Opinion pieces: On data and models: Is more always better?

Abstract

With regards to data and model sophistication, the new mantra for financial services and FinTech seems to be ‘the more the better’, supported by attractive business cases in risk underwriting, fraud detection, customer lifetime value management, conditional investment risk/return optimisation and improving customer journeys, to name a few. However, more data and more sophisticated models are not always a universal panacea and may lead to bad business outcomes if not managed appropriately in the context of the desired business outcome. This paper summarises the evolving business cases for increasing data and models in the risk management domain and their associated risks. Making the associated risks transparent naturally leads to the conclusion that a timeless risk mitigation approach — common sense — is critically necessary to complement the more structured model risk management (MRM) framework that is evolving.