In the last financial year, Australia’s General Insurance industry profits fell 12% YoY, largely as a result of unfavorable claims experiences. Claims inflation – how the size and frequency of claims are changing – represents a significant cost within the insurance industry that needs to be effectively managed. Forecasting how claims patterns are likely to change is a vital input for any insurer or underwriter to be able to plan current pricing, future pricing, and reserving decisions. Claims inflation estimates need to be able to effectively decompose inflation into changes in claim rates due to an evolving customer book and product mix and “underlying” inflation, and other factors like economic, regulatory changes, changing use patterns and business process changes.
Despite being an essential component for the smooth operation of an organisation, many insurers find it challenging to get accurate estimates of claims inflation with robust confidence intervals. The traditional actuarial technique incorporates many assumptions about the problem that don’t tend to hold in the real world, can be time-consuming to create, and struggle to incorporate additional supporting data sources.
In an article by Deloitte, “From Mystery to Mastery: Unlocking the business value of Artificial Intelligence in the insurance industry”, they identified:
“AI offers great opportunities for optimizing existing and enabling new procedures, giving a competitive edge to businesses today and in the future”.
At Intellify, our experience with major Australian Insurers shows that the latest range of Machine Learning and statistical methods provide an alternative approach to the problem and allows for a wider range of non-linear forecasting capabilities. These provide robust projections of future claims rate and include specialised techniques like multi-level/ hierarchical models, Bayesian statistical time series methods, and Recurrent Neural Networks (RNNs).
Intellify is a market leader in providing machine learning solutions. If you are interested in discussing how we can help your organisation, please don’t hesitate to reach out to us at firstname.lastname@example.org