Description: As the portion of inforce business valued under PBR grows, it becomes increasingly important to not only calculate reserves for point-in-time valuation but to also project reserves. This is especially critical for actuarial functions such as pricing, business strategy/forecasting, and asset adequacy testing. In this session, we will review PBR projection techniques based on work completed by the American Academy of Actuaries when developing the PBA Projection Practice Note. In addition, this session will examine specific challenges that can arise when actuaries use models to project reserves at future dates. Topics covered include mortality improvement, credibility, scenario generation, starting asset projections, and various other technical issues related to nested modeling requiring both inner and outer loop projections.

Description: Results of survey, conducted recently by the Taxation Section of the Society of Actuaries, on how federal income tax is reflected in the projection work being performed.

Description: According to the Actuarial Standards Board’s Actuarial Standard of Practice No. 49 (ASOP 49), actuarially sound Medicaid managed care capitation rates will include a provision for underwriting gain in order to provide compensation for the risks assumed by an MCO. However, there is no generally accepted method for quantifying underwriting gain in managed Medicaid capitation rates. This article discusses a new model developed to address this void and calculate underwriting gain. The model was developed by researching existing financial literature and performing a statistical analysis of industry data to build out a framework that can be leveraged to estimate underwriting gain.

Description: While many know that machine learning is a big deal and even have a general idea of what it is, many do not know what machine learning can do for them. As an alternative to citing achievements or explaining algorithms, this article outlines the challenges that occur in statistics and the solutions that machine learning offers. Some of those challenges include correlated predictors, missing data, data interaction, and prediction intervals. Corresponding solutions include regularization technics, missing value imputation with KNN, tree based models, and cluster analysis for backtesting.

Description: This 1979 session is about recent advances in Prediction Theory. Topics included: 1. Statistical methods to predict path properties of a Time Series. 2. How and to what extent Box-Jenkins Time Series methods can be modified to predict two or more time series simultaneously. 3. Workable techniques that are already being successfully used in practice.