top of page



Risk Models Development
Banks have been using risk models in their credit decisions for a long time. Over time, linear regression models evolved to logistic regression models. Today, the democratization of artificial intelligence and machine learning world with open source libraries and increased computational power changed the paradigm from "not feasible" to "must-have for competitive advantage". We help our clients develop and adopt explainable and of high prediction power machine learning models for all areas of risk practice such as PD (including application and behaviour), LGD and EAD. We bring our deep industry experience and holistic approach for end-to-end risk management to our engagements in order to ensure high returns for investments made.
Strategy Consulting
Machine Learning & AI Teams
Risk Models Validation and Monitoring
We believe that model validation and continuous monitoring can be positioned as a portfolio management function for risk models. This function can identify and prioritise the improvement areas for all models not just changes from development state while fulfilling regulatory requirements. With the introduction of risk models developed with machine learning methodologies, traditional validation approaches will need to change in terms of dimensions covered under qualitative and quantitative areas. We believe this is a good opportunity for banks both to transform their validation methodologies and to evaluate how they position the validation function as a whole.
Strategy Consulting
Machine Learning & AI Teams
Credit Risk Appetite Engine
Over our years of experience, we came to the conclusion that all credit underwriters strive to make the best credit decision for their banks in given situation, but due to many internal and external factors, it has been usually very difficult to structure all decisions and produce a central risk appetite engine at bank level. Such risk appetite engines will allow both CEOs and chief risk officers to pull the levers for their underwriting machines aligned with their business objectives, to shape their future asset quality and even to control price levels of business units. These risk appetite engines range from sophisticated models for cut-off decisions in individual banking to models for group level limit package decisions in business banking. Our engagements produce implementation ready structures with tested prototypes and cover organization and governance aspects.
Strategy Consulting
Limit Structure and Credit Experience
Time-to-yes for credit all decisions has become one of the leading drivers in capturing customers of targeted risk profile. The experiences of customers, sales people and credit underwriters until time-to-yes are somehow equally important to this end-to-end duration. We believe most of the issues in existing credit experience and long time-to-yes can be associated with one root cause: internal limit structures of banks. For individual banking usually credit products and limit products are 1-1 match, but for business banking, most of the banks can utilize credit products of different maturity, currency and other specifics under a single limit product. This creates non-standard procedures, heavy manual involvement and even leakages in some cases. We start designing end-to-end credit experiences from limit structures and continue with all related aspects such as sub-allocation rules, authority levels and knock-out / escalation cases. All these designs aim to fix pain points along customer / employee journeys. We produce implementation ready journeys and structures covering process, organization and governance aspects of this key function of banks.
Strategy Consulting
Experience & Service Design
Early Warning, Monitoring and Collection Structure
We believe most banks leave money on the table with siloed structures for loan monitoring and collection. Early warning signal detection should be the responsibility of the whole organization with varying levels of involvement. Risk related metrics in branch personnel scorecards will not suffice by itself. We approach this important subject with a holistic view and design structures that minimize leakages. Our structures include tools for collective early warning detection, standardized procedures and recommended actions for transition to worse, work-out strategies for different profiles of customers and exposures including restructuring plans. We produce results based on data driven approaches and leverage predictive models where applicable.
Strategy Consulting
Machine Learning & AI Teams
bottom of page