Loss reserve variability and loss reserve errors

Abstract

Purpose – The purpose of this study is to assess the level and variability of Ghanaian property and liability insurer’s reserve estimates to examine its sources and ascertain if reserve errors are random or not (i.e. manipulated or not). Design/methodology/approach – It uses information on insurer claim reserve provisions, claims outstanding, claims incurred and claims paid for the period of 2000-2010. Categorizing the sources of variation as endogenous and exogenous, the authors use the panel correlated standard error regression model to determine sources and magnitude of industry reserve error. Findings – The study finds that size, age, lag of loss reserve error, inflation rate and real gross domestic product are significant in determining the degree of reserve error variation. Type of ownership (domestic or foreign) is, however, not a significant source of variation. Further, the authors found that industry reserve errors are random (not manipulated) across firms, suggesting that sampled insurers act independently on reserve error decision making and are not influenced by industry trends and competition. Research limitations/implications – The main research study limitation is the difficulty involved in obtaining annual statements from insurance companies in Ghana. Reluctance of companies to make statements available impeded on the smooth flow of the study during data collection. Practical implications – Policy-wise, this suggest that regulatory bodies can uniquely set reserve error levels for existing firms with little influence on competition. Further, the Ghanaian insurance regulator does not to focus on the type of ownership (foreign or local) when setting regulatory standards. However, size of the company and age (length of operation) should be considered. Originality/value – This paper is the first empirical study to examine the loss reserve error and loss reserve variability of Ghanaian property and liability insurance companies.