Accelerating the evolution of risk analytics

Raiffeisenlandesbank Niederösterreich-Wien AG gains speed and accuracy with SAS^® High-Performance Risk

Modern financial institutions must assess their risk exposure with speed and accuracy to respond to market volatility. And they must do so while grappling with skyrocketing data volumes, shrinking decision-making windows and ever-changing regulatory requirements. The urgency for rapid assessments across all risk types has never been greater.

To address these challenges, Austrian bank Raiffeisenlandesbank Niederösterreich-Wien AG deployed SAS High-Performance Risk. The addition of in-memory grid technology and a Hadoop server to the existing SAS platform marked a significant evolution of the bank’s risk management capability.  

Accurate results for faster, better decisions

Founded in the 19th century, Raiffeisenlandesbank is the top Raiffeisenbank in the state of Lower Austria. While the Vienna division of the Raiffeisenlandesbank serves around 270,000 customers in the Austrian capital, the independent local Raiffeisen banks serve nearly a million, forming the leading banking group in Lower Austria.

“Our main priority is getting results quickly to react to business changes,” says Herbert Radl, Head of Risk Methods and Analytics. “Banks need a stable and secure information platform with the right level of computing power to aggregate structured and unstructured data from a variety of internal and external sources.”

Krzysztof Widelka, Head of ICAAP (Internal Capital Adequacy Assessment Process) and Limit Management, elaborates on platform requirements. “We need consistent, high-quality data for different risk calculations, particularly when our supervisors want to compare our current situation with last year or correlate financial and risk information,” he says. “We also need flexibility, velocity, accuracy and the ability to guarantee transparency across the entire analytics process.”

Adding to these requirements is the need to manage the ever-growing volume of models used for financial reporting standards, such as IFRS 9.

One platform for credit and market risk assessment

Raiffeisenlandesbank now relies on SAS Risk Management for Banking in concert with SAS High-Performance Risk. Driven by the speed of in-memory grid technology, the solution provides a scalable, common computing platform for both credit and market risk assessment. And since the new platform uses the same server, data interfaces and Hadoop cluster as the previous SAS environment, implementation was completed in just six months.

“Hadoop provides the ability to store and process huge amounts of any kind of data quickly,” Widelka says, “while SAS has the power to handle iterative and advanced analytic tasks. Together, the complimentary technologies give the bank consistent and uniform data for risk analysis, reporting, stress testing and ICAAP assessment. They also provide flexibility to manage fluctuations in valuation methods and data volumes.

“We deployed SAS High-Performance Risk to accelerate simulation times and get faster and more accurate intra-day results. Because of the Hadoop cluster, we can now calculate incremental VaR and perform stress testing in 90 minutes rather than four hours. And we can execute 80,000 credit risk simulations in seven hours instead of 24. Because we can perform more simulations in less time, the results are extremely accurate.”

Radl appreciates the self-service analysis of SAS Visual Analytics. “No matter what their skills, users can easily and autonomously drill down into more granular levels of detail to explore aggregate risk exposures, asset positions and capital allocations – saving time and getting faster decisions when assessing a loan, for example.”

Artificial intelligence to redefine risk management

The benefits of SAS High-Performance Risk are already significant, but the project is evolving. Next steps include integrating other risk categories such as regulatory capital and liquidity risk, continuing the transition to SAS Visual Analytics, achieving the ability to analyze big data to develop new risk models, optimizing the infrastructure and reducing operating costs.

The bank also has plans to combat fraud with the new platform. “In the future, we look forward to implementing new analytical tools like fraud detection thanks to our ability to combine data sources, such as financial and behavioral information,” Widelka adds.

Looking ahead, Radl believes artificial intelligence will transform the risk management process. “In the past, we spent a lot of time collecting information, maintaining models and improving data quality,” he says. “Now these procedures are being automated thanks to AI and machine learning. Soon we’ll be able to fully focus on analysis and simulation, though this will require a profound mindset shift across the organization.”