Formation of a scoring model for assessing the creditworthiness of a corporate borrower. Automated scoring system for assessing the creditworthiness of borrowers Credit scoring of legal entities example

Let us consider the models of bankruptcy of an enterprise, and in more detail methods for assessing the solvency of an enterprise.

What is a scoring model for enterprise valuation?

The scoring approach to assessing the solvency of an enterprise consists in analyzing statistics on enterprises on their fulfillment of obligations to creditors, information about which is contained in the credit history bureau. Therefore, scoring models are sometimes referred to in the literature as credit scoring models ( credit-score) or credit scoring models. Thus, we can say that credit scoring models are statistical models for assessing the solvency of an enterprise.

History of the scoring approach to evaluation

Previously, scoring models were developed solely to assess the creditworthiness of individuals for the purpose of issuing loans by banks. This approach was first proposed by D. Duran in 1941 to classify bank customers into two classes: creditworthy and non-creditworthy. To determine the class, indicators were calculated to make a conclusion about its risk of bankruptcy. Scores for scoring models are calculated using the logistic regression tool. On its basis, by the way, logit-models for assessing the risk of bankruptcy of individuals and enterprises are also built.

The task of the scoring approach for assessing the solvency of an enterprise

The task of the scoring model for assessing the solvency of an enterprise is to classify it according to the degree of financial risk. The scoring approach is similar to the rating approach for assessing an enterprise, since it also has a rating (class) for an enterprise, in addition to this, there is a scoring and assigning a rating to financial indicators.

The difference lies in the fact that as a result a rating is assigned and the company belongs to the solvency class, i.e. In addition to evaluation, classification is also carried out. Also, as a result of scoring, a rating is obtained for the enterprise and a rating for financial ratios describing the enterprise.

Scoring models for assessing the solvency of an enterprise

Consider domestic scoring models for assessing the solvency of an enterprise. Let us analyze two domestic scoring models by Dontsova-Nikiforova and Savitskaya. These models are designed to assess the risk of bankruptcy of domestic enterprises. So, let's begin.

Scoring model of Dontsova-Nikiforova (1999)

Dontsova L.V.

Economists Dontsova L.V. and Nikiforova N.A. offer a scoring model for assessing the solvency of an enterprise, which allows the company to be assigned to one of the six classes of solvency, based on the assessment of six financial ratios.

Indicator	1 class(score)	Grade 2(score)	3rd grade(score)	4th grade(score)	5th grade(score)	6th grade(score)
Absolute liquidity ratio	0.25 and more (20)	0.216	0.15(12)	0.1(8)	0.05(4)	Less than 0.05(0)
Quick liquidity ratio	1 or more(18)	0.9(15)	0.8 (12)	0.7(9)	0.6(6)	Less than 0.5(0)
	2 or more(16.5)	1.7(120	1.4(7.5)	1.1(3)	1(1.5)	Under 1(0)
	0.6 and more(17)	0.54(12)	0.43(7.4)	0.41(1.8)	0.4(1)	Less than 0.4(0)
Working capital ratio	0.5 and more(15)	0.4(12)	0.3(9)	0.2(6)	0.1(3)	Less than 0.1(0)
Reserves coverage ratio	1 or more(15)	0.9(12)	0.8(9)	0.7(6)	0.6(3)	Less than 0.6(0)
The minimum value of the border in points	100	64	50	28	18	–
Grade 1>100 points	The company has a good margin of financial strength
Grade 2>64 points	The company has an insignificant probability of paying off debts, in general, there is a risk
Grade 3>50 points	Troubled Enterprise
Grade 4>28 points	The company has a high risk of bankruptcy
Grade 5>18 points	The company has a very high risk of bankruptcy, recovery measures are unlikely to help
6th grade<18 баллов	The company is financially insolvent

Note:

In the valuation model, the main emphasis is on liquidity ratios (, quick liquidity ratio, absolute liquidity ratio), as well as on turnover ratios (equity ratio, inventory ratio).

Odds	Formula	Payment
Absolute liquidity ratio	(Cash + Short-term financial investments) / Short-term liabilities	p.1250 / (str.1510+str.1520)
Quick liquidity ratio	(Current Assets - Inventories) / Current Liabilities	(p.1250+p.1240) / (p.1510+ p.1520)
Current liquidity ratio
Coefficient financial independence	Equity / Assets	p.1300 / p.1600
Working capital ratio	(Equity - Non-current assets) / Current assets	(p.1300-p.1100) / p.1200
Reserves coverage ratio	Inventory turnover ratio= Sales revenue / Average inventory	p.2110 / (p.1210 np.+p.1210 kp.)*0.5

n.p. and k.p. - the value of the balance line at the beginning of the period and the end of the period, respectively.

Savitskaya scoring model (2007)

Savitskaya G.V.

Professor G.V. Savitskaya offers her own scoring credit model for assessing the financial condition of an enterprise. The difference lies in the fact that in the model the enterprise is classified according to five classes and three financial ratios are used for this.

Indicator	1 class	Grade 2	3rd grade	4th grade	5th grade
Return on total capital, %	30 and above (50 points)	29.9-20(49.9-35 points)	19.9-10(34.9-20 points)	9.9-1(19.9-5 points)	Less than 1(0 points)
Current liquidity ratio	2 or more (30 points)	1.99-1.7(29.9-20 points)	1.69-1.4(19.9-10 points)	1.39-1.1(9.9-1)	1 and below (0 points)
	0.7 or more (20 points)	0.69-0.45(19.9-10 points)	0.44-0.3(9.9-5 points)	0.29-0.2(4.9-1 points)	Less than 0.2(0 points)
Class boundaries	100 points	99-65	64-35	34-6	0 points
Grade 1>100 points	An enterprise with good financial strength
Grade 265-99 points	The company has a small risk of non-repayment of debts
Grade 335-64 points	Troubled Enterprise
Grade 46-34 points	The company has a high risk of bankruptcy. Lenders risk losing their invested funds
Grade 50 points	The enterprise is insolvent

Note:

Two of the three financial ratios determine the solvency of the enterprise, where the current liquidity ratio determines the short-term liquidity, and the financial independence ratio determines the long-term liquidity of the enterprise.

