We offer Scoring and Integrated Credit Risk Workflow to Our Partners

Products Offered

Short term loan with rates from 0.05% to 0.8%

Sound credit risk management is crucial to a lenders’ stability, growth and future profitability. With so many variables to balance, managing the credit risk can be a challenge. gkCredit’s Analytics & Scoring services help our partners optimize their credit risk analysis with high quality data, systems and strategies to help minimize the risks and improve recoveries.

gkCredit’s Credit Analytics services focus on the pillar of application processing, and help our partners automate and manage their processes from end to end. Credit scoring is a statistical analysis to identify a person’s or businesses’ creditworthiness. Companies use credit scoring, among other things, to decide on whether to extend or deny credit. Our solution offers a dynamic credit scoring module and an integrated credit risk workflow which will maximize the customer value managing risk and helping you avoid potential bad debt. gkCredit’s strength comes from its strong knowledge-base in specific segments and market combinations. We first bring model and scoring approaches that are tested and validated for similar markets and segments instead of building from the scratch. Then, we setup a self-learning system which constantly adapts for better credit returns and lower NPLs.

Our Approach

We apply a 4 step approach to setting up a credit analytics solution for a new partner:

Gathering of data

Deploying an initial

Performance observation

Continuous model
building & improvement

Initial scorecard reducing break-even to as short as 7 months

At step 1, we gather all the data points available at gkCredit’s credit application system. Depending on the dynamics of the market, new data points can be acquired through integrations with local credit bureaus or data providers. An example to this is the transactional bank account data which has become available and accepted in the European markets with the PSD2 regulations.

Secondly, we deploy a basic model from our past learnings in similar market. The model is kept plain and simple with very high approval rules, yet eliminating a small group up high-risk customers. The intention is to collect data for a small period of time while effectively managing cost and recovery performance for the lender. We have helped one of our partners, -myKredit- achieve profitability and break-even in less than 7 months, which could not have been achieved in the absence of an initial model.

After we observe the performance of the initial model, we setup a self-learning machine learning algorithm that constantly updates its rules for better predicting which customers to approve or deny. The model not only changes the importance of its variables, but also adapts to the number of cases observed by adding and removing the predictors to an optimum balance. Currently, gkCredit’s self-learning scorecards are running in 6 markets with no manual intervention, and constantly improve the recovery rates from customer’s first, second and third loans.