Description: The American Academy of Actuaries and the Society of Actuaries have joined resources to maintain this Economic Scenario Generators Excel Workbook tool for use in regulatory reserve and capital calculations. Please see the SOA website research pages for further information on this tool and how to use it.

Description: This session will explore governance considerations for insurance carriers leveraging machine learning models. During the session, panelists will share insights on the evolving predictive model and AI regulatory landscape, as well as techniques for developing or enhancing AI governance frameworks that comply with emerging regulatory guidance.

Description: Claims are insurers deepest and richest transactional data – yet we rarely unleash the full potential of this data. In this session, we will discuss potential sources and attributes of claims data, traditional approaches alongside new analytic techniques and how to integrate models into existing claims management processes. During this session, the presenters will discuss different areas of how machine learning can drive significant value unleashing the potential in claims data, these include data capture, claims management and workflow design as well as behaviour prediction. In addition, the audiences will also have opportunity to learn different use cases of how claims models taking advantage of machine learning, such as foreign death fraud detection and critical illness triage.

Description: This presentation discusses a process for determining the minimum reinsurance retention level consistent with defined parameters. The conclusions reached were: 1. Under PDS dependence, optimal reinsurance strategy is multivariate stop loss form. 2. Explicit solutions could be derived in certain bivariate model. 3. Binding and unbinding constraints are equivalent. From Actuarial Research Clearing House, 2012, Vol. 1.

Description: This is a presentation on computing tight bounds for insurance payments with nonlinear risk. Semi-definite Programming is defined, some examples are presented, and non-linear applications are presented. From Actuarial Research Clearing House, 2012, Vol.1.

Description: Authored by a MIB Solutions team of Tom Rhodes, Stuart Klugman, Marianne Purushotham and Stacy Gill, the report summarizes the results of a study on the application and adoption of credibility theory within the life insurance and annuity industry. In addition to the report, numerical examples related to the application of credibility to company mortality and lapse data are provided in the four Excel files. This excel spreadsheet is a numerical example relating to the application of Credibility for lapses by policy for limited fluctuation. Please see the SOA website research section for the full report on Credibility Theory Practices.

Description: Asset cash flow modeling and asset assumptions are growing in importance with PBR especially considering the low interest rate environment. The Academy’s Life Practice Council and Life Valuation Committee have formed a Low Interest Rate Asset Adequacy Testing Task Force which has conducted an industry survey to raise awareness and summarize the actuarial best practices for financial reporting actuaries practicing in today’s very low interest rate environment. As the scrutiny on the asset side ratchets up, actuaries must consider how best to develop the asset assumptions as well as model the reinvestment and disinvestment strategies. Coherent assumptions and modeling practices across modeling exercises are expected while respecting the purpose of each calculation.

Description: The Mortality Improvement subgroup of the American Academy of Actuaries’ Work Group and the Society of Actuaries’ Mortality and Longevity Oversight Advisory Council released a recommendation for a set of improvement factors that vary by gender and attained age to be used in conjunction with the 2008 Valuation Basic table or the 2015 Valuation Basic table for AG-38 purposes. This recommendation has been adopted by the NAIC’s Life Actuarial Task Force.

Description: Hear about some applications using quantile regression-derived individual and group claim distributions and how they may be used in various traditional health care modeling situations. The session will not focus on the technical use of quantile regression to obtain estimates of claim distributions, but more so is intended to focus on the uses of the distributions in actuarial modeling and the additive framework in assessing claims distributions. Presenters will show how actuaries can better assess risks by focusing on the individual- and group-specific claims distributions as opposed to traditional point-estimate-centric models.

Description: Over the last 10 years, Artificial Intelligence (AI) healthcare foundation models have made tremendous strides when it comes to delivering highly accurate predictive results when compared to traditional models. This is especially true for models developed and trained on the ‘language’ of healthcare. When trained on healthcare data sequences (claims, EHR, CPT, ICD, pharmacy, lab, etc.), these models deliver far more accurate results as to WHAT beneficiary-level health events will occur and even WHEN they will occur.