Financial independence ratio = autonomy ratio.

Calculation of financial ratios in the scoring model

Odds	Formula	Payment
Return on total capital	Profit before tax / Liabilities	p.2300 / p.1700
Current liquidity ratio	Current assets / Current liabilities	p.1200 / (p.1510+p.1520)
Financial Independence Ratio	Equity / Assets	p.1300 / p.1600

Summary

Let us sum up the analysis of credit scoring models for assessing the solvency of an enterprise. One of the indisputable advantages is that these models were developed for domestic enterprises. One of the difficulties in estimating using such models is the large complexity of the calculations and often the incomprehensibility in using the scoring of financial ratios. Their use is well combined with other methods of assessing the financial condition.

Thank you for your attention! Good luck!

Lending

scoring of individuals

in the consumer credit market

i.n. rykova,

doctor economic sciences Moscow city

Lately Special attention given to creditworthiness assessment individual using a scoring (scoring-based) system for selecting key financial indicators1.

Thus, for the first time proposed in 1941 by the American economist David Duran, such a system makes it possible to estimate the "weight" of financial and economic factors affecting competitiveness. It should be noted that each key factor (indicator) receives a numerical value in points corresponding to the level of its significance. In the future, based on the results of such ranking, a scoring scale is compiled in the form of a table grouped by factors.

Thus, scoring highlights those characteristics that are most closely related to the creditworthiness of individual borrowers, so it is important to ensure the correct selection of such characteristics and determine the appropriate weighting factors for them. In addition, a distinctive feature of the scoring method is that it should not be applied according to a template, but should be developed independently by each bank based on the characteristics inherent in it, taking into account the traditions of the country, changes in socio-economic conditions, etc. In order to widely implement scoring, the bank must analyze the effectiveness of the current model and, if necessary, modify the set of characteristics and the scale of their numerical assessments.

Most credit institutions use credit scoring to assess individuals

1 Proskurin V. A. Scoring method for assessing the creditworthiness of individuals//Business and banks. -2000.-No. 30.

persons in the consumer lending market (Table 1).

Let's immediately determine that with a final indicator of 1.25 points or more - the client belongs to the group of insignificant or moderate risk, less than 1.25 points - an undesirable client.

The next step in assessing the creditworthiness of a borrower in credit institutions is to conduct a more detailed analysis of an individual, taking into account its characteristics (Table 2).

One of the main advantages of express loans is that often banks do not require collateral. With express lending, the solvency of the borrower can be assessed either "manually" by bank employees (it takes 2-3 days to process), or using a special computer program - a scoring system (the decision is made in one day).

The principle of assessing solvency for express car loans, as well as for consumer lending: the program contains the data of the borrower's questionnaire, each indicator is assigned a certain value (point), and based on their amount, a decision is made to grant a loan or refuse.

The scoring method can significantly reduce the time for consideration of an application, however, it increases the risks of the bank, which accordingly compensates for them with high rates. Typically, one-day loans are 3-4 percentage points more expensive than manually assessed express loans. In addition, the maximum amounts of scoring loans are usually very limited; this assessment method is the most technologically advanced product that allows you to reduce bank costs and rapidly increase your loan portfolio.

Table 1

Duran credit scoring technique

Points indicator Maximum amount points

Age 0.01 points for each year over 20 0.3

Gender Female - 0.4 Male - 0 -

Length of residence in the area 0.042 for each year of residence in the area 0.42

Profession Low risk - 0.55 High risk - 0 Other professions - 0.16 -

Work in the industry Public sector enterprises, government agencies, banks, brokerage firms - 0,21 -

Employment For each year of work at the enterprise - 0.059 0.59

Finance Having a bank account - 0.45 Owning real estate - 0.35 Having a life insurance policy - 0.19 -

table 2

credit scoring system for assessing the creditworthiness of individual borrowers

Client characteristics Points Client characteristics Points

1. Age of the client: 6. Profession, place of work:

less than 30 years 5 manager 9

less than 50 years old 8 skilled worker 7

over 50 years old 6 unskilled worker 5

student 4

pensioner 6

unemployed 2

2. Availability of dependents: 7. Length of employment:

no 3 less than 1 year 3

one 3 less than 3 years 4

less than 3 2 less than 6 years 7

over 3 1 over 6 years 9

3. Living conditions: 8. Availability of bank account:

own apartment 10 current and savings 6

rental housing 4 current 3

other (lives with friends, family) 5 savings 2

at the present address: 2 (including other financial institutions):

less than 6 months 4 one 3

less than 2 years 6 more than two 5

less than 5 years 8 no 1

over 5 years

5. Client's income (per year), $:

more than 50 000 9

Let's consider the methodology for assessing the borrower's solvency, which is used, for example, in the Autoexpress Credit program, and we will carry out a calculation based on it. We denote the financial capabilities of the client conditionally in Table 3.

Based on this, the share of the monthly payment on the loan will be calculated according to the formula (1):

Table 3

Financial capabilities of the client - an individual

1. Living wage in the lending region Pm

2. Persons in custody, number L

Income: 3. Average salary for the last 3 months. W

4. The annual amount of regular income taken into account as sources of loan repayment Pd

5. Final average monthly income Sd = W + Pd/12

Costs: 6. Maintenance costs Rs = (L + 1) x Pm

7. Monthly payment for an apartment (when receiving, renting) PC

8. Pu annual tuition fee

9. The annual amount of contributions for voluntary insurance Sun

10. Payments to repay the current debt on loans, credits, interest on them (average for the last 3 months)

11. Other expenses (alimony, court deductions, etc.), average for the last 3 months. Etc

12. Total average monthly consumption Cp=Rs+Pk+Pl+Pr + (Pu+Sun) /12

13. Average monthly disposable income Рd = (Сд-Ср)

The assessment will be made according to the criterion of 100 x (1 - Dp). The maximum score for the criterion is 30.

