TW201327453A - Small enterprise financing risk assessment method - Google Patents

Abstract

A small enterprise financing risk assessment method implemented with a computer system includes the following step: reading historical score data from a computer system, and classifying risk groups of the historical score data and calculating default probability of each risk group with the computer system. Then, a new customer risk rating group is calculated by imputing a score sheet of the new customers, and a financing contract of the new customer is determined according to the risk group thereof.

Description

Micro-enterprise financing risk assessment method

The invention relates to a micro-enterprise financing risk assessment method, in particular to an application computer system for micro-enterprise financing risk assessment method.

With the advancement of society, the demand for capital turnover by individuals and businesses has increased with time. The emergence of credit financing shows that people attach importance to credit, which is undoubtedly of epoch-making significance for raising people's attention to credit. For an advanced society, the binding force of credit must be greatly improved to enhance the credit of individuals and enterprises. It is often said that "people do not believe in standing."

In general, banks and society may be concerned about the risks of credit financing. In fact, this is a two-pronged one. How to effectively manage the credit of individuals and enterprises, and reduce the risk of financial institutions when lending, not only can effectively increase the lending profits of financial institutions, but also reduce the interest of individuals or corporate lending.

Since enterprises in emerging countries mostly operate in a family mode, the financial statements of enterprises often disguise through different methods to beautify the data on the statements, thus greatly reducing the traditional financial crisis warning model constructed by considering only financial variables. Predictive effectiveness. In particular, for financial statements that are not on the shelves or listed companies, and because of the lack of effective management, they are often out of touch with reality, so that the credibility is too much.

However, in the process of financing risk assessment, it is often found that individual financing risk assessment data and corporate financing risk assessment data often have completely different behavior patterns. What is more noteworthy is that the results of large-scale corporate financing risk assessments are often not applicable to micro-enterprises, and often have very different results.

Therefore, it is necessary to establish a reasonable assessment method for the financing risks of micro-enterprises to effectively manage the financing risks of micro-enterprises.

In view of the above-mentioned prior art, since financial institutions need to be able to correctly judge the repayment ability of enterprises in order to reduce risks, how to judge the repayment ability of micro-enterprise materials and design Appropriate repayment methods will effectively reduce financing risks.

One of the objectives of the present invention is to provide a micro-enterprise financing risk assessment method for use in a computer system to effectively assess a new micro-enterprise customer and to fully establish a financing risk assessment method suitable for micro-enterprises. .

According to the above-mentioned purposes, the present invention discloses a micro-enterprise financing risk assessment method, which is used in a computer system to effectively perform a default risk assessment on a new micro-enterprise customer and determine the amount and duration of financing.

This micro-enterprise financing risk assessment method is used in a computer system to conduct a default risk assessment for a new micro-enterprise customer, which includes the following steps: reading the historical score data of the micro-enterprise of the database of the computer system; The computer system calculates a risk grading group; uses the computer system to calculate a default probability of each risk grading group; inputs a new customer rating table; uses the computer to calculate a risk grading group of the new customer; and the computer is based on The risk rating group determines the financing contract for new customers.

Among them, the above-mentioned risk classification group for calculating new customers uses a computer system to calculate significant variables of historical score data by stepwise regression analysis, regression analysis, stepwise logistic regression analysis or logistic regression analysis, and the new customer score table. Enter the computer to calculate the risk rating group for the new customer. Significant variables include category variables and continuous variables. For example, category variables include buying and selling retail, personal insurance is normal, for example, property insurance contributions are normal or other life insurance contributions are normal, and no credit card pre-borrowing cash, while continuous variables include Pre-tax net profit/total assets, current turnover/total assets, turnover growth rate, interest expenses/current turnover, and operating margin/current turnover.

