Methodology
For the Variable Annuity Benefit Utilization Study that we released in December 2009, RCL used the classical Limited Fluctuation (LF) credibility method. We believe that this method is reasonable for this purpose, as it reflects both withdrawal frequency and severity and is computationally straightforward. In this context, severity measures the amount of a withdrawal in relation to a guaranteed Benefit Base.

The LF method utilizes the Central Limit Theorem, and assumes that withdrawal frequency and severity are independent, and that withdrawals are mutually independent. RCL believes that these assumptions are reasonable for this study, and that investors’ regular review of their financial plans and product performance with their financial advisor tend to mitigate the potential dependency effects of systematic withdrawal programs from year to year.

The LF method ascribes full credibility when the number of withdrawals is such that, with a high degree of probability, the actual total withdrawals should fall within a narrow confidence interval around the expected total withdrawals. The number of withdrawals is a natural choice for this purpose, as it implicitly reflects key factors that intuitively affect credibility – both credibility and the number of withdrawals tend to increase when exposure increases and when withdrawal frequency increases.

To read more, please refer to the complete white paper contained in the pdf document below.

Limited Fluctuation Credibility and Variable Annuity Benefit Utilization Study