Credit rating models for dummies, Part 1 : Altman Z-Score model
In one of recent our posts, we had talked about why we should not be allergic to Credit Ratings and right after, we had stopped being allergic to these useful tools. Now, why not talk about the Credit Rating Models with details? It seems quite difficult to explain all models here in only one post, but I try to give a framework with basic characteristics of frequently used models in the credit risk world. Even though you know it well, or maybe have read millions of times, we can remember the definition of credit ratings first:
Let’s define them with their features:
- They are opinions about credit risks;
- They are assessments of an entity’s ability to pay its financial obligations (SEC definition);
- They are predictive and have advanced views since they state the likelihood that the evaluated part will go bankrupt;
- Usually the parties which assess the ratings (e.g. rating agencies, financial institutions, etc.) utilize their own methodology to measure the creditworthiness and use a rating scale to declare its rating evaluation and mostly it is expressed in letters such as “AAA to D”.
In spite of recent criticism, credit ratings still remain the most common and widely used measure of corporate credit quality. There are plenty of credit risk models utilized by companies, institutions and banks. However we don’t aim to run a comparison of all of models but only the well-known and readily available credit rating models. Hence, here I have chosen only the mostly utilized ones just to have a general information about their basics :
- The Models Based on Financial Statement Analysis (e.g. Altman Z score, Moody’s Risk Calc),
- The Models Measuring Default probability (such as Structural Models)
- Machine Learning Models (e.g. Maximum expected Utility).
Today, we start with Altman model!
Financial Statement Analysis Models: These models provide a rating based on the analysis of financial statement items and ratios. The right example for this type of models to talk about can be well-known Altman’s Z-Score.
Altman Z- Score:
Edward Altman developed a model using financial statement ratios and multiple discriminant analyses to predict bankruptcy for publicly traded manufacturing corporate. The advantage of discriminant analysis is to be able to use many characteristics that can be combined into a single score. The discriminant function transforms the individual variable values into a single discriminant score or Z-value which is then used to classify the analyzed company. This model uses five financial ratios which are derived from the financial statements as reported by both bankrupt and non-bankrupt companies. The ratios are then combined in a specific way to produce a single number so called z-score that is a general measure of corporate financial health. (Source).
In the American business environment, Prof. Altman used those five indicators that have enabled the prediction of 72% of the companies’ bankruptcies with two years prior their occurrence. He initially used a sample of 66 companies of which 33 were distressed and 33 were financially healthy (distressed group of companies comprise the ones filed a bankruptcy from 1946 to 1965). After the initial groups were defined and companies were selected, balance sheet and income statement data were collected. A list of twenty two potentially important indicators of bankruptcy signs were collected for evaluation process. Then, the variables were classified into five categories: liquidity, profitability, leverage, solvency and activity. The ratios were chosen on the basis of their popularity in the literature and their applicability to study. Out of 22 variables, five were chosen which are best in prediction of bankruptcy. The resultant model is presented as follows:
Where “Z” indicates an index of bankruptcy. The discriminant coefficients represent the share of economic and financial indicators in assessing the bankruptcy risk, the level of an indicator being the best as the highest absolute values. The overall value of z-score demonstrates as follows :
Zones of Discrimination:
Z < 1.81 = Zone I – Distress Zone – High probability of bankruptcy for the company;
1.81 < z < 2.99 = Grey area – uncertain zone;
z > 2.99 = Zone II – Safe zone – Low probability of bankruptcy for the company.
The concern about original Z-score model was that the model was solely applicable to publicly traded companies (since needs stock price data), so that Altman tried to find solutions to apply model to the companies in private sector. Hence, the z-score model was revised, substituting the book value of equity with the market value. The result is exhibited in the following z’-score:
The only variable changed observed was with X4 which became:
X4 = Book value of equity / Book value of total debt.
X3 and X5 were virtually changed. The overall value of Z’-score is as follows:
Z’ < 1.23 = Zone I – Distress Zone – High probability of bankruptcy for the company;
1.23 < Z’< 2.90 = Grey area – uncertain zone;
Z’ > 2.90 = Zone II – Safe zone – Low probability of bankruptcy for the company.
Altman realized a further revision with the model was necessary in order to adapt the model for non-manufacturer companies and emerging markets. The modification aimed to analyze the model’s accuracy without X5 = sales/total assets . This was carried out by having the minimum potential industry impact which is more likely to take place when an industry-sensitive variable like asset turnover is included. Altman, Hatzell and Peck (1995,1997) applied this revised Z’’ Score model to emerging markets’ companies (Source: “Revisiting Credit Scoring Models in a Basel 2 Environment”, Edward I. Altman, May 2002). This time, the book value of equity was used for X4. The resulting model new Z’’ score model is as follows:
Where Z’’ Scores below 1.10 indicate a distressed condition.
The advantages and disadvantages of Altman Z score will be discussed at the end of this blog series.
Next week we will have a look at Structural Models.
For now, please stay tuned…