To determine the assessment according to the criterion "Financial capabilities of the client", the following documents are required from the client:

■ certificate from the employer on the client's income for the past year and for all full months of the current year, signed by the chief accountant and stamped;

■ documents confirming additional income. The next step in assessing the solvency of an individual is the sufficiency of the client's unmortgaged property (Table 4).

Based on this, when determining the assessment for this criterion, the sufficiency of property is assessed according to the formula (2):

It is also necessary to assess the loan collateral (Table 5).

The maximum score for this criterion is 25.

The next stage in assessing the solvency of an individual is the criterion for lending conditions (Table 6).

When determining according to this criterion, the client is required to have an extract from the client's account in the bank

Table 4

Adequacy indicators of the client's unmortgaged property

The assessment will be made according to the criterion of 5 x Di. The maximum score for the criterion is 5. When determining the grade, the following documents are provided:

■ documents confirming the existence of ownership;

■ insurance policies on property.

Characteristic Legend

1. Contributions to

2.1. Securities of the Central Bank

2.2. Grade valuable papers Ocb \u003d Cb / 2

3.1. Own apartment Sq.

3.2. Sum insured Ks

3.3. Apartment appraisal Ok = min (Kv, Ks)

4.1. Own house SD

4.2. Sum insured Ds

4.3. House valuation Od \u003d min (Sd, Ds)

5.1. Dacha Dch

5.2. Sum insured Dhs

5.3. House valuation Odh = min(Dh, Dhs)

6.1. Car A

6.2. Sum insured Ca

6.3. Car rating Oa = min (A, Ca)

7.1. Other property of Ii

7.2. Sum insured C

7.3. Valuation of other property Оi = min (Ии, Сi)

8. Property Im=V+Otsb+Ok+Od+Odch+Oa+Oi

Table 5

Loan security

Characteristic name Legend

1. Estimated value of collateral Oz

2. Collateral discount, % Zd

Security Ok \u003d Oz x (1 - Zd) / Kr x (1 + 2 x St / 12) 100 x (1 - Ok)

Table 6

Terms of lending to individuals

Feature Value Score by criterion

1. Financing the purchase by the client F 7 x ((F/ (F + Kr))

2. Loan term, months Wed 3 x (Ms - Wed) / (Ms - 1)

Final score according to the criterion Sum of parameter scores

ke. The maximum score for the criterion is 10. Depending on the points scored, the loan falls into one of the quality categories (Table 7).

■ the bank's client does not permanently reside in the city (suburb) of the location of the bank's lending unit or the period of his permanent uninterrupted residence in this city (suburb) is less than one full year;

■ evaluation according to the criterion "Customer Character" is not positive;

■ the evaluation according to the criterion "Financial opportunities of the client" is negative;

■ the score according to the criterion “Credit security” is equal to zero.

We will carry out a further assessment of the client of the credit institution according to the criteria presented in Table. 3 - 7. We present the data in Table. eight.

Final credit score

Table 7

The number of points scored in assessing the quality of the loan Quality category Score

From 30 to 65 inclusive 2 Application is inadequate for the requested loan

Table 8

Assessment of the client's solvency

Lending options

Financial possibilities of the client

Z (rub.) * 67 500

Pd (per year) -

Pu (per year) 20,000

Sun (per year) -

Pl (per month) -

Evaluation by the criterion (points) Ots. 1 - 41. Maximum - 30

Sufficiency of unmortgaged property

The end of the table. eight

Evaluation according to the criterion Ots. 2 - 31.25. Maximum amount by criterion - 5

Loan security

Evaluation according to the criterion Ots. 3 - 10.

Maximum score - 25

lending

Evaluation according to the criterion Ots. 4 - 2.9. Maximum - 10

Sum of points = Ots. 1+Es. 2+Es. 3+Es. 4+Es. 5 - - 74.15 points.

Conclusion on the creditworthiness of the client - over 65;

* Based on 1 US dollar - 27 rubles.

most fully reflect all the characteristics of a potential client - an individual when receiving consumer loans.

The practice of mass application of scoring methods in Russian conditions can lead to a sharp increase in loan defaults2. The positive experience of their successful use in economically developed countries was formed in a completely different way. economic environment. In Russia, in the absence of credit bureaus, low credit culture of the population, and a single information space in the financial sector, the massive use of foreign scoring technologies will undoubtedly increase the credit risks of the retail banking business. In this regard, the improvement of methodological approaches to assessing the creditworthiness of individual borrowers, the adaptation of existing foreign experience to Russian characteristics seems to be a very important task.

To assess the solvency of the client, loan officers need to analyze a huge number of documents. Their list is quite large and includes about 15 items. Their obligatory provision by the client, on the one hand, limits the range of potential borrowers of the bank, and on the other hand, allows you to create a higher quality loan portfolio and reduce credit risk.

One of the advantages of this technique is the use of special formulas and correction factors that make it possible to simplify the work of the bank's credit department employees and calculate the solvency of a potential borrower. However, indicators for it should be obtained in each specific situation separately, and the result should not be considered as something that indicates

2 Voroshilova I. V., Surina I. V. On the issue of improving the mechanism for assessing the creditworthiness of individual borrowers. Materials of KubGAU, Krasnodar. - 2005.

which is unambiguously in favor or against the issuance of a loan. After all, even if at the time of consideration loan application financial indicators of the client are at an acceptable level, do not forget that the risk of non-repayment of the loan still remains, since it is impossible to completely eliminate it in principle. The indicators will only help to assess the degree of credit risk and, unfortunately, this technique does not allow predicting the position of the borrower in the future.

