Investment in big data analytics and loss reserve accuracy: evidence from the U.S. property-liability insurance industry

Abstract

This study explores the impact of big data analytics investment on loss reserve accuracy in the U.S. property-liability insurance industry. Utilising a dataset of 1243 insurers from 2002 to 2016, we find a significant association between higher investment in big data analytics and more accurate loss reserve estimates. Our analysis distinguishes between over-reserving and under-reserving behaviours, revealing that big data analytics contributes to the reduction of both. The study employs entropy balancing, internal instrumental variable estimation and errors-in-variables regressions to enhance the robustness of the findings. This research not only fills a gap in the academic literature but also provides practical implications for enhancing the precision of loss reserve estimates through technological investments.