Credit scorecards are created with the help of statistics. First, all past loan applications of interested consumers are collected.

Then these applications are divided into two essential categories.

The first one deals with the people who repaid their loans in due time without much hassle.

The second one deals with those of the defaulted.

It is mandatory to compare the first group with the second one to prepare an appropriate scorecard.

Credit scorecards provide an accurate measurement of the likelihood that a customer will repay the credit amount back in the allowed amount of time.

Logit or probit are estimation techniques which are statistically used to predict the probability of default of new clients based on this historical database.

The default probabilities are then compared to a “credit score.” This score will rank the potential client by their height of risk without explicitly identifying their probability of default.

It is to be noted that the procedure of credit scoring was not always fit enough and it did have drawbacks. Then newer and improved techniques were applied to maintain this method of comparing credits.

These measures are: hazard rate modeling, reduced form credit models, or logistic regression.

The essential differences from credit scoring involve both the database and the ability to calculate the financial value of a loan, given its risk from a credit perspective.

The database includes all of the available observations on both defaulted and non defaulted clients. This makes it much easier to see the effects of macro-economic factors like stock prices, auto prices, interest rates, and home values on the default rates of retail loans secured by automobiles or homes.