The most important point in the underwriting process is the assessment of the client's solvency in terms of the ability to make timely payments on the loan. To perform this assessment, information is consolidated on employment and the receipt of income by the borrower, as well as on its expenses. After that, it is concluded whether he will be able to repay the loan. At the same time, an opinion is issued whether the pledged property is sufficient security for granting a loan or not.

As noted in the proposed methods for credit institutions, among the quantitative characteristics is the ratio total amount monthly obligations of the borrower to the total family income for the same period, as well as the sufficiency Money(based on maintenance costs). Qualitative characteristics include the borrower's income, employment stability, credit history, credit security, etc.

Evaluating the underwriting methodology, we can conclude that a systematic approach to the analysis of the borrower is applied here. The positive side of the methodology is the ability of the bank to develop an individual approach to any potential borrower, within which the required number of characteristics will be taken into account. The negative side is the laboriousness of its implementation, which requires special qualifications of bank employees. Most banks prefer

FINANCE AND CREDIT

offset credit risk by raising interest rates. Other methods are also used, the application of which does not require large expenditures of time and labor3.

It should be noted that the understanding of the expediency and relevance of using more advanced methods occurs most often in those banks in which lending to individuals is implemented as a mass service. If the bank plans to launch a large-scale program, then in order to succeed in the market in the face of constant toughening of competition and, as a result, a decrease in profitability, it is necessary to look for ways to reduce operating costs and minimize risks.

A prerequisite here will be the correct construction of the mechanism that will carry out this activity. Figuratively speaking, it is necessary to create a kind of pipeline consisting of a certain number of employees interacting with borrowers and among themselves according to certain clearly defined rules and algorithms. Such algorithms include methods for analyzing applications and making decisions on issuing a loan.

In the process of analyzing data on borrowers and loans, various mathematical methods are used to identify factors and their combinations in them that affect the creditworthiness of borrowers, and the strength of their influence. The discovered dependencies form the basis for making decisions in the corresponding block.

The proposed approaches to improving the organization of the process of lending to individual borrowers at the stage of assessing their creditworthiness will make it possible to unify the procedure, on this basis to speed up and reduce the cost of it, to obtain a more accurate and reasonable result, which will ultimately reduce lending risks, ensure the necessary stability of the bank and a given level of profitability.

Thus, the basic issue of lending to individuals is a reliable classification of potential borrowers into "good" and "bad". The following problems in the field of lending to individuals facing banks today can be distinguished:

1. The absence of special legislation regulating relations in the field of consumer lending. Relations in this area

3 Voroshilova I. V., Surina I. V. On the issue of improving the mechanism for assessing the creditworthiness of individual borrowers. Materials of KubGAU, Krasnodar. - 2005.

are regulated by the Law of the Russian Federation of December 19, 2000 No. 238-FZ “On Banks and Banking Activity” and the Law of the Russian Federation of February 7, 1992 No. 2300-1 “On Protection of Consumer Rights”.

2. Absence credit history. This gives a lot of opportunities to unscrupulous borrowers who can get several loans from different banks without any verification of their previous credit exploits.

3. Used salary schemes of enterprises. Employers often prefer "gray" schemes for remuneration of their employees. The borrower cannot officially confirm the level of income, and the bank loses a solvent client.

4. There is no simple mechanism to return money to the investor in the event of the borrower's insolvency. The cost of such errors is very high: loss of principal, legal fees, administrative costs, lost time, etc.

5. Problems of classification. Reliable assessment of a potential borrower, cutting off "bad" borrowers is necessary. Misclassification creates the problem of enforcing borrower repayment.

6. The problem of collateral. The mechanism for the implementation of collateral is an inconvenient and expensive exercise. Lack of pledge registration movable property allows you to sell or re-mortgage the pledged property by an unscrupulous borrower.

There is also the problem of assessing the real possibilities of guarantors, which is due to the fact that most Russian banks solve the issue of reducing their credit risks by simply transferring them to the borrower's guarantors. At the same time, various legal entities (both large and medium and small enterprises) are often guarantors, especially for large loans. In the context of future plastic loans, this practice will be applied everywhere, since it is convenient to give the borrower plastic card, and in case of any difficulties with the return of the loan, claim it from the guarantor - the enterprise where he works. On the face of it, this should solve the problem, but on a broader view, this lending policy does not guarantee success to the extent that banks rely on.

Having studied the main problems of lending to individuals, we can propose the following ways to solve them:

1. For the development of the mortgage lending market, it is necessary, first of all, to reduce

interest rate for a loan by eliminating the risk of non-payment from it. It is also necessary to introduce a number of amendments and additions to some legislative acts. Russian Federation aimed at creating an affordable housing market.

2. For the development of the educational lending market, it is necessary:

The legislative framework providing financial assistance for all who want and are able to get an education;

A loan repayment guarantee by the state, allowing it to take a significant part of the risks.

3. In a competitive environment, the one who minimizes risks will win by reliably determining which client is “good” and which is “bad”, and will offer borrowers more profitable terms.

4. Banks need to take advantage of advanced technologies and apply them to evaluate potential borrowers. Thanks to this, it will be possible not to be afraid of the upcoming competition in this market. In such a situation, banks that decide to develop this market should have the following:

Consolidated customer information presented in a unified form. Information must be periodically updated with data from all branches of the bank. Such a vault would act as a credit bureau;

Reliable method of classification (reliability should be more than 90%) of potential borrowers and cutting off "unreliable". This method will reduce the risks of default to a minimum, which will allow you to issue cheaper loans and, accordingly, attract more borrowers. At the same time, profit from lending to individuals will increase significantly;

The borrower classification model should have the properties of replicability and adaptation to the state of the market, to each branch of the bank. That is, built on the basis of general patterns, the model should be adjusted to the particular features inherent in each branch. This will take into account local features, which will further reduce the risk;

The classification model should be periodically rebuilt, taking into account new market trends. This achieves its relevance.