The micro-enterprise financing risk assessment method of the present invention further includes using a computer to calculate the survival period and the solvency. The risk rating group is confirmed by P1, P2 and P3, which can further accelerate the loan procedures. When the risk rating group is P4 and P5, the financing contract is determined according to the survival period and the solvency. When the risk rating group is P4 and P5 and the survival period is more than six months, the customer is at risk of default and further limits the loan ceiling according to the solvency. When the risk rating group is P4 and P5 and the survival period is less than six months, the customer is highly risky and refuses to undertake the loan when it is unable to provide the guarantee of solvency.

The above-mentioned micro-enterprise customers are preferably a small and medium-sized enterprise with a capital of no more than 10 million Taiwan dollars and a loan amount of not more than 4 million Taiwan dollars.

Therefore, the micro-enterprise financing risk assessment method of the present invention can not only effectively judge the default risk of new customers, but also update the historical score data by accumulating data accumulated over time to more accurately estimate the risk of customer default. In addition, since the present invention performs customer risk assessment based on objective significant variables, the artificial subjective judgment is reduced, which facilitates the realization of the automated evaluation of the micro-enterprise loan, and reduces the disadvantages and risks caused by the artificial nuclear loan.

The invention discloses a micro-enterprise financing risk assessment method, which can effectively evaluate the risk of micro-enterprise financing, and further evaluate the solvency and the estimated time-to-date of the default for the micro-enterprises with high risk classification, according to the estimated default. The time limit and the amount of the financing contract, effectively reduce the risk of financing financial institutions. The spirit and scope of the present invention will be apparent from the following description of the preferred embodiments of the invention. The spirit and scope of the present invention.

Referring to Figure 1, there is shown a schematic diagram of a micro-enterprise financing risk assessment method of the present invention. The micro-enterprise financing risk assessment method of the present invention first reads the historical score data input in the computer database, includes known historical score records and actual repayment records, and can be continuously updated according to new data. A database to build a more realistic assessment method.

As shown in the figure, in step 110, the historical score data in the computer system is first read. In step 120, the risk classification groups P1 to P5 are calculated. Next, in step 130, the probability of default for each risk rating group is calculated.

The micro-enterprise financing risk assessment method of the present invention is mainly used for judging the accuracy of the loan made by the small and medium-sized enterprises whose capital amount is less than NT$10 million and the loan amount is less than NT$4 million. .

For example, when the loan information of the micro-enterprise that meets the above conditions and the related scores and other related materials are read from the database of the computer system, the classification of the risk classification group is performed, and the micro-enterprise that meets the above conditions will be The loan data and the score sheet are grouped using an analysis of a logit model.

For example, the risk rating group is divided into P1 to P5, where P1 represents the safest risk rating in the risk rating and P5 represents the most dangerous risk rating in the risk rating. In addition, the predicted probability value 0 represents a default case, and the predicted probability value 1 represents a normal case, and the predicted probability value of P1 is set to 1.0 to 0.9, and the predicted probability value of P2 is set to 0.9 to 0.8, and the predicted probability value of P3 is set. For 0.8 to 0.7, the predicted probability value of P4 is set to 0.7 to 0.6, and the predicted probability value of P5 is set to be less than 0.6. It is based on historical score data including historical score records and actual repayment records, and uses statistical methods such as stepwise regression analysis, regression analysis, stepwise logistic regression analysis or logistic regression analysis to calculate significant variables of each group. To obtain the appropriate risk grouping and establish a risk assessment function. Furthermore, the significant variables of the present invention preferably include class variables and continuous variables.

Next, in step 130, the actual probability of default (PD) of each risk classification group P1 to P5 is further calculated. In actual data and simulation results, the micro enterprise can be found, and the revenue is less than NT$25 million. Yuan and the actual probability of default of NT$25 million or more have different results. See Table 1 and Table 2.

In order to make it more accurate to predict the chances of new customers guaranteeing repayment and default, it is better to increase the correctness by evaluating the cases with revenue less than NT$25 million and NT$25 million or more.

For example, revenue is less than NT$25 million, which has a lower risk of default. In general, cases with revenues greater than or equal to NT$25 million have a higher risk of default. However, the micro-enterprise financing risk assessment method of the present invention can use different databases to calculate different significant variables to apply the appropriate loan amount.

