What is credit scoring

The word means account. In fact the scoring – is the common name of mathematical models, by means of which is determined by the probability of defaults on issuing loans.

The general method of scoring models (hence the name of them) is built on the principle of counting: on the basis of the applicant’s credit (profile – the main but not the only source of information), each type of data is assigned a score. For example, the borrower’s age from 18 to 23 years – X points, from 23 to 35 – Y points, more than 35 – Z scores. Further, all points scored by the borrower are added and the final result is obtained (or score). And then runs a simple rule: if it is greater than a predetermined value – the loan is granted, otherwise – no.

This example is very conditional. Each bank – its recognition of the risk model, which uses a variety of indicators evaluated different ranges are assigned to various points. The decision to grant or refuse the grant funds may also be taken not simple linear condition.

Recently, not only banks build their model and calculate scores – these have become engaged and the credit bureaus, and even mobile operators. BCI is calculated likelihood of problems with loans based on the data of millions of contracts. Cellular operators – on the basis of billing millions of their subscribers: how the subscriber spends on communication, how often shall make payments, what amounts enjoys any deferred payment, how often, what discipline extinction. This data is also used by banks in their scoring models as additional parameters to assess the likelihood of non-payment on the future of the borrower.

What information is used

First of all, it’s the information you provide about yourself in the application for the loan. Each indicator – a formula for the input parameter. Date of birth – it is considered to be of your age. Date of registration of residence – duration in days of permanent residence in one place. The industry in which your company operates, the employer – for the evaluation of the likelihood “of the objective of non-payment on the loan” – for example, risks that have decreased due to the recession in a particular industry. Agree – is the difference between public sector employees (civil servants) and employees engaged in the import of fruit and vegetables from the European Union.

Moreover, these profiles are used for internal verification – assess the reliability of the information provided by you. Industry and your position in the questionnaire there a reason not to satisfy the curiosity of bank employees. On the basis of questionnaires Bank calculated the average income of workers of different sectors of posts. And then compares the income that you declare in the questionnaire, with their calculated values. Indicate much more – the formula will increase your risk of credit default.

Variety of models

Algaritmy applied depend on the amount of data that the bank can get at the moment of receiving the application.

Express-scoring used in “fast” lending stores for the purchase of certain goods, when issuing consumer loans in cash and plastic to a minimum of documents (for example, only the second passport and identity document). In assessing the probability of default of the bank is forced to rely only on questionnaire data, the message “word of honor.” It is clear that the assessment will be more “tough” – to evaluate the accuracy of the information difficult.

Another thing, when you provide more and proof of your income, employment and financial condition – meaning an income statement, a copy of the work book, the documents on the property. It is included in the work of a more complex model, which assesses the “softer”, as banks have more confidence in the documents with the signature blue and blue print than “dry” mathematical formula.

Why You Need

In fact, this is a very convenient way to solve many problems at once Bank

1) To compare two borrowers with each other. How to determine which of the borrowers in the future more likely will be problematic? some data in the questionnaire – a few sheets? The calculation of risk leads all, they say, to a common denominator. Who he is more – that and refuse.

2) Rapid assessment. All models are automated and integrated into the information systems of banks. It is this decision allowed the development of rapid lending stores. Task Manager – the right to make the data of your application (well, with your words bring) to the computer. The decision essentially adopts the robot – the computer.

3) eliminates the possibility of cheating employees of banks – excludes “human factor” in the decision “to give the money anyway.”

4) Flexible management extradition. Depending on the liquidity situation in the banking and financial plan of the bank is very convenient to manage extradition. Need more to issue loans – the bank reduces the borderline value for the cut-off may potentially troubled borrowers. It is necessary to reduce the issuance – the bank “bullies” in the model of “good” scoring points – failures in the credit go one by one.

5) The rate of risk-based. Many banks in advertising do not declare a particular interest rate, and are limited to phrases like “the rate of …”. It’s not just because they want to attract more customers advertising. Just in their “formula” calculation of the rate laid down for the client based on his score. The greater the likelihood of the risk – the higher the rate. And vice versa. Agree – very reasonable approach – if I am a reliable borrower – so me and the rate should be lower.

What you need to know the ordinary borrower on this procedure.

Any scoring model – a complex mathematical apparatus, over which employs experts in higher mathematics. So guess what data point in the questionnaire that “certainly gave the loan” – will not succeed. Moreover, any false and contradictory information are “calculated” the formula itself.

Some models take into account such factors as “the speed with which you fill fields of the form online in the online questionnaire” and you enter data manually or “copy-paste something.”

So do not try to trick the computer!

Second, the model is constantly being refined and through the expansion of statistical sampling – the more profiles and granted loans from the bank – the more accurate the model parameters can be estimated.

In the long run, the borrower does not need to know – there is a bank scoring model or not, and what’s she like. You possess this knowledge or not – loan or you approve or not.

  • Mortgage or consumer credit
  • How to Get a Better Auto Loan if You Have Bad Credit + video
  • Types of letters of credit and payment
  • Credits for the purchase of equipment
  • Commercial Mortgages credits purchase of real estate business
  • What is urgently needed if a large amount of the credit
  • Obtaining consumer credit

39 queries in 1.375 seconds.