What is credit scoring? About types, model and method.

Before a bank grants a loan to a customer, a thorough customer check for creditworthiness must be applied.

Credit scroing method

Determining the level of customer reliability with regard to the timely loan repayment is one of the key elements of credit risk assessment. This is done on the basis of a credit history analysis and scoring, based on the customer's characteristics. Information on how the customer has repaid, and continues to repay, their dues is often provided by credit bureaus. While the customer assessment itself is based on credit scoring models.

Credit scoring is one of the methods used for estimating the risk associated with granting a loan, or rather the probability of its non-repayment. It is based on the calculation of the customer score according to data provided in the loan application or obtained from other sources. The more similar the profile of a borrower is to profiles of those repaying their loans on time, the higher the rating it will receive.

Credit scoring result is usually presented in points, and the number of the points allows for assigning the customer to appropriate risk category (e.g. reliable customers or customers who may have problems with loan repayment). Credit scoring, regardless of how it is calculated and what characteristics it takes into account, eliminates the human factor and ensures objectivity in the process, which reduces risk and speeds up the credit process.

The use of scoring models in credit processes is very popular due to its numerous benefits:

• shorter processing time of applications, which also leads to lower costs
• objective assessment of credit risk
• improved employee productivity
• possibility of using appropriate financial collateral
• monitoring the credit portfolio for bad credits
• better forecasts and credit strategies

What influences credit scoring?

Many factors are taken into account in the process of granting a loan. These include the characteristics of the borrower (who they are), their economic situation, the amount of the loan applied for, its purpose (i.e. what is to be financed by the loan) and the type of collateral. The variety of these factors means that the risk is estimated using elements of quantitative and qualitative analysis.

The quantitative analysis includes, first of all, an assessment of the financial standing of the customer based on their income and monthly expenses. It may also include cash flow analysis of the customer's accounts and credit history. While the qualitative assessment takes into account, among others, marital status, education or employment form - for natural persons, and for enterprises - legal form, industry in which they operate or the way of keeping accounts.

Equally important are past customer behaviors that adversely affects credit scoring:

• late payment of installments and other liabilities,
• exceeding credit card limits,
• large number of commitments entered into,
• no credit history of any kind.

Credit scoring model. How is it calculated?

Banks usually grant loans based on a credit scoring model that combines qualitative and quantitative analysis. Credit scoring is based on statistical methods, thanks to which it is possible to predict the probability of a certain event occurring in the future - in this case a loan default.

The scoring process uses information about the customer collected at the application stage - mainly data characterizing the customer, but also information about their past behavior. Each credit institution considers a different set of features and assigns different point values to them. For example, a highly educated person will usually score higher than a college dropout, but the exact point value and its impact on the final score may vary from bank to bank. The sum of points from particular characteristics is usually the final score. The range of the possible score is determined by each bank / institution - the most popular credit scoring in the USA (FICO) gives results between 300 and 850, while the scoring provided by the Polish credit information bureau (BIK) can reach a maximum of 100 points.

What are the types of credit scoring?

Scoring models can be classified according to different criteria. Thus, we can talk about a scoring of individuals or companies (division based on the assessed entity) or credit card, cash or mortgage scoring (depending on the type of product applied for by the client). Taking into account who created and managed the scoring model, we can talk about internal scoring (prepared by banks for their own needs) or external scoring (created and made available by specialized institutions, e.g. credit information offices).

There is a very clear dividing line between application and behavioral scoring. The first one is designed to evaluate new customers on the basis of data provided by the customer in the credit application. Behavioral scoring, on the other hand, is determined on the basis of the history of the customer's behavior concerning the service of financial products. Therefore, it is calculated for regular customers, mainly in order to resell new products or change the terms and conditions of existing products (e.g. increase the credit card limit).

For the most part, the goal of the scoring models is to determine the risk of debt default. However, more and more emphasis has been placed recently on using this method for other purposes:

• profit maximization (what credit terms should be offered to the customer to be accepted, i.e. risk-based pricing)
• increasing the effectiveness of marketing campaigns by investigating whether the customer will be interested in a given product,
• fraud scoring,
• attrition scoring,
• improving debt management by determining whether the customer will be able to repay the loan in the event of financial problems.

Regardless of the type, the scoring models enable an objective assessment of credit risk, which is a key element of the credit granting process. In order to make the credit calculation as accurate, transparent and low-risk as possible, banks increasingly automate it and use ready-made systems that allow for performing a credit assessment model in a point system. The use of such tools reduces the probability of granting doubtful loans and allows for accelerating the entire credit process while reducing the risk of human error.

Individual customer scoring

The risk assessment of natural persons is based on their personal characteristics. The most commonly used are: age, marital status, education, number of dependents, seniority, form of employment, occupation, etc. The risk assessment of natural persons is based on their personal characteristics. Financial characteristics also play a very important role, such as monthly income, possible additional income or information about expenses (repayment of other loans, living costs, bills paid, etc.). If possible, data on the financial history of the customer are also taken into account, such as the number of previous commitments made and the history of their repayment or information about possible overdrafts.

Depending on the type of credit applied for, the list of characteristics to be taken into account in the risk assessment may vary. The scoring model for a mortgage loan is usually more comprehensive than that for a credit card. When buying a house on a loan, the scoring also includes a number of information about the property itself (including its value, LTV level, etc.). It also puts more emphasis on the income data.

Enterprise Scoring (SME)

In the case of small and medium-sized enterprises, the risk of bankruptcy and insolvency is examined by assigning points. Their capital, debt and development strategy are taken into account. In the case of enterprises, the credit calculation is made on the basis of: the characteristics of the industry, the characteristics of the company, its previous financial results. Credit risk analysis is also performed taking into account the company's structure, source of financing, competition, and even the qualifications of employees, mainly those at high levels, are examined. The credit assessment model in the case of a company also takes into account the financial condition of the company, its planned projects, liquidity, financial liabilities and industry risk assessment.

In the case of the smallest companies, however, very often, in addition to taking into account the parameters characterizing the company, the owner of the examined company is also the owner. It turns out that the profile of the owner and their personal credit history is more important than the numbers describing their business. This is particularly often used in the case of the smallest entities operating on the market for a short period of time.