Head of Strategy and Controls
Faster and more consistent credit insurance assessments
Requests for private loans are on the rise, increasing the workload of insurance underwriters. This causes avoidable stress and creates more room for human errors. Automating the underwriting process can ease the workload; the problem is determining how to ensure that each request is still processed correctly and evaluated thoroughly. The Belgian credit insurer Atradius and SAS worked out a solution and managed to automate 60 percent of the underwriting decision-making process. This efficiency gain enabled underwriters to refocus on more complex cases and improve Atradius' business decisions.
We appreciate the autonomy the SAS tool provides. It has enabled us to automate 60% of our underwriting decisions and improve our business decisions.
Automated credit scoring improves more than just underwriting efficiency
The Atradius Instalment Credit Protection (ICP) business unit primarily covers financial institutions against the risks of defaults involved in multiple instalment agreements with private individuals and B-to-C companies in Belgium and Luxembourg. Atradius ICP processes around 15,000 credit insurance requests per month with a turnover of approximately 30 million euros.
Relieving the pressure on underwriters
Atradius used to dispatch all the requests to underwriters, who processed each of them and communicated their decision to the client within fifteen minutes. However the number of credit requests from private individuals has grown dramatically in recent years. As a result, the workload increased, becoming a major source of stress for Atradius underwriters and increasing the risk of human errors.
"Back in 2007, we felt that hiring more underwriters might not be the most sensible choice," recalls John Decleyn, Head of Strategy and Controls at Atradius ICP. "We feared that this market trend could change precipitously—which is in fact precisely what happened during the recent worldwide financial crisis. So we turned to SAS to help us streamline our underwriting process."
"It was crucial to us that the functions of the new tool faithfully reflected our credit decision criteria, and not the other way round," says John Decleyn. "For example, it was important that the number of requests accepted and refused after the implementation of the SAS tool remained consistent with the figures recorded prior to implementation. We therefore spent several months with the statistical and underwriting departments developing a detailed scorecard. After the initial round of tests, the development team introduced filters into the SAS solution in order to precisely reflect our underwriting decision process."
Efficient combination of automation and human expertise
Financial institutions submit their credit insurance requests mainly online via a host-to-host network. Today, instead of Atradius dispatching each new case to underwriters, the data recorded on screen are fed automatically to the SAS tool. In a matter of seconds, the tool simultaneously processes large quantities of data from three databases: the case data fed directly to the credit scoring system, the Atradius internal client database, and databases from the Belgium National Bank.
Requests from private individuals represent around 75 percent of Atradius' volume and the SAS-based system automatically accepts or refuses 60 percent of them. This high level of automation assures consistency in the decision process, relieves stress on underwriters, and frees up time that they can spend on addressing more difficult cases. The remaining 40 percent are typically more complex requests that require additional information. In these cases, the SAS solution suggests a credit scoring and sends the files to the underwriter, who then handles them manually.
Customized reporting fine-tunes risk management
The SAS solution not only improves the efficiency of the underwriting process, it also delivers valuable information for strategic and tactical business decisions. For instance, the tool enables Atradius to develop their own monthly reports based on specific credit data and market data. Based on these reports, Atradius can fine-tune risk management in accordance with the type of insurance policy, the type of financial intermediary, or market conditions.
The system's analytical functions also have a significant commercial impact. "The tool provides us factual data that we can use to support our pricing decisions and commercial negotiations," declares John Decleyn.
Atradius involved the underwriters from the very beginning. They helped define credit scoring variables, participated in tests, and played an active role in shaping the final credit scorecard. "Although the SAS tool runs in the background, working with a scorecard was a radical change for underwriters. Involving the main users in every step of the automation process also made it their project, which increased their willingness to adopt the tool," observes John Decleyn. It also helped reduce training time at the end of the implementation project.
Developing new value-added functions
"We plan to increase the ratio of automated underwriting decisions above the current level of 60 percent, as well as the quality of the decisions," states John Decleyn. In addition to the current customized credit scoring model based on the evaluation of default probability, Atradius plans to develop an additional model that will anticipate the expected percentage that can be recovered in the event of default. The internal Atradius project management team will lead its development.
"We appreciate the autonomy the SAS tool provides. Combining both models (Probability of Default and Loss Given Default) in a matrix will bring valuable improvements to the processing quality of credit insurance requests."
Streamline the underwriting decision process.
SAS® Credit Scoring
- Improving business process efficiency. 60% of credit insurance requests are now automatically processed, enabling underwriters to focus on more complex cases.
- Fine-tuning risk-management. Customized reporting provides data that improve credit risk analysis.
- Improving commercial policies. Factual data support pricing decisions and commercial negotiations.