The financial crisis can be traced to poor risk management. Even though financial institutions are supposed to be well-versed in the skills of risk management – why did so many of their risk management policies and procedures fail and why do some succeed?
The answer is a complex one. One of the main reasons for the failures was that most financial firms were unable to apply risk management coherently across the entire organisation. The key disciplines – credit, market and operational risk management – often remained in separate silos. On the boundaries, where they should have overlapped, there were frequently gaps where major risks went undetected.
In other words, they did not have an enterprise risk management (ERM) strategy – an integrated approach that aligns strategy, processes, people, knowledge and IT so that risk is better understood and controlled throughout every part of the enterprise. To make matters worse, many firms did not make it clear who had overall responsibility for overseeing risk management activities.
So how should financial institutions strengthen their risk management? Certainly, they will need to plot their own course and take its own measures.
Findings show that there are major failings in silo-based risk assessment; accurate and complete data is essential to sound risk management; and it is important to have a firm-wide strategy and platform to measure and monitor risk, according to the Global ERM in Financial Services Survey 2008.
Reducing Customer Risk
Take the risk management strategy by one of the world’s 100 largest banks with total assets of approximately US$180 billion. The Australia and New Zealand Banking Group (ANZ) has utilised technology from SAS Data Mining Solutions as a key component of its credit scoring systems.
According to Andrew Wilson-Annan, ANZ Senior Portfolio Modeling Manager, the use of credit scoring assists in three key areas: identifying those applicants who should receive credit, determining the amount of credit they should receive, and the steps that should be taken – on an individual basis – should there be a failure in commitment.
"SAS has given us the ability to develop credit scoring models that support a fully automated assessment system," said Wilson-Annan.
Prior to the introduction of SAS and the credit scoring systems for retail lending, ANZ branch staff had the complex task of assessing each loan/credit application manually on a consistent basis. This assessment process is now largely automated and incorporates SAS Enterprise Miner to develop models, which assist the bank in assessing 20 million credit decisions annually.
“By automating the more straightforward lending decisions, assessors can spend more time with customers who have complex borrowing requirements – aligning assessment resources and credit risk.”
By basing credit approvals on specific data, such as assets, liabilities and stability, personal biases that might otherwise affect decisions are completely removed from the process.
Quite apart from protecting ANZ's exposure to risk, the credit scoring system is regarded at all levels as a critical component in the bank's customer protection mechanisms. Credit scoring models developed by SAS help the bank identify credit applicants who could be at risk of over-extending themselves.
More details on how ANZ reduces customer risk are available here.
Besides ANZ, many other financial institutions accept that they need to strengthen their risk management and take a holistic approach. It is now up to the firms to ensure that whatever steps are taken to improve risk management are in the context of an enterprise-wide approach.
How are they going about it? The article Enterprise risk management: The culture of the future for financial services provides further insight.
Sign up for Computerworld eNewsletters.