After all, one and the same cannot be used.

same approach 5 years ago and now.

At the moment, banks, to one degree or another, have experience in each of these items, but the methods underlying them are either too inert to adequately respond to market dynamics, or too expensive (proposed foreign solutions are comparable to income from consumer lending in present form). That is why loans are so expensive and the demand for them is not so great. An increase in reliability and a decrease in cost will make it possible to abandon the practice of transferring risks and costs to borrowers. Then everyone will benefit - both banks, keeping their specific profitability at the same level, and borrowers attracted by more favorable conditions. All this becomes more relevant in view of the future booming consumer credit market and future competition.

One of the main measures to prevent possible losses is a correct assessment of the borrower's ability to fulfill its obligations. The choice of criteria for it was relevant in all periods of the development of banking and has already entered the economic literature as one of the main tasks in determining the creditworthiness of a borrower. No less important is the problem of proper organization of the procedure for assessing creditworthiness as the most important in the credit process.

The creditworthiness of the client in the world banking practice appears as one of the main objects of assessment in determining the feasibility and forms of credit relations. The ability to repay a debt is associated with the moral qualities of the client, his art and occupation, the degree of capital investment in real estate, the ability to earn funds to repay loans and other obligations.

The list of elements of the borrower's creditworthiness and indicators characterizing them may be broader or shorter depending on the objectives of the analysis, types of loans, loan terms, the state of the bank's credit relations with the borrower. The optimal or acceptable values ​​of such indicators should be differentiated depending on the activities of the borrower, the specific terms of the transaction, etc.

To date, there are several basic methods for assessing the creditworthiness of customers. The systems differ from each other in

the number of indicators that are used as components of the overall assessment of the borrower, as well as different approaches to the characteristics and priority of each of them.

Scoring models are used mainly when providing loans for the purchase of goods (express loans) and when issuing credit cards. Scoring is a mathematical (statistical) model by which, based on the credit history of existing customers, the bank determines how likely it is that a particular customer will repay the loan on time. Scoring highlights those characteristics that are most closely related to the reliability or, conversely, to the unreliability of the client.

The credit scoring technique is an assessment in points of characteristics that make it possible to determine with sufficient reliability the degree of credit risk when providing a consumer loan to a particular borrower. The most significant indicators for predicting credit risk may be such indicators as age, number of dependents, profession, income, cost of housing, etc.

The advantages of scoring models are obvious:

1) decrease in the level of loan default, speed and impartiality of decision-making;

2) the possibility of effective management of the loan portfolio;

3) lack of long-term training for employees of the credit department;

4) the ability to conduct an express analysis of the loan application in the presence of the client.

However, despite the positive aspects, the use of credit scoring is associated with a number of difficulties. One of them is that the evaluation characteristics are determined only on the basis of information about those clients to whom the bank has already provided a loan. Another and most significant problem is that scoring models are built on the basis of a sample of the most "early" customers. Given this, bank employees have to periodically check the quality of the system and, when it deteriorates, develop a new model.

It should be noted that about ten characteristics are taken from the application form filled out by the borrower for evaluation, and the rest of the data is stored in the statistical database for further updating and scoring analysis. At the moment, Russian banks evaluate such characteristics as income, number of dependents, ownership of a car (at the same time, a car of domestic and foreign production is distinguished, necessarily taking into account the period that has elapsed since its release), the presence land plot(its area and distance from the city center are considered), work experience, position, education.

Credit risk is the main banking risk, the management of which is a key factor in determining the effectiveness of the bank's activities. Typically, banks form a significant part of their income through lending activities, so the assessment of potential profit in relation to the probability of loan default is of particular relevance. In a narrow sense, credit risk is defined as the risk for a lender that a borrower will not repay borrowed funds.

Recently, in our country, there has been an intensive growth in the lending market and, in particular, in the sector of lending to individuals. This inevitably leads to an increase in credit risks, which are assumed by both individual financial institutions and banking system the country as a whole.

The growth of risks is caused simultaneously by the expansion of the contingent of borrowers and the increase in lending volumes. In this situation, the quality of credit risk management in retail lending is of particular relevance and becomes one of the factors for increasing competitiveness. credit institution in the banking services market.

When issuing a loan, the bank, first of all, is interested in the creditworthiness of a potential borrower, that is, the ability to fully and on time pay off its debt obligations. It is the task of choosing creditworthy borrowers that scoring systems mainly serve.

The word scoring itself is translated from English as “scoring” and literally means a point-based assessment when making a decision. This is a special system invented by cunning overseas financiers and programmers. It allows you to quickly and efficiently assess the degree of risk of a potential borrower. In other words, the scoring program decides whether or not to give this person a loan, depending on whether the borrower scores a certain number of points or not. Of course, this decision is not made in a vacuum. It is backed up by comparison with cases already in the database. Simply put, the program identifies all borrowers with such data and calculates the percentage of returned and non-repaid loans, as well as the degree of possible risks.

Priority in formulating the concept of credit scoring is usually given to David Duran, who, in his study published in 1941 in the National Bureau of Economic Research, used 7,200 "good" and "bad" credit histories of loans with regular repayments provided by 37 firms. He applied the chi-square test to identify characteristics that markedly distinguished "bad" from "good" and developed an efficiency index designed to show how effective this characteristic to differentiate the degree of risk ("good"/"bad") among loan applicants. He then used the discrimination function to develop credit scoring models.

According to Fair Isaac, over 90% of banks in developed countries currently use credit scoring. It is becoming more and more popular among Russian banks.

So, the primary task that is solved when lending with the help of scoring is risk assessment and risk management based on the forecast, with what probability a particular borrower can overdue loan payments, that is, a statistical method for assessing the borrower's creditworthiness.

Secondly, scoring is a process of automating decision making. In the process of granting a loan, banks are interested in studying the solvency of the future consumer of the loan. The purpose of this study is to model or predict the likelihood that a loan applicant can be classified as an attractive or unattractive client.

