Credit evaluation method

Credit evaluation method refers to the skill of analyzing and judging the credit status of the evaluated object, which runs through the whole process of analysis, synthesis and evaluation. According to different signs, credit evaluation methods have different classifications, such as qualitative analysis and quantitative analysis, subjective rating and objective rating, fuzzy mathematical rating and financial ratio analysis, factor analysis and comprehensive analysis, static rating and dynamic rating, prediction analysis and default rate model. The above classification is only a simple list, and various industries also have rating methods.

These methods cross each other, have their own characteristics and develop continuously. For example, in subjective rating methods and objective rating methods, subjective rating depends more on the qualitative analysis and comprehensive judgment of the rated institutions by rating personnel, while objective rating is more based on objective factors.

In the development of rating industry, rating companies constantly sum up their own experience, and rating indicators are also constantly refined, so it is necessary to compare different design methods. According to different methods, there are different understandings of the elements, mainly in the following ways.

5C Factor Analysis: This method mainly analyzes the following five credit factors: the borrower's personality, business ability, capital, asset mortgage and economic environment.

The 5P factor analysis includes personal factors, purpose factors, payment factors, creditor's rights protection factors and enterprise prospect factors.

5W Factor Analysis The 5W factor analysis method includes the borrower (who), the purpose of the loan (why), the repayment period (when), the collateral (what) and how to repay (how).

The factor analysis method of 4F method mainly focuses on the following four aspects: organizational factors, economic factors, financial factors and management factors.

Campari-Campari method analyzes the borrower's moral character from the following seven aspects: debt repayment record (personal character), the borrower's debt repayment ability (ability), the profit (deposit) obtained by the enterprise from the loan investment, the loan purpose (amount), repayment method (repayment) and loan insurance.

LAPP method LAPP method analyzes the following elements: liquidity, activity, profitability and potential.

CAMEL evaluation system Camel evaluation system includes five parts: capital adequacy ratio, asset quality, management level, profitability and liquidity. Its initials are "Camel", which is named after the same English name as "Camel".

The above rating methods are similar in content, and they are all based on qualitative analysis and quantitative calculation when necessary. What they have in common is the rating of moral quality, repayment ability, capital strength, guarantee and operating environment conditions, borrowers, loan purposes, repayment period, collateral, and how to repay, etc., but all aspects of corporate credit influencing factors must be included and cannot be omitted, otherwise the credit analysis will not meet the requirements of comprehensive reflection. The traditional credit evaluation factor analysis method is an expert analysis method used by financial institutions to analyze customer credit risk. In this index system, the focus is on qualitative indicators, and the credit level of customers is judged through frequent contact between customers. In addition, several credit evaluation companies in the United States believe that credit analysis is basically qualitative analysis, although some quantitative financial indicators are also attached importance, but the final conclusion depends on the subjective judgment of credit analysts and is finally decided by the rating Committee. This is the most widely used method in credit evaluation at present. The general practice is to give or set the standard weight of each specific index according to its different position in the overall rating target, and determine the standard value of each specific index at the same time, then compare the actual value of the index with the standard value to get the rating index score, and finally summarize the index score to get the overall weighted evaluation score.

The biggest advantage of weighted scoring method is that it is simple and easy to calculate, but it also has three obvious shortcomings.

First, there is no distinction between different attributes of indicators, which will lead to unscientific calculation of comprehensive index. There are often some indicators in credit evaluation that belong to the national indicators. For example, the asset-liability ratio is not as big as possible, nor as small as possible, but as close to the standard level as possible. For state indicators, the weighted scoring method is easy to get wrong results.

Second, it can't dynamically reflect the changes of enterprise development. Enterprise credit is continuous, and the weighted scoring method only investigates for one year, reflecting the state of the enterprise at that time, and it is difficult to judge the status and trend of credit risk.

Thirdly, the interval regulation of weight function is ignored. Strictly speaking, the complete weight range should be between the highest and lowest values of the index, not the average or the highest value. When using the weighted scoring method to calculate the comprehensive index, the actual value of the index value is compared with the standard value, and then multiplied by the weight. This ignores the scope of the weight, which will lead to the error of the evaluation results. It can be seen that the weighted scoring method is difficult to meet the basic requirements of credit evaluation. Based on the principle of fuzzy mathematics, this method uses membership function to make comprehensive evaluation. The general steps are as follows: firstly, the corresponding value of each index in the closed interval [0, 1] is given by using the membership function, which is called "single factor membership degree", and each index is evaluated separately. Then, the membership degree of each single factor is weighted arithmetic average, and the comprehensive membership degree is calculated to get the comprehensive evaluation index value. The closer the result is to 0, the worse it is, and the closer it is to 1, the better.

The membership function scoring method is more reasonable than the weighted scoring method, but the lack of effective treatment methods for state indicators will directly affect the accuracy of the scoring results. At the same time, this method fails to fully consider the dynamic changes of various indicators of enterprises in recent years, and the rating results are difficult to fully reflect the real situation of enterprise production and operation development. Therefore, the membership function evaluation method is still not suitable for scientific credit evaluation. According to the principle of multi-objective programming, the efficiency coefficient method determines the satisfactory value and impermissible value for each evaluation index respectively. Then take the impermissible value as the lower limit, calculate the degree that the index reaches the satisfactory value, and convert it into the corresponding evaluation score, and finally calculate the comprehensive index through weighting.

Because the satisfaction value and impermissible value of each index are generally taken from the optimal value and the worst value of the industry, the advantage of the efficiency coefficient method is that it can reflect the position of the enterprise in the same industry. The efficiency coefficient method also fails to treat different indicators differently, and does not fully reflect the economic development of the enterprise itself, resulting in unreasonable rating conclusions and unable to fully realize the rating purpose of credit evaluation. The analysis, attention, integration and judgment of credit status are an inseparable organic whole, which is also the rating process of the two-dimensional judgment analysis method of multiple credit risks.

Multivariate feature is a standard model derived from quantitative statistics with financial ratio as explanatory variable. Using this model to predict the possibility of an event of a certain nature, rating personnel can find the signal of credit crisis as soon as possible. After long-term practice, the application of this model is the most effective. Multivariate analysis is to select the variable that can provide more information from a plurality of variable values (financial ratio) that represent the characteristics of the observed object, and establish a discriminant function to minimize the misjudgment rate of the derived discriminant function in classifying the observed samples. Determine the critical value according to the discriminant score and locate the credit risk of the research object.

Two-dimensional judgment is to simultaneously examine the changes of credit risk from two aspects:

One is space, which correctly reflects the position of the evaluated object in the industry (or the whole industry);

Second, time, as far as possible, to investigate the possibility of credit risk of the assessed object within a period of time.