Through the historical scoring data and the customer repayment situation, after the simulation of the evaluation model, the new customer can use the micro-enterprise financing risk assessment method of the present invention to accurately judge the risk classification group of the new customer, and can further understand the situation. The probability of default is effective in increasing the discriminating ability of new customers, and the risk of microfinance loans is reduced by the financial industry.

Step 140, when a new customer wants to make a loan, firstly, a score sheet must be filled in, and the score sheet is input into the computer. The score sheet mainly uses objective scoring items and simulates corresponding coefficients for customer risk. Judging, in turn, can reduce the risk of subjective operations of business personnel, further reduce the risk of deliberate deception of customers, and gradually establish the ability to automate loan evaluation.

See Table 3 for the new customer's required score sheet, which contains at least the following items.

Step 150: After the new customer's score table is input into the computer, the risk classification group of the new customer may be calculated according to the significant variable, including the category variable and the continuous variable, and using the risk grading function.

For example,

Among them, the mixed variable value is a regression method, which combines the five financial variables in the eleventh item into a variable value obtained by “financial variable mixing”, which is also through stepwise regression analysis, regression analysis, and stepwise logistic regression. Statistical methods such as analysis or logistic regression analysis, calculate significant variables to produce continuous significant variables.

It can be seen from the above-mentioned predicted normal probability of 0.83 that this new customer is a customer of the second group of risk classification group. According to Table 1, the actual default probability of this group is about 12.23%, which is an acceptable risk. range. That is, in step 160, when the risk rating group is a new customer of P1 to P3, the financial institution's business personnel can directly sign a financing contract with the micro-enterprise according to the current loan contract, see step 170.

However, if a risk rating group is a new customer of Level 4 and Level 5 (P4 and P5), then the survival period and the solvency of the new customer need to be further calculated to determine the loan amount and repayment. For the time limit, see steps 180 to 200.

In step 180, the survival period is calculated. When the risk classification group is a new customer of the fourth level and the fifth level (P4 and P5), the survival period must be further calculated. Referring to Table 5, the method for calculating the survival period is the same. Land, through the computer system to analyze historical score data and related repayment results, using stepwise regression analysis, regression analysis, stepwise logistic regression analysis or logistic regression analysis and other statistical methods, can calculate significant variables of survival, the present invention Significant variables of survival preferably include the variables listed in Table 5.

Therefore, after calculating the survival period of this new customer, it is estimated that its survival period is about 7.36 months, so its solvency can be further evaluated. If the new customer is estimated to have a survival period of less than 6 months after the calculation of the survival period, it will be regarded as a high-risk customer, and an estimated survival period of more than 6 months will be considered as a risk of default. client.

If new customers have their risk rating groups at P4 and P5 levels, their solvency needs to be further assessed. See step 190, asking the customer to fill out or fill out the relevant financial information and enter it into the computer system at the same time as applying for the loan.

See Table 6 for the relevant financial information score sheet.

According to the above scores of the financial information score sheet, customers with default risk, that is, P4 or P5 customers with a survival period of more than 6 months, will limit the amount of their loans according to the score results.

When the score reaches 15 points or above, the maximum amount of loan that can be undertaken is about 2 million yuan. When the score is less than 15 points, the maximum amount of loan that can be undertaken is about 1.5 million yuan.

In addition, customers with high risk, that is, P4 or P5 customers with a survival period of less than 6 months, will also limit the amount of their loans or ban their loans based on the results of the ratings.

When the score reaches 30 points or above, the maximum loan amount that can be undertaken is about 1.5 million yuan. When the score is less than 30 points, the financial institution's business personnel will directly reject the relevant loan application.

Subsequent financial data scoring can further expand the source of customers and effectively avoid defaults. In addition, this financial information score can also be used by P1 to P3 customers. When the score reaches 30 or more, the loan contract can be accelerated and the customer can get the required funds.