In practice, scoring looks like decision trees - this is a model built on a logical chain of rules that try to describe individual relationships between data relative to the expected result. The structure of decision trees openly shows the reasoning behind the rules and therefore makes it easy to understand the decision making process.

If a credit institution uses credit scoring correctly and adequately, it will gain an effective competitive advantage to maintain and improve its competitive position in the market and survive in the fight against competitors for a long time.

Professional bankers have a saying: "A scorecard is only as good as the data."

When scoring, almost everything is taken into account: gender, age, profession, work experience, family and children, the amount requested as a loan ... The age of a creditworthy borrower is usually 25–45 years. Theoretically, banks promise to issue loans to those who are 21 years old and those who are already 50. But in practice, everything looks very different.

The same is true with marital status. The Bank wishes to see respectable and responsible citizens as its clients. Having a family is a sign of responsibility. Therefore, ceteris paribus, the bank is more likely to give a loan to a married client than to a single one.

A huge role in the scoring assessment will be played by the profession. It's one thing to be a teacher who peacefully explains to the kids how to calculate the discriminant, and another thing is to be a firefighter who, at the risk to the lives of these same kids, if anything, saves. With the same income, age and marital status, a loan is more likely to be given to a representative of a less dangerous profession. Banks also consider specialties related to the entertainment industry to be “unreliable”, primarily gambling and show business. True, banks are often loyal to employees of cafes and restaurants - people always want to eat. But the empty belly is deaf to spectacles. So the waiter will have a better chance of getting a loan than a singer playing live music in the same restaurant.

Do not forget about the financial crisis, great and mighty. If earlier builders and realtors were desirable and reliable clients, now their areas of work are considered very unstable. The same goes for brokers, traders, workers in the insurance and tourism industries. Therefore, do not be offended by a bank that refused a loan - the risk is too high.

But banks are now loyal to the "state employees". Doctors, teachers, lecturers, some police officers (primarily investigators and interrogators), even railroad workers have a good chance of getting a loan. "Highly respected" by many banks and representatives of "timeless" working professions - hairdressers, cooks, car mechanics. The main thing is that the borrower should be an employee.

Banks are not too happy with mid-level businessmen: not only can their business go bankrupt, but entrepreneurs also cause more trouble in servicing a loan. In the event of the loss of an “armchair” with a large income, it is not easy today to quickly find a new job with a similar salary. It is also difficult for officials to take out a loan with their small salary: unfortunately, “kickbacks” are not indicated in the 2-NDFL certificate.

Any self-respecting bank also checks information about the place of work. But he does this no longer with the help of a scoring program, but through databases pension fund(like Sberbank). Or with the help of their own security service, which can at the same time check other circumstances of your life. For example, the ideal borrower should not have a criminal record or "dark spots" in the biography, such as being captured by Somali pirates.

To become an ideal borrower, you sometimes need to meet rather original requirements. For example, it is desirable to work in the specialty that you received at the institute - especially if you just recently graduated from this same institute. By the way, with the arrival financial crisis many banks no longer inspire confidence with a work experience of 3 months in one place - at least six months, and preferably several years. At the same time, career growth in the biography will also be a big plus.

The more qualities that match the portrait of an ideal borrower, the more chances for a positive verdict. But sometimes the logic of the scoring program produces solutions that seem at least funny to an inexperienced client. The girl manager, smiling, says that she was given a consumer loan without any problems, and her common-law husband was refused in the same bank, although their age, field of activity and income are exactly the same. The scoring analysis showed that men are more likely to default, and the program ruled in favor of the girl.

The ideal borrower today is a person 30-40 years old, receiving a “white” salary, with a spouse (or spouse), who has a higher education and a solid work experience. “And if this person also has property, then this can be regarded as a“ lifeline ”in case of force majeure situations, from which no one is immune today.

It should be remembered that banks do not forgive deception. Never. If the security service finds out that potential borrower provided deliberately incorrect data (for example, the form 2-NDFL of the guarantor “drawn”), they will not give a loan.

The main advice that bankers give to borrowers is to tell the truth about yourself, the whole truth and nothing but the truth. According to them, it is not so scary to have five children as it is scary to indicate in the questionnaire that there are four or three of them.

However, if one bank refuses a loan, another may well issue it, because each bank has its own credit strategy.

Thus, a potential client is assessed using a scoring model - a kind of tester that determines the mathematically expressed characteristics of the borrower, indicating its ability to repay the loan on time. Since the basis for credit scoring is statistical laws, the evaluation of a potential borrower is not always correct. In the event of an error, the bank loses money - either the borrower does not repay the loan, or the bank unfairly refuses the potential client.

Benefits of scoring:
– optimization of costs for consideration of the application by automating the process of decision-making and issuance of a loan;
- reducing the time for consideration of an application, increasing the number and speed of processed applications;
- the absence of subjective opinion of an expert when making a decision on issuing a loan;
– determination of the level of profitability and risk loan portfolio etc.;
– detection and prevention of fraudulent attempts.

Disadvantages of scoring:
- the program evaluates real person, and the information that he reports about himself, and a well-prepared client can present data about himself in such a way that he is almost guaranteed to receive a loan;
- creditworthiness is assessed on the basis of data on those borrowers to whom the loan was issued. One can only guess about the behavior of borrowers who were denied loans.

In addition, scoring models require constant refinement and updating, as both socio-economic and lending conditions and people themselves change over time.

In Russia, the development of scoring (in its original meaning) is limited by the still low volumes of lending by Western standards, as well as rapidly changing socio-economic conditions. Russian banks and rating agencies do not have sufficient information about clients to build effective mathematical models that ensure the demand for retail lending, on the one hand, and minimize banking risks, on the other.

To solve this problem, banks can be offered to act in two ways.

The first way is to use a model developed abroad, with its obligatory "calibration" adaptation to Russian realities, which will require both time and money.

