Reserving Styles – Are Actuaries In-Sync with their Stakeholders?

Motivation: Reserving actuaries are constantly faced with forming estimates that inherently reflect consideration of data and information that spans from initial expectations to actual claims experience. The actuaries and their stakeholders (e.g., members of management) may implicitly or explicitly apply different perspectives on the relative merits of projections based on actual experience or initial expectations, or projections that reflect a blending of the two. As an actuary associated with an audit firm, Mr. Littmann encounters these situations frequently, primarily in a reserving context. Apparently subtle differences in perspectives among actuaries and among various stakeholders when actual experience diverges from expectations (generating divergent projections of unpaid claim estimates) can generate substantial dialogue. The paper presents an exploration of historical progressions of recognizing accident year losses, casts light on certain implications of common actuarial methods, and provides insight on the notion of a reserving cycle akin to an underwriting cycle. The investigation provides a framework for dialogue among stakeholders to the reserving process, as well as identifies areas where actuaries may be able to enhance the technical aspects of, and their communications from, their work processes.

Method: The paper provides examples of the historical progression of accident year loss ratios booked by the industry in aggregate and for a sample of companies. A model is presented to demonstrate the extent to which a combination of cyclical accident year loss ratios and alternate views from stakeholders on their ‘best estimates’ to be adopted at a point in time can create differences in the estimates of unpaid claims liabilities.

Results: The outcomes are a framework for expressing views on responsiveness to the emerging claims data in relation to initial expectations, as well as illustrations that provide actuaries with insights on the implications of differing views on loss picks. The paper identifies matters for actuaries to discuss among themselves and with their stakeholders. Discussions around these concepts and implications in advance of the periodic reserves meetings may help the meetings go more smoothly.

Conclusions: Apparently small differences in styles for making loss picks from among projections that span from initial expectations to extrapolations from actual data can yield noticeable differences in reserve estimates. Differences in selection approach between stakeholders do matter and create the need for discussion, transparency and documentation.

Keywords: Reserving Methods, Management Best Estimate, Reserve Variability, Credibility