of financial distress.About financial distress at home and abroad,forecasting model mainly include the accounting model and market model.Accounting model mainly include Z-score model and Logistic model.Market model mainly include Merton model,as well as the KMV company of KMV model and Na?ve DD model proposed by Bharath & Shumway(2008),which model is better has not yet formed a unified conclusion.Based on 86 financial distress listed companies and 86 financial health listed companies of 2011-2016 as research samples,the article systematically compares the two kinds of model for prediction ability.The paper first chooses 12 indicators consisting of debt paying ability,operation ability,profit ability and development ability as primary index of Logistic model for prediction of default.Then,through a normal distribution test,significance test,single variable Logistic regression and step wise regression method,we constitute a Logistic model for prediction of default.Four indexes are: cash flow ratio,asset-liability ratio,total assets profit margins and growth rate of total assets.Then we compare the Logistic model and the Z-score model of Altman(1968)as the accounting model with Bharath & Shumway Na?ve DD model(2008)and the improved Na?ve DD model(1)as a market model.We use ROC curve analysis to compare the predictive power of default,finding the area under the ROC curve order from large to small is: Logistic financial distress prediction model,Z score model,improved Na?ve DD model and Na?ve DD model.The results show that the accounting model is superior to the market model,which make investors and creditors should more refer to the accounting information provided as a reference in making decisions.The empirical results also show that the prediction ability of Na?ve DD models in different default point is no significant difference,enlightening scholars should focus on the asset value and its volatility when they study in the market model,and should not focus on the improvement in the default point.