The second way is to abandon the use of scoring at the initial stage and issue loans to everyone on the basis of a standard check in order to accumulate the necessary credit history. After that, banks will be able to develop their own scoring model based on these peer review data, which is rather intuitive, very effective, but also more expensive for a bank or agency.

So, scoring is an automated credit risk assessment system that is widely used in the United States and Western Europe, and is also gaining more and more popularity in Russia. The source material for scoring is a variety of information about past clients, on the basis of which, using various statistical and non-statistical classification methods, a forecast is made about the creditworthiness of future borrowers. Scoring systems, having their pros and cons, allow bank employees make quick decisions on lending, adjust lending volumes depending on the market situation and determine the optimal ratio between the profitability of lending operations and the level of risk.

The concept of scoring customer creditworthiness

Definition 1

Scoring is a statistical or mathematical model that uses data from the credit histories of bank customers, and ultimately it is possible to calculate the probability that the next potential borrower will return the funds received on time.

This borrower assessment methodology is a weighted sum of a certain set of characteristics in a very simplified form. This is necessary for the formation of a summary indicator. This indicator is further compared with the so-called break-even line.

Such an assessment of the borrower's solvency is needed to determine the integral indicator of each potential client, and the result obtained must be compared with the above line (accordingly, only those borrowers who have this indicator above the break-even line will be able to receive a loan).

Usually in national economy banks use adapted models of scoring assessments of the creditworthiness of an individual, which are adapted to Russian conditions.

First, a preliminary assessment of the possibility of obtaining a loan is given, based on the data of the questionnaires-applications of borrowers. Based on the results of the completed application forms, protocols for evaluating the possibility of granting loans are signed.

Example 1

If the score is less than 30, the protocols record a refusal to provide a loan, but if more than 30 points were scored, then at the next stage the risk is assessed more carefully, taking into account additional surveys.

Advantages and disadvantages of credit scoring

Scoring methods and models allow:

• reduce the risk of loan default;
• make decisions on issuing a loan quickly and impartially;
• allow you to effectively manage your loan portfolio;
• no need to spend a lot of time training employees of the credit department;
• it is possible to conduct an express analysis of the loan application in the presence of the client.

The limitations of the scoring methodology include the fact that it can only be applied to information about those customers to whom the bank has already issued a loan. Also, bank employees have to periodically check the quality of the methodology and analysis and develop a new scoring methodology.

Further improvement of the scoring methodology will expand and change the list of assessed characteristics of loans.

At mortgage lending citizens use the underwriting of the borrower, the most important thing is the assessment of the timely installment of loan payments. The ratio of the size of the monthly obligations of the borrower to the total family income for the same period, etc. is estimated.

The process of conducting a scoring assessment of the creditworthiness of borrowers

Typically, to analyze the creditworthiness of a potential borrower, the following are requested:

• a copy of the documents proving the identity of the borrower;
• confirmation of the client's income: certificate in the form 2-NDFL, copy tax return in the form 3-NDFL;
• Additionally, they may also request property ownership documents and other documents that can confirm the solvency and business reputation of the client.

The bank's specialists analyze the solvency of an individual borrower based on the data on the average monthly income and the amount of deductions for the previous six months, as well as information based on the questionnaire. The result is calculated as the average monthly income minus all mandatory payments and is adjusted by a correction factor that varies depending on the amount of income (from 0.3 to 0.6). The greater the income, the greater the adjustment.

Remark 1

At the moment, the most universal method of assessing creditworthiness is the method of assessing the financial position of the client.

To reduce and control risks, banks must assess the financial condition of the borrower on a quarterly basis.

As an improvement in the assessment of the creditworthiness of individuals, it is proposed to use a scoring system when determining the volume of loans issued.

Credit scores are designed to measure the risk of default by a potential borrower, taking into account various factors of credit history. Formulas for calculating credit scores are usually Western banks are not disclosed, but, in general, the following components are used, which can be considered as applicable experience:

1. 35% is credit history - the presence or absence of compromising information. Bankruptcy, pledges, judgments, agreements, confiscations, foreclosure of property, late payments can be the reason for refusal to issue a loan.
2. 30% Leverage – This category looks at a number of specific dimensions of leverage, including the number of overdraft accounts, existing debt obligations, installment purchases.
3. 15% share falls on the term of the credit history - the average period of lending and the term of the original loan.
4. 10% is an assessment of the types of credit used (installment plan, overdrafts, consumer loans, mortgages), shows the history of managing various types of loans.
5. A 10% share of the score falls on the number of loan requests - the borrower's rating is reduced if requests have been made in large numbers recently (14-45 days).

Scoring models should be based on up-to-date data and quickly reconfigure when changes are made. credit policy jar.

The credit bureau plays an important role in the work of the scoring model. It is necessary to study the credit history of the potential borrower and the spouse of the applicant. All types of income and expenses of the borrower must be documented.

Example 2

Loans should not be issued to citizens who have payments under executive documents in the amount of 50 percent or more net income. Also, a guarantee of an individual whose deductions from wages equal to or greater than 50 percent of net income.

Figure 1 presents information developed foreign banks to obtain information about the purpose of the loan, the personal characteristics of the borrower and the credit history of the borrower.

Figure 1. Variables used in scoring models for assessing the creditworthiness of borrowers. Author24 - online exchange of student papers

When assessing the credit risks of potential borrowers, a number of factors are taken into account: age, marital status and education, the number of his/her dependents, the client's place of residence, profession, length of service, current work experience. As well as the following financial information: client's regular income and liabilities; credit history, which includes facts such as high-quality loan repayment; previous positive cooperation with the bank, if the client is already a client of the bank.

When issuing loans, banks seek to maximize profits and guarantee the return of funds transferred to the borrower. In order to reduce the risk of delays, financial institutions carefully analyze all applicants and approve only applications, the obligations of which will be fulfilled with a high probability.

