Tag Archives: скоринг
Nowadays, an expansive growth of past due debts in retail crediting is one of the most important problems of financial-credit system of Russia as well as of other countries. Collection activity may be optimized by implementing collection scoring systems. The paper regards one of the main problems of collection scoring – evaluating of collection financial flows and collection probability. The solution with the use of Tobit Type II model is offered, which allows determining the probability that debtor will make payments, dividing debtors to the groups of payers and defaulters and estimating the probable sum of payment for each debtor. Factors that have the most influence on the probability of payment (surety amount, loan duration, residence region etc.) are detected. As a classification result, the category of debtors who will not return the debt is defined. For other debtors on the basis of linear model the probable monthly sum of payment is calculated. The results of this research may serve as a foundation for planning work with the debtor, will allow identifying hopeless debts and in accordance with that effectively allocating collection resources and reducing transaction costs.