Case Study: DKB
(Aryza Evaluate)

Situation

A central challenge of the project was to establish a methodological framework capable of:

• automating parameter-based calculations for large portfolios comprising more than one million accounts;
• incorporating expert assessments for individual large exposures.

At the same time, a highly performant data infrastructure had to be built to ensure data quality at reporting dates.

Calculations needed to be completed within tight financial close timelines. In addition to methodological accuracy, performance, traceability, and reproducibility of results were critical, as these had to be demonstrated to both external auditors and internal audit functions.

Objective

The objective was to implement IFRS 9-compliant risk provisioning and ensure its stable integration into monthly and annual financial statements. The solution was required to be deployable in the short term and operable without in-house development. Long-term operation needed to be sustainable with a reasonable level of staffing.

Implementation

“The initial implementation process was characterized by a strong hands-on approach,” says Felix Fliegel, Head of Credit Consult at DKB. Required areas of action were identified and addressed promptly.

During the project, the necessary data interfaces to the bank’s core systems were established, and calculation runs were integrated into the monthly and annual closing processes.

In a subsequent project phase, functionalities for calculating migration losses across all rating classes were implemented in a separate application, enabling compliance with requirements under the Capital Requirements Directive (CRD).

Today, Aryza Evaluate is used both in risk controlling and in the workout unit responsible for managing distressed exposures.

“The implementation process has since reached a high level of efficiency and professionalism,” Fliegel adds. Business configurations can be adjusted directly within the application without requiring changes to program code. Updates and new versions can be tested in a structured manner and subsequently transferred into live operations.

Operational Use

In day-to-day operations, DKB primarily uses three functional components of Aryza Evaluate:

• Parameter Engine: periodic calculation of collateral values, as well as custom models for determining risk parameters such as PD, LGD, and CCF;
• Good Book: automated calculation of risk provisions for large performing loan portfolios based on risk parameters (PD, LGD, CCF);
• Bad Book Processing: structured individual case handling of notable or defaulted large exposures within the same system.

This enables both automated valuation and manual case-by-case assessment within a unified working environment.

Outcome

The introduction of Aryza Evaluate has reduced DKB’s manual reconciliation effort in ongoing operations. Valuations are transparent and can be consistently explained to both internal audit and external auditors.

Adjustments can be coordinated and implemented with flexibility. Risk provisioning is now embedded as a permanent component of the regular financial closing process, rather than being managed as a project-based activity.

From DKB’s perspective, the stability of operations is reflected in:

• structured bug reporting;
• predictable maintenance and support costs;
• reliable delivery timelines for risk provisioning calculations;
• consistent test results during version changes.

Customer Feedback

“Our collaboration with Aryza has consistently been constructive and results-oriented. Particularly noteworthy is the professional and collaborative communication, which has made a significant contribution to our joint success. We value the high level of commitment as well as the ability to respond flexibly and in a solution-oriented manner to our specific requirements. Overall, we look back on the partnership with a high degree of satisfaction and look forward to continuing our collaboration.”

Felix Fliegel, Head of Kredit Consult  DKB