Credit Scoring and Its Applications

From MaRDI portal





Credit scoring is any procedure to assign probabilities of default to loans to individual customers of a credit institution. It is understood that the probability of default depends on a (relatively large) number of variables that determine the ability of the counterpart to pay back. These are variables like amount of loan, maturity of the loan, salary, other liabilities, marital status, number of years as a customer, etc. This credit quality depends exclusively on the selected variables. Of course, the selection of these variables is based on a knowledge expert, with the help of certain statistical issue. Let us call this vector of variables the financial situation of the loan. The scoring procedure really assigns a probability of default conditional to the financial situation. This being the case, given the values of this vector of variables for a given new loan, the scoring procedure looks in a historical data base of loans with similar financial vectors and deduces this probability of default based on the history of default of these nearby vectors. This book ``Credit Scoring and Its Applications, gives surveys comprehensibly the main and most used scoring procedures, from simple discriminant analysis to the more sophisticated neural networks, and describes practical issues ranging from how to imput economic cycles to how to measure scorecard performance. All in all this is an excellent introduction (and more) to credit scoring. This is a useful book for practitioners, and quite timely, since the forthcoming banking regulation (the Basel Accords) are forcing banks and credit institutions anywhere to refine and improve their present scoring procedures.




Cited in
(92)








This page was built for publication: Credit Scoring and Its Applications

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3151804)