Having engaged in AML- and IRB related assignments with SIFI:s as well as banks just below the threshold, a few common observations stand out and where we at FCG believe that the industry, with guidance from the FSA, can take further steps to enhance the battle against money laundering:
Firstly, the collaboration between the dedicated AML / FC units and the risk model validation departments can be improved. Particularly, the discipline of validating credit risk models is very mature, confirmed by the advancements we observe cross-institutions and the requirements set by the regulators. The balance between requirements and expectations from the FSA on one hand, and the ramp up of technology and methodology within the institutions is well managed from both sides, and where it is expected that the implementation of the next generation of IRB models and their validation features will output robust and well calibrated credit risk models.
This is not our experience in the area of development and validation of AML models, where the approach can wary significantly between institutions, and where the expectations from the FSA is only vaguely formulated in comparison to the requirements in the IRB-framework. It is a privilege for FCG to help the industry on the mission to establish an industry standard, but we believe that one root cause for this is the limited interaction between AML-units and the (other parts of the) risk departments. By deepening the collaboration, the AML initiatives could draw upon all the current and past experience from the risk departments in developing and validating models. This would include features such as ensuring a correct scope of application for the models, techniques and methods for data management and data quality assurance, as well as hard core quantitative tests for predictive power, model stability etc.
Secondly, it is expected that institutions are establishing governance structures that support a risk-based approach to risk management and the fight against criminal organizations’ money laundering attempts. To this point, it can be surprising to see the resource allocation in regard to model validation programs and the various purposes and portfolios that are covered. Here, we would encourage the FSA to acknowledge that there are limits to how large an (validating) organization can grow and encourage, by means of harmonized guidelines that encompass all validation activities (i.e. including traditional risk models as well as AML models), institutions to focus on the models that are the most important. Marginally small credit portfolios should not require yearly validations. By approaching it this way, we believe that the quality of validations, irrespective of the purpose of the models, will improve which will be of benefit for the institutions, the FSA and the society as a whole.
We would be happy to hear your opinion on this matter!