Asset and risk management

International Asset Management business must maintain accurate forward market valuations for the cars on its balance sheet. With around 10,000 models, and around 100 combinations required for each model, the business struggled to scale up to the millions calculations required using a traditional matrix approach. Variations in local expertise and processes made scaling even more of a challenge.

We built a centralized cloud-based valuation modelling engine, leveraging a portfolio of internal and external datasets. AI techniques (specifically the XGBoost technique) enabled accurate, dynamic long-term residual value predictions, across all models and variants, taking local market characteristics into account. By combining AI-driven predictions and expert local knowledge in the field, we were able to scale the solution quickly and effectively. This significantly improved asset valuations across all markets and enabled pre-emptive risk mitigation measures to be put in place.

In the first year, we recorded an increase ranging up to  25% in the accuracy of asset valuations, driving a significant improvement in predicted value of the total fleet. This in turn added value to the balance sheet.