Comparative Study of Predictive Analytics Algorithms and Tools on Property and Casualty Insurance Solvency Prediction

Abstract

The Insurance Regulatory Information System (IRIS) is a collection of 13 financial ratios used primarily by regulators to determine the solvency of an insurance company. Predicting the IRIS values can help companies to stay compliant with the IRIS regulation. Knowing ahead of time, the company can take actions to prevent potential IRIS unusual values and ensure its financial health. In this article, we compare the prediction accuracy and calculation speed of the current mainstream machine learning algorithms and libraries in predicting the IRIS ratios. Based on the best models selected from the comparison of the algorithms, we develop a Shiny R web APP for companies to predict their IRIS ratios for future years.