Micro-Level Loss Reserving Models with Applications in Workers Compensation Insurance

Accurate loss reserves are essential for insurers to maintain adequate capital and to efficiently price their insurance products. Loss reserving for Property & Casualty (P&C) insurance is usually based on macro-level models with aggregate data in a run-off triangle but many limitations exist in these models despite of their simplicity. A run-off triangle is a summary of underlying individual claims data. In recent years, a small set of literature has proposed reserving models that are based on underlying individual claims data, analogous to approaches used in the life insurance industry. They are referred to as “micro-level models”. We demonstrate the application of a micro-level reserving model in a large portfolio of workers compensation insurance provided by a major P&C insurer. A hierarchical model is specified for the individual claim development process, with multiple blocks to model various events (claim occurrence, notification, transaction occurrence and payment amount) during the claim development. The model is estimated with historic data, validated with a hold-out sample, and compared with commonly-used macro-level models. We show that the micro-level model provides a more realistic reserve estimate than that given by the macro-level models, and the estimation error is largely reduced through the use of individual claims data. The micro-level model is more likely to capture the downside potential in reserves and to provide adequate allowance when extreme scenarios occur. We conclude that micro-level models provide valuable alternatives to traditional models for loss reserving. The data used in this study consist of over five million records and have the nature of “big data”. Our work is similar to data analytics in that we also search for patterns and use these patterns for predictions. It is different from data analytics in the techniques adopted for pattern searching: while big data analytics typically