Data analytics models in risk management: impact on corporate strategy

What it is about

Allan Calderón Moya explained at the Third Annual Risk Conference (September 2025) how data analytics has evolved within a financial institution, from the first risk models to anticipatory analytics, and what impact it has on corporate strategy.

The 40 slides go through four stages (2002-2008 first models, 2009-2014 consolidation, 2015-2019 beyond risk and 2020-2022 pandemic and post-pandemic), with cases of credit, market and operational risk models, collections, anti-money laundering with network analysis, cross-selling and an alternative origination model, and close with data quality and lessons learned.

Objectives

• Show how the use of data analytics evolved from risk management toward the business.
• Present concrete cases of models and their contribution to strategy.
• Highlight the role of data governance and quality.

Key takeaways

1. The creation in 2019 of a data analytics and governance directorate made it possible to offer the business advanced analytics tools, customer experience assessments and data-driven decisions.
2. Models went from measuring risks (credit, market and operational VaR) to generating opportunities: automatic origination scores, empathetic collections, customer retention and cross-selling.
3. In anti-money laundering, network analysis (customers as nodes and transactions as edges) was used to calculate centrality scores and scores for participation in risky communities.
4. For populations with little information, such as digital credit card customers, an alternative two-stage origination model was built.
5. Lessons include: the scope of models is limited by the lack of a robust data quality scheme, one need not wait for a "super model" to meet a business need, and strengthening validation and challenge of assumptions is essential.