HomePortfolioBuilding a System for Scoring Creditworthiness

Building a System for Scoring Creditworthiness

Finance
AI / ML

A Belarusian commercial bank asked Altoros to create a software system for scoring creditworthiness of the bank’s potential customers.

Building a System for Scoring Creditworthiness

The customer

The customer is one of commercial banks operating in Belarus. The bank’s main activity is providing small loans (up to $100,000) to small- and medium-size businesses as well as consumer loans to natural persons (up to $10,000).

The need

The main task was to develop methods and algorithms that could be used to classify borrowers according to their creditworthiness. We were to create a software system for scoring creditworthiness of the bank’s potential customers.

The solution

For our research we developed two creditworthiness scoring software systems:

  • CS JurPers is a system that classifies legal entities based on their balance sheet ratios.
  • CS NatPers is a system for scoring creditworthiness of natural persons based on the data in their personal profiles.

“Credit Scoring of Juridical Persons” (CS JurPers) is a system designed to classify potential borrowers from the bank (legal entities) according to the level of their creditworthiness. It implements the main mathematical algorithms designed for scoring creditworthiness of legal entities (the mechanism of linear discriminant analysis of Gaussian random vectors and the algorithm based on the logit model of binary selection). The system also employs the expert methodology developed by the Ministry of Finance of the Republic of Belarus to evaluate creditworthiness of borrowers.

To classify new borrowers, the user can employ the following:

  • the expert methodology of the Ministry of Finance of the Republic of Belarus
  • the algorithms based on linear discriminant analysis (the Fisher model)
  • logit models of binary selection with pre-evaluated ratios

The outcome

CS JurPers is currently used by the above mentioned Belarusian commercial bank to evaluate creditworthiness of borrowers. The implementation of the system reduced the time required to make decisions from 3-5 working days to 1 hour. Thanks to this system, the bank was able put new credit products on the market, minimize credit default risks, and, of course, increase the turnovers and returns on their loans.

/

Want to develop something similar?

Preloader
Yauheni Starchak

Yauheni Starchak

Artificial Intelligence Practice Head

y.starchak@altoros.com650 265-2266

4900 Hopyard Rd. Suite 100 Pleasanton, CA 94588