Credit score

A credit score (of English to score points, score -. Score ) is a numerical value based on a statistical analysis, which represents the credit worthiness of a person. With credit scoring companies try the creditworthiness of the customer or partner companies according to a predetermined method to determine more or less automated.

On the basis of borrower characteristics, such as " customer since ", "residence", " professional ", " security " points are awarded, these weighted and then combined into a single credit score to help with this total score lending. Is the creditworthiness sufficient credit may be granted. However, scores can not be used only for credit decision itself, but also for setting interest rates and credit lines.

Motivation is to avoid risks and to obtain a statistically based method relined objectified decisions. The better the underlying scoring model reflect reality, the less credit losses, there will be. Scoring models and there influent characteristics must be constantly maintained.

The specific rules and algorithms of a scoring and weighting are called " score card ", after the homonymous term from the sport. There are several techniques in order to develop appropriate scorecards, such as logistic regression, discriminant analysis, artificial neural networks and other data mining methods.

  • 4.1 advantages
  • 4.2 disadvantages

Internal and external Scores

Credit scores can on our own data of a company (such as employee master data, credit application data) based or external data, consider some of bureaus.

Internally Detected credit scores are not necessarily consistent with external ratings, so that there are may different default probabilities. This can have many reasons:

  • Different inputs
  • Various information aggregation method
  • Different rating scales
  • Create banks internally "point in time" ratings, that is, a default prediction for one year after assessment time point, while external credit ratings on a "through the cycle" approach, ie a default prediction over the business cycle based.

If a bank both the internal value and the external rating is available, it may either:

  • Believe your own system
  • Believe the foreign system
  • Both systems
  • Do not use system

Credit scoring in retail banking

Credit scoring is used as a statistical method of lending institutions in order to perform a risk classification for private standardized installment loans and small loans. Such loans are usually unsecured and are awarded based solely on the personal credit of the borrower or.

The aim is a fast credit decision in the processing of personal loans, the detailed financial situation of the borrower can be resolved only to a small degree.

There are used personal characteristics (such as occupation, employer, marital status, account management in-house, positive and negative features of the credit reports ) and economic conditions ( disposable income and financial circumstances, expected spending ). With own customers can be to draw on experiences in their customer relationship; the credit decision is made after a personal interview, although done on a subjective, intuitive judgment, but also provides a holistic impression of the traditional form of the loan officer.

The detected features are by a score standardized ( credit scoring ). Valuation rules that classify the data to be collected and assign a point value ( the score ) can be deposited in various methods. In addition to stand-alone applications spreadsheet applications or paper-based process descriptions are common.

Admissibility pursuant to § 28b BDSG

On 1 April 2010 an amendment to the Federal Data Protection Act ( BDSG) came into force, according to (28b § BDSG) scoring for the purpose of deciding on the establishment, implementation or termination of a contract with the data subject is permitted if the data used by a scientific recognized mathematical and statistical methods for the decision are proven significantly, in the case of bureaus, a transmission of the data used would be allowed for the calculation not only address data will be used and been informed of the use of such data in the case of the use of address data of the person concerned before is.

Information according to § 34 BDSG

Since April 1, 2010 are companies that use the scoring, according to § 34 paragraph 2 BDSG obliged to inform the affected information on the determined in the last 6 months probability values ​​on an individual basis, the data types used in the calculation as well as the existence and importance of the probability values ​​and traceable to give in general terms.

Credit bureaus are required by § 34 paragraph 4, to provide corresponding information about the probability values ​​received in the last 12 months to third parties.

Credit scoring Schufa

SCHUFA offers its customers since 1997 together with the credit information on individual consumers a score value based on the data it stores to. This is a value from 1 to 1000, which is assigned to the respective consumer and indicates the likelihood of a credit failure. The lower the value, the greater the probability of default. The score value is dependent on the purpose for which it is requested - so get as insurance other score values ​​as mobile carrier. Each consumer may prohibit at the Schufa the score submission to his person. Whether this proposal has a negative impact on future credit decision is open. Since the beginning of 2007 can be seen in % values ​​in the credit report (online) the own base score value.

SCHUFA industry scores were revised in 2001. The score method is based on the logistic regression model, which models the probability of a random event with two possible outputs. Approximately 6.7 million anonymized data sets were evaluated on a " maturation period " of 15 months for the process in 2001. From the industry score, there are 7 different types. These are: mortgage bank, mail order, retail, telecommunications, cooperative and savings banks, banks and the credit bureau Business Line.

Since April 1, 2010, consumers can according to § 34 BDSG, Section 4 some information about the historical probability values ​​- that is SCHUFA scores that were to grant information within the past 12 months SCHUFA contractors - receive.

Among the features that takes into account the SCHUFA include actual, so for example, entries on outstanding loans and validly become reminders or enforcement orders, but also turn a statistical scoring.

SCHUFA keeps accurate calculation formula of their scoring system under lock and has to date with all requests to disclose these resists.

Until 2001, obtaining a credit report has been included as a negative feature in the scoring; after massive protests turned Schufa this practice, by its own account.

From the lender's so inquired and automatically determined, a risk classification is determined and prepared the credit decision.

Credit rating in the corporate business

In the corporate sector, the economic data will be analyzed in more depth; with an analysis of the financial statements for information processing and evaluation in the foreground. Tendency statements are made as well as qualitative, forward-looking factors taken into account ( for example, the potential of human capital ). Then a rating classification is made. As a result, the banks took account of risk factors are similar to those of the major rating agencies. They take into account the financial situation, market position and management quality. A long-standing relationship with the borrower ( bank relationship) may give banks an information advantage over credit rating agencies, which have only external information.

Pros and Cons

The credit scoring model has over conventional methods advantages and disadvantages:

Benefits

  • Standardization, personal preferences of the agent's credit be turned off
  • Empirically validated ( objectively comprehensible )
  • IT technical refinement possible
  • For the lender (which does not necessarily have to be banks, since the method is also applicable to the goods financing ), the credit decision process more efficient through automation.
  • Acceleration of the credit decision
  • Time and cost savings

Disadvantages

  • The personal experience of the loan officer does not flow. A long-standing business relationship with the borrower often provides an information advantage dar. The loan officer decides on its holistic impression of the borrower. However, some technical scoring solutions take into account such data.
  • The data may be problematic ( Privacy Policy )
  • The decision is possibly based on outdated or erroneous data taken ( data quality)
  • Disclosure or trading data are possible
  • Query without customer consent
  • Insufficient consideration of qualitative personal data
  • Constant updating required

Costs

Two types of costs are to be observed in connection with the credit check:

  • Costs from the type 1 error: lending to customers poor credit ratings that result in a loan default.
  • Costs from the type 2 error: no lending to borrowers with good credit ratings, loss of interest income
Discriminant function analysis Data-Mining
488200
de