Along with the trend towards liberalization and globalization in financial markets, the capital markets in different countries are rapidly developing, enterprises are raising funds through a greater variety of channels, and international financial markets are moving increasingly closely together. Once credit is put at risk, the resulting domino effect, if relatively light, will harm enterprises, financial institutions, investors and creditors and, if more serious, will affect the national economic system, spreading to international financial markets.
The outbreak of the subprime crisis in the U.S. in July 2007 that was triggered by rising interest rates that caused many borrowers to default on subprime loans in the housing market is the most obvious example. With highly everaged financial products being packaged through securitization and distributed to every country, the financial crisis has been spreading to the whole world, resulting in a financial tsunami that has no precedent in history.
The literature on the assessment of the credit risk of enterprises may basically be regarded as having developed in two directions. The first relates to the accounting-based model that is mainly concerned with the firm’s historical financial statements, while the second has to do with the market-based model that is by and large based on information provided by the securities markets. In the early years, the evaluation of credit involved the use of the accounting-based model, for example Beaver’s (1966) univariate analysis approach, Altman’s (1968) multivariate discriminant analysis method, Ohlson’s (1980) Logit analysis method, Zmijewski’s (1984) Probit analysis method, and Atiya’s (2001) neutral network analysis. The structural model, which was based on information provided by the equity market, regarded equity and debt as the firm’s asset value options, with the equity option pricing model being used to assess the firm’s credit risk (Black and Scholes (1973) and Merton (1974)).
Based on the assumption that markets were efficient, any change in the quality of a firm’s credit would always be immediately and fully reflected in the market value of its equity and debt, If the hypotheses in models like the one above did not give consideration to market transaction costs, then the market value would reflect all available information, and for this reason the financial statements would not include information other than that related to the market prices of securities. In addition, the financial statements would only reflect past information and were non-forward looking, so that any financial reports that were disclosed were characterized by a time lag. Furthermore, through window dressing, the financial statements’ effectiveness in terms of forecasting credit risk was reduced.
It was for these reasons that many studies believed that the market-based model ought to be superior to the accounting-based model. However, while the addition of accounting variables into the market model served to increase the model’s ability to explain the underlying information (Benos and Papanastasopoulos (2005)), the results of these studies indicated that the market value of a security was inadequate in terms of encompassing all of the relevant information regarding the firm’s credit risk in its financial statements, and hence the accounting variables still served a purpose in assessing the firm’s credit risk.
