Pricing retail automotive loans

Date18th March 2022

Case study

We worked closely with one of Australia’s leading Automotive Finance firms to build sophisticated Artificial Intelligence and Machine Learning algorithms for pricing retail auto loans.

The Challenge

One of Australia’s top 5 auto finance organisations with 50% of Australia’s auto loan market share needed to replace their current cost base pricing methodology with an efficient risk based pricing structure.  

The business’ current pricing strategy was focused on covering internal costs and adding a margin. However, once the price was put into the marketplace, there was no understanding of whether customers would “take-up” these internally set interest rates. In contrast to this, the risk based pricing model would use advanced machine learning to calculate a default probability using a range of customer related data.

The Solution

The solution involved developing customer elasticity models to predict “take-up” at any given interest rate per customer, allowing the business to price loans to increase conversion and margin at a customer level. We provided consulting services to build Pricing Governance Frameworks, Business Objectives and Constraints, Pricing Committee Charters, Reporting Packs and resource allocation for Pricing and Data Science teams. The project also involved the implementation of pricing software and optimisation using behavioral modelling to increase business benefits on their loan portfolio.

The Business Impact

After a 3-month proof of concept and project implementation, the annual business benefits to the business net profit was estimated to be $20M.