FINANWARE™ Credit Risk IRBa was designed and built to provide a support framework for the implementation of the Basel Committee recommendations with regard to the appropriate management and handling of credit risks, via Internal Ratings Based approach (IRBa). The design flexibility of the solution makes it possible to connect with the definitions and specificities established by the regulatory agencies of each country and with the different forms of self-regulation.

The management support provided by FINANWARE™, enables the building of bases for the measurement and technical analysis of credit risks and comprises flexible and parametric models which, upon adjustment to the institution's defined methodology, expedite a static and dynamic analysis of the main credit risk indicators both for compliance with the applicable legal rules and for the determination and management of non-compliance probabilities, the exposure at the time of non-compliance, the expected losses due to non-compliance and the expected recovery rates. FINANWARE™'s basic models were developed under standardized financial engineering principles and rules tested and successfully applied on the market, with facilities to define and weigh the variables to be used for the generation of performance and analysis and results projection models.

With the support of the statistical tools, FINANWARE™ Credit Risk IRBa sets up an environment for the flexible generation of several analysis scenarios that facilitate the simulation and viewing of the impact and results of the assumptions applied to the value of assets and/or profits, and for determining the amount of provisions and capital required to cover losses. It also facilitates the analysis of the best institutional strategies to reduce credit risks or mitigate its effects. The model's structure lends itself to a dynamic analysis of information using the following dimensions: products, offices, areas, interest rates, currencies, time periods amounts, type of customer, customer segment, economic sector, financial use of credit, credit status, credit rating, provisions and types of guarantees.

The models enable a flexible definition of the variables to be considered in the quantitative and qualitative criteria for scoring and rating prior to the extension of loans and also for follow-up and control.

Traditionally, the rating and calculation of provisions were performed by transactions systems through the automation of the operating processes. However, the need to implement new internal rating methodologies and models, with the ability to evaluate and predict expected losses in respect of loan contracts, required that this function be managed by systems specializing in information gathering and processing with this approach.

Performing the rating process in transactions systems has certain difficulties having to do with the availability of the information required, either because it is dispersed in more than one portfolio management system or because it is not integrated with the customer's information or with information in other credit risk sources, such as credit cards, overdrafts, contingent reserves, or because changes in the definitions, models, variables or weightings involved require significant programming efforts in order to adjust the rating system.

FINANWARE™ Credit Risk IRBa is a module specifically designed to execute this process because it is built on architecture allowing for the definition and flexible adjustment of the models, variables and parameters to be applied in the customer rating process.

This flexible solution makes it possible to select and change the variables considered for rating purposes to determine as parameters the amounts, ranges, weights and weightings to be applied in the process and the rating scales, evaluation of results and calculation of provisions (expected losses) to be generated by the process. It additionally provides analysis and scenario facilities to generate several rating stress-testing templates, thus reducing the need of back testing in order to calibrate the models.