When the survival and solvency calculations of the above-mentioned P4 or P5 customers are completed, the financial institution's business personnel can sign a financing contract with the customer based on the results of the computer system calculation and the proposed contract period and loan amount.

Therefore, the micro-enterprise financing risk assessment method of the present invention can not only effectively assess the default risk of micro-enterprise customers, but also update historical score data by using new data to more accurately estimate the risk of customer default. In addition, since the present invention mainly scores according to objective items, the subjective scoring items of the business personnel are eliminated, and the logistic regression analysis is used to estimate the loan risk of the new customers by using significant variables, and the accuracy of the system prediction is increased. It is worth noting that when the micro-enterprise financing risk assessment method of the present invention estimates that the new customer is risk of default or high default risk, it can calculate the survival time and solvency of the new customer, and then according to the time of survival and debt service. Ability to adjust the repayment method to sign financing contracts with customers to effectively reduce financing risks. In addition, the assessment of solvency can also accelerate the loan speed of low-risk customers at the P1 to P3 level, so that customers can obtain the required funds.

The above description is only the preferred embodiment of the present invention, and is not intended to limit the scope of the invention. Equivalent changes or modifications made without departing from the spirit of the invention are intended to be included within the scope of the appended claims.

110~200. . . step

The above and other objects, features, advantages and embodiments of the present invention will become more apparent and understood.

The first figure is a schematic flow chart of the micro-enterprise financing risk assessment method of the present invention.

110~200. . . step

Claims (10)

A micro-enterprise financing risk assessment method is used in a computer system to conduct a default risk assessment for a new micro-enterprise customer. The micro-enterprise financing risk assessment method includes: a historical score of a micro-enterprise that reads a database of computer systems. Data; using the computer system to calculate a risk rating group; using the computer system to calculate the probability of default for each risk rating group; inputting a new customer rating table; using the computer to calculate a risk rating group for the new customer; The computer determines the financing contract for new customers based on the risk rating group. For example, the micro-enterprise financing risk assessment method described in item 1 of the patent application scope, wherein the above-mentioned computer is used to calculate a risk classification group of new customers, and the computer is used to perform stepwise regression analysis, regression analysis, and stepwise logistic regression analysis. Or logistic regression analysis, calculate the significant variables of the historical score data, and enter the new customer score form into the computer to calculate the risk classification group of the new customer. For example, the micro-enterprise financing risk assessment method described in claim 2, wherein the significant variables include category variables and continuous variables. For example, the micro-enterprise financing risk assessment method described in item 3 of the patent application scope, wherein the above-mentioned category variables include: buying and selling retail business, personal insurance normal and no credit card pre-borrowing cash. For example, the micro-enterprise financing risk assessment method described in item 3 of the patent application scope, wherein the above-mentioned continuous variables include pre-tax net profit/total assets, current turnover/total assets, turnover growth rate, interest expenses/ Current turnover and operating margin/current turnover. For example, the micro-enterprise financing risk assessment method described in item 1 of the patent application scope includes the use of the computer to calculate the solvency of the risk classification group for P1, P2 and P3 to accelerate the loan procedure of the new customer. For example, the micro-enterprise financing risk assessment method described in item 1 of the patent application scope includes using the computer to calculate the risk grading group as the survival period and solvency of P4 and P5, and determining the financing according to the survival period and the solvency. contract. For example, in the micro-enterprise financing risk assessment method described in item 7 of the patent application scope, wherein the survival period is greater than six months, the customer is at risk of default, and the loan ceiling is further restricted according to the solvency. For example, in the micro-enterprise financing risk assessment method described in item 7 of the patent application, in which the survival period is less than six months, the customer is highly risky, and when the customer is unable to provide the guarantee of solvency, the loan is refused. For example, in the micro-enterprise financing risk assessment method described in item 1 of the patent application scope, the above-mentioned micro-enterprise customer is an enterprise with a capital of no more than NT$10 million and a loan amount of not more than NT$4 million.