Risk score

The Ocrolus Risk Score offers a highly analytical approach to predict small business borrowers' probability of default using cash flow analytics. By training a model to learn the relationship between 500+ Ocrolus cash flow features and business loan performance data, the Ocrolus Risk Score transforms how lenders assess credit risk using transactional bank data.

The Risk Score uses a Gradient Boosting Model trained on tens of thousands of matured business loans to reveal risk patterns not found in traditional data sources. Customized risk scores can also be created based on your unique target values and customer data.

Key benefits

• More informed decision making: The Risk Score model has been trained to identify unique patterns in granular transaction data, helping lenders rank risk more effectively than possible with traditional data alone.
• Streamlined credit assessment: Quickly assign borrowers to different risk categories to simplify your underwriting process and provide faster lending decisions.
• Predictive insights to mitigate risk: Anticipate potential risks and take proactive measures to mitigate them using historical data and relevant indicators.

Custom Risk Score

Work with Ocrolus to fine-tune the Ocrolus Risk Score to your lending portfolio, tailoring it to predict the best risk from cash flow data within your unique customer segment and product structure. Custom scores are deployed through the Ocrolus API and can be actively monitored and refreshed periodically as new data becomes available.

To start evaluating the Risk Score, please reach out to your Account Manager or [email protected] today. To learn more about the Risk Score, see the API documentation of Risk Score.

📘

Generic model default definition

It is a charge-off with at least 10% of loan principal unpaid. You can choose alternate dependent variable definitions in a custom risk scoring exercise.