Data and analytics-driven approach for smart credit risk decisioning
Improved data governance and personalized reporting
Building a sound credit risk culture
Company achieved this using • SAS® Visual Analytics
SAS Visual Analytics helps ČSOB bank improve credit risk management and regulatory compliance
Československá obchodní banka, a. s. (CSOB) provides its services to all client segments, i.e., individuals, small and medium-sized enterprises, and corporate and institutional clients. CSOB offers its customers a wide range of banking products and services, including products and services of other ČSOB Group companies.
The ČSOB Department of Credit Risk Management serves as a second line of defense in managing the credit risk of all Group entities. This involves advising top management on credit risk-related topics, challenging local credit risk partners, ensuring compliance with regulatory requirements, and building a sound credit risk culture. The department also aims to capture early warning signals based on detailed portfolio monitoring, carrying out portfolio stress tests, oversight of the expected credit loss calculation, etc.
The Department of Rainbow and Data Quality Management is tasked with collecting, organizing, and maintaining accurate and relevant financial and customer data. This involves creating efficient data storage systems and ensuring data quality and integrity. The department also works on data analysis to extract valuable insights that can inform strategic decisions and improve operational efficiency. Data management helps the bank meet regulatory requirements, enhance internal customer experiences, and make informed decisions based on reliable information.
Both departments are working together to create a solid data foundation and high-quality analysis for management and board decision-making. To improve credit risk management processes, a new data architecture was sought that would be much faster, more flexible, and accessible to a wider range of users.
Credit risk reporting data relied on a single table
In the old solution, most of the underlying data for credit risk reporting and analytics were based on one underlying table containing a complicated logical structure with a lot of necessary data transformations. Relying on a single inflexible base data structure posed significant challenges. This setup became a roadblock when changes were needed, as it lacked adaptability. Moreover, the complexity of computations was difficult to understand, leading to time-consuming data reconciliations, and the overall structure posed a significant challenge for using advanced analytics. All these reasons led to an unnecessarily slow process and hindered decision-making.
“Before we started using SAS, we were experiencing some issues in the area of data governance and system flexibility. It was like a black box to us. The whole process required many time-consuming updates and data checks, and we were not able to do time-series analytics efficiently.” Jaroslav Bělka Executive Data Director Československá obchodní banka, a. s.
Faster data processing, better quality, and governance
In the old process, calculations were delayed by an inefficient base table that wasn't optimized for computational tasks. Thanks to the SAS solution, this bottleneck was addressed by transitioning to a relational database system. By leveraging the relational structure, the new system significantly accelerated the entire process, thanks to the improved design and indexing capabilities of the relational database. This transformation not only enhanced performance but also streamlined operations and paved the way for more sophisticated data analysis. The interconnectivity of SAS Visual Analytics and Python provided additional capabilities for the implementation of advanced data quality checks based on machine learning techniques as well as automatic notifications about portfolio development and personalized reporting.
Improved analytics and retrieval of information
The switch to a relational database brought several advantages. Firstly, it made calculations much faster and more efficient, improving the overall speed of the process. Complex calculations that used to take a considerable amount of time are now being completed swiftly and accurately. Additionally, the new database structure allows more efficient information retrieval, which led to the creation of a detailed interactive reporting structure providing easy and user-friendly access to the data even to non-technical users. Automatic notifications, together with personalized reporting, translate the data into distilled information, which helps to discover potential risk signals and leads to data-driven decisions and better oversight.
ČSOB Group Facts and Figures*
4 mio. +
customers
8000+
employees
200+
branches
Personalized reporting and data democratization
Rainbow and Data Quality Management department, together with the Credit Risk Management department, implemented a new data architecture that is much more flexible and faster, enhances data quality by design, and is tailored to the analytical needs of credit risk management. On top of this new data architecture, comprehensive interactive reporting and automatic notifications about portfolio development were built. The interconnectivity between SAS Visual Analytics and Python provided a base for the usage of time-series analytics and machine-learning techniques to enhance the identification of relevant changes in the ČSOB Group credit portfolio. Furthermore, risk-related data are now accessible to a much wider user base within the bank.
“SAS solution saves time and resources for the entire team, which no longer needs to spend time on manual data processing. It provides better oversight and improved accessibility for non-technical users. Now it’s easier for our business to succeed and our organization to become fully data-driven.” Jaroslav Bělka Executive Data Director Československá obchodní banka, a. s.
