Business Challenge:

The client was one of the biggest banks in the GCC region with an asset size of around 60 Billion USD. The Bank wanted to adopt Foundation Internal Ratings Based (F-IRB) approach of BASEL II.

For compliance with the F-IRB approach the bank needed to estimate probability of defaults (PD) which in turn needed credit rating models. Aptivaa's scope of engagement in the bank was (i) To develop the bank's rating Landscape (ii) To validate the existing large corporate rating models (iii) PD estimation using historical default experience & (iv) Development of new rating models for SME, project finance , private banking , securitization and real estate fnance porfolio

Our Approach

For an exhaustive coverage of all the aspects of credit rating model development and validation, we used a modular approach for the overall engagement. Outlined below are the key modules, our approach and the deliverables.

Portfolio Review: This module included the analysis of the bank's portfolio and using the results to determine the optimum number of rating models required by the bank. The deliverable of the module was a rating landscape document providing a mapping of banks portfolio and the corresponding rating model.

Model Validation: The client had been using large corporate rating model for last five years but it has not undergone any validation since inception. For validating these models we collected data for the last five years and conducted statistical tests inline with BASEL working paper 14 guidelines to assess the discriminatory and predictive power of the rating models. Benchmarking of these models against external rating agencies was also done. Based on the results of the tests refinements were made to the existing model. A detailed validation framework was prepared for regulatory compliance

PD Estimation: For estimating the probability of default for the bank, data of actual historical defaults was collected for the last five years. The data was used to prepare a transition matrix and the average historical default rate was used to estimate PDs. It was observed that the bank has experienced very few defaults in the last five years leading to a skewed data set for a robust model development. This problem was overcome by calibrating the PDs using the FSA approach for low default portfolio.

Model Development: Rating models were developed for SME, Private Banking, Securitization, Project Finance & Real Estate Finance exposures of the bank. Based on several statistical techniques appropriate rating models were developed for each of the aforementioned exposures. Excel based templates for each of the rating models was provided to the bank. 

Business Benefits

The coverage of rated obligors in banks corporate banking portfolio was increased and substantially improved the risk rating capabilities of the bank. The regulatory requirement related to F-IRB implementation was also complied with and helped bank get the regulatory approval.