Coface implements a new company rating system

With SAS®, Coface provides evaluation indicators for credit insurance, company information and debt management

Recently, Coface has reinforced its product line with the @rating score, a unique evaluation tool developed with SAS technologies. The @rating score is an external credit-risk rating that evaluates worldwide insurable trade debt and grades companies in compliance with Basel II reform.

From local scoring to a worldwide rating system
To improve the quality of its financial products, Coface has been working with SAS since the mid-1990s to perfect the administration and operation of its risk evaluation systems. SAS solutions were first used to improve the company's methods of scoring small- to medium-sized businesses. With SAS, Coface improved this process by increase calculation efficiency, reduce production times, minimize costs and offer customers an immediate view of short-term risks for their entire portfolio of suppliers.

After completing this project, Coface pressed forward and turned to SAS to evaluate all French companies. Nicolas Lemettre, director of Coface's Rating Division, describes the project's progress: "Having already developed an expert system for integrating scores generated by outside agencies, we wanted to internalize the rating activities and production processes. For greater efficiency, we also needed to classify this population of companies into multiple segments. We chose SAS solutions so that we could produce our models more quickly and monitor their performance," continues Lemettre.

Coface began the project in 2000 by preparing a reliable statistical base, followed by a study phase to evaluate the system. Two years later, the company implemented a complete SAS architecture to support the international @rating system.

"Our system's strength lies in the interaction between the data study and data production services," says Lemettre. "This structure allows us to make very precise and rapidly updated predictions, which is vital for our customers. To calculate a valid score and make reliable predictions, for example, the age of the data must be considered. Therefore, it is important for us to be able to archive our data correctly."

Today, the Coface information system manages 44 million companies around the world via a shared online database. As the first insurable company rating system with worldwide coverage, it includes a commercial credit recommendation and the available @rating score, which is based on political factors, devaluation risks, debt levels and more. The result is a unique world-class scale for evaluating company credit risk.

Basel II: a strategic challenge for Coface
Now that Coface is managing its production processes better, its new challenge is to respond to the limitations imposed by the Basel II Accord. To comply with this reform, banks must improve their tracking and risk assessment practices, as well as financial communications, so that they can adapt their equity capital requirements and align them with actual risk levels. This means, for example, that banks must associate a client's or partner's risk assessment with each loan commitment, especially when that partner is a medium-sized or large company.

The @rating score has proven particularly useful for banks since it is a credit-risk measurement that meets Basel II requirements. This rating, which indicates the potential risk involved if a company were to default on its loans, can be used as an external rating or as input for an internal rating.

"With the SAS solutions, we are two years early in responding to the national controller's requirement for implementing performance measurements that go back at least three years," says Lemettre. "With SAS, we are now capable of running simulations over several years that integrate every modification of the algorithm."

Impressive risk measurement
According to Lemettre, the @rating score is impressive on many levels. Its precise assessment of a company's likelihood to default can be used for arbitration aid, better risk monitoring, partially automated decision making and risk model calculation.

Using the score will also generate growth within a company's information division, since Coface is known as a reliable data supplier through its information subsidiaries and partners. Additionally, the rating can be compared between one country and another as an objective measure for evaluating intercompany credit. Many companies consult the Coface rating, for example, to confirm that a buyer is solvent.

Automating the process
Currently, the Coface rating system manages approximately 4 million customer tracking inquiries per year. "We integrate a total of 7 million new pieces of information per year with SAS, requiring verification of the ratings for French companies," says Lemettre. "We update every company's rating at least once a year, automatically or with the aid of an expert."

To respond to demand better, Coface uses SAS solutions to automate its scoring process. "We quickly built data marts that can access SAS and then made our models," says Lemettre. "The SAS solutions allow us to develop, update and back-test scores on a country-by-country basis, as well as conduct multiple studies, such as the impact of economic cycles on ratings. SAS gives us the methods, and we use known statistical techniques such as sampling methods, data processing, modeling, and validation. In this way, we can more easily respond to the Basel reform's transparency requirement."

Confirmed results
With SAS, Coface has developed an operational system on time, internalized its risk modeling work, developed better interactions between consultants and risk modeling teams, and finally, sped up the sample programming phases.

"Overall, SAS helped us improve our productivity," says Lemettre. "The modeling phase was completed in just a few months instead of two years, so the technique we used was obviously the best. We are convinced of the value of a good partnership with SAS, and today, we are using its solutions to develop new activities like analytical marketing."

Copyright © SAS Institute Inc. All Rights Reserved.