The assessment of the creditworthiness of a borrower - an individual is often carried out using (from English scoring - "scoring"). scoring model analyzes the factors influencing the risk of non-repayment of the loan, and issues recommendations for approval or refusal of the application. When applying for a loan, the borrower is first asked to fill out a questionnaire. It is on the basis of these data that the assessment is made. For each parameter, the client receives a certain number of points, there are increasing and decreasing coefficients. The final result was previously calculated manually bank employees, today this is done automatically in special programs.

Where is scoring applied?

The scoring model is widely used in the field of microfinance and express lending, where reviewing the data of a potential borrower and making a decision takes less than 1 hour. To check creditworthiness in special program enter information from the completed application. The system automatically compares the data provided by the potential borrower with statistics. So, if the database contains information that people of the same age or profession often do not repay the loan, then the decision on the application may be negative. In such cases, a bank or microfinance organization usually refuses a potential borrower without explanation.

Benefits of the CREDIT SCORE SCORING SYSTEM

Decision making speed. If a bank employee analyzes the solvency of the borrower, this will take a lot of time. The specialist needs to independently check each parameter, manually enter all the results obtained and draw a conclusion. With the help of modern scoring systems for assessing creditworthiness, data is processed quickly, which means that a decision is made promptly.

Objectivity. Even an experienced and qualified specialist can make a mistake in the calculations or form a biased opinion due to a personal attitude towards the client. The score is a much more objective indicator of creditworthiness because it is calculated automatically. A bank employee cannot influence the operation of the algorithm.

Financial benefit. The use of a scoring model for assessing creditworthiness can significantly reduce the share of default. This not only increases the bank's profits, but also enables it to offer better rates to customers. The level of default directly affects the interest on loans, so conscientious payers are also interested in reducing it.

What determines the results of scoring

The final assessment when using any scoring model consists of a number of indicators. First of all, the passport details of the borrower, information about the place of residence and other contact details are checked. This is a preliminary stage at which applicants with invalid documents are screened out. Then there is the analysis of other factors.

• Personal information about the client. The scoring assessment takes into account the marital status of the borrower and the presence of minor children. The length of service at the last place of work is also taken into account.
• Applicant's solvency. One of the most significant factors influencing the scoring score. To get approval, it is important to prove that you have not only sufficient funds to repay the loan, but also regular payments. To assess the financial situation and creditworthiness in most cases (especially when issuing large loans), it is required to provide documents from the place of work: a 2-NDFL certificate or in the form of a bank. Sometimes the applicant's expenses (for the maintenance of dependents, utilities, etc.) are also taken into account.
• Credit history. When scoring the creditworthiness of customers, it is mandatory to check debts and delinquency on previously taken loans. The Bank may, if consented to, obtain data on the applicant from the credit history bureau (BKI), which reflects all the necessary information. The system also takes into account the presence or absence of regular payments for existing loans. The BCI records the history of applications made by the applicant: a big percentage waivers from other financial institutions may lower the score.

• transactional behavior. If the borrower is payroll client or has a deposit in a bank, the credit score may be increased. This takes into account the amount of savings on the account and their dynamics.

The scoring system checks all data separately and compares them with each other in order to identify possible contradictions. The reliability of the specified information is confirmed by the existence of a connection between the income and expenses of a potential borrower, place of work and address of residence, etc.

Analysis of scoring data

Based on the result, the system makes a decision:

• OK- the score is high, the application can be transferred to the next level;
• refusal- the applicant scored too low a score, so the consideration of the request is terminated;
• additional analysis required– the system does not have enough data to give an adequate assessment. In this case, the bank specialist independently studies the applicant's questionnaire and clarifies the information. To confirm the controversial aspects, the applicant may be required to provide additional documents. After a manual review of the application, a final decision is made.

How to get a high score

Avoid loan arrears. To increase the chances of a good assessment and approval of the application, you need to have a clean credit history. This means that the applicant must not be in arrears on other loans or outstanding debts. Therefore, even if there is financial difficulties It is important to keep track of your credit history. It is better to provide the bank with documentary evidence of temporary insolvency in time and develop a debt restructuring or deferral scheme. This will close the current loan and increase the likelihood of approval of a new one.

Open a bank account. In most banks, you can get additional scoring points if you have an account, so it is better to make a deposit in advance.

Indicate only real information in the application. The accuracy in filling out the questionnaire also affects the assessment. The information must be objective and truthful: doubts about the accuracy of the information may be the reason for refusing a loan.

Pay attention to the relevance of contacts in the questionnaire. In order to increase the scoring assessment of creditworthiness, it is necessary to indicate only real contact details in the questionnaire. A bank employee should be able to call all subscribers whose phones have been entered by a potential borrower. If it is not possible to contact them, the specified data may be considered unreliable. This is one of the reasons to refuse credit.

If the score was too low and the application was rejected, this may indicate that the model and algorithm specific bank not suitable for the borrower. Financial institutions often use their own systems, which take into account a different set of factors.

What to do in case of refusal

With a low score, the system usually simply rejects the application, while the client is not informed of the reasons for such a decision. Bank employees often recommend repeating the appeal after a few months. Alternatively, you can try applying to another financial institution. However, this should be done with caution: all refusals are recorded in the credit history, and if there are too many, the score is reduced. In order to find out about the presence and number of rejected applications even before contacting the bank, you can send a request to the CBI.

The scoring model does not provide objective and relevant results if the client is applying for a loan for the first time. For such cases, some banks use only manual processing of applications by specialists. At the same time, in fact, such clients are often offered less favorable conditions, increased interest rates and reduced loan amount. So the bank reduces losses from a possible default. However, if you repay the first loan on time and without delay, this will be reflected in your credit history, so the next time you can count on a higher rating.

To use the services of the NBKI for the development and/or use of scoring system methods, fill out the application form on the website.

Facebook

Twitter

In contact with

classmates

Google Plus