Description: Mortality Model Selection and Combination in RAlternative approaches for selecting mortality models and forecasting mortality have been proposed. The usual practice is to base mortality forecasts on a single mortality model selected using in-sample goodness-of-fit measures. However, cross-validation and other alternative out-of-sample measures are increasingly being used in model selection, and model combination methods are becoming a common alternative to using a single mortality model. Such methods combine predictions from multiple mortality models to reduce model choice uncertainty and improve forecast accuracy. Explore techniques to select and combine mortality models using the freely available open source R package CoMoMo (https://github.com/kessysalvatory/CoMoMo), which extends the widely used R package StMoMo (https://CRAN.R-project.org/package=StMoMo). We will use data for the USA and other countries to discuss and demonstrate four different model combination approaches including simple model averaging, Bayesian model averaging, model confidence set, and stacked regression ensembles.

Description: "Living to 100 in the time of COVID-19: A Study of Late –Life Mortality Trajectories, Leonid A. Gavrilov, Ph.D., Natalia S. Gavrilova, Ph.D., Academic Research Centers, NORC at the University of Chicago Swedish Mortality Does Not Decelerate in Recent Cohorts, or Does it? – Roland Rau, Ph. D., Professor, University of Rostock & Max Planck Institute for Demographic Research, Rostock, Germany; Marcus Ebeling, Ph.D., Karin Modig, Ph.D., Associate Professor, Anders Ahlbom, Ph.D., Professor Emeritus, Karolinska Institute, Stockholm Sweden; Kai Kaufhold, Aktuar DAV, Ad Res Advanced Reinsurance Services GmbH, Cologne, Germany; Annika Ziegler, Lukas Stake, Hochschule Koblenz & Ad Res, Germany"

Description: This paper applies the research in Prospect Theory of Kahneman and Tversky to explain why sales of annuities lag sales of life insurance products. The author theorizes that life insurance is easier to sell because customers perceive themselves to be in the domain of losses, while potential annuity customers perceive themselves to be in the domain of gains. Kahneman and Tversky showed us that people are mostly risk averse in the domain of gains and risk seeking in the domain of losses. Purchase of an annuity or life product is a risky investment. Customers will find it easier to take the risk in the life insurance domain of losses than the annuity domain of gains. The author then proposes a hybrid life annuity product that might address both mortality and longevity risk in one product that would allow both products to be more competitive than stand-alone products.

Description: This article helps companies better understand the currently proposed upcoming VM-22 regulation by comparing the VM-22 measurement model to LDTI MRBs. The article highlights key considerations as companies plan for adoption of these two measurement models. These considerations should help inform a company’s strategy related to the business impacted by these standards.

Description: Term products exhibit a wide range of behavior as policyholders can elect different term periods and conversion options. While term products have been around for decades, it's only recently that longer level term products are reaching the end of the level term period. Term conversions have historically been an afterthought and are rarely modeled. Discuss the use of predictive modeling techniques such as LASSO regression and cross-validation to build models for level term lapses and post-level term shock lapses, as well as the extrapolation of experience for shorter term products to longer term products (e.g., extrapolating 20-year term experience and applying to 30-year term). Explore the use of traditional techniques to develop term conversion rates and a framework for modeling the assumption. Challenges and limitations of setting lapse and conversion rates will be discussed. By attending the session, you will: - Understand the complexities of term policyholder behavior. - Understand the challenges in developing and implementing assumptions for lapses and conversions. - Understand the application of traditional and predictive modeling techniques for term policyholder behavior.

Description: It has been a little over a year since the release of the Society of Actuaries (SOA) Research Institute's Mortality Improvement Model, MIM-2021. Since then, waves of COVID-19 have continued to occur in the U.S. The uncertainty around the COVID-19 pandemic and the slowing of U.S. general population mortality improvement prior to the pandemic are just a few of the reasons that make developing MI projection scales difficult. Uncoverthe results from intercompany surveys on mortality improvement, including one that addresses COVID-19, discover insights from a mortality expert panel on COVID-19 and future mortality, and discuss enhancements to MIM-2021. By attending the session, you will: - Understand MIM-2021 and 2022 enhancements. - Understand SOA Research Institute activities around COVID-19 and impact on mortality improvement. - Apply research results and insights presented to daily work.

Description: Insurance carriers today must balance a multitude of factors to remain relevant and innovate- -process improvements and efficiency (automated or traditional); -customer experience (less intrusive, more data); -distribution (aging traditional, product focus, emerging opportunities); and -mortality. While these factors are not new, true innovation in these areas require a convergence of engagement among actuaries, underwriters and data scientists to determine the specific pain points and appreciate both the benefits and consequences of various paths. Explore examples of how best to approach these topics, engage different disciplines, avoid pitfalls and ensure a holistic view. Real life examples and lessons learned will be shared during this session.

Description: This paper applies the research in Prospect Theory of Kahneman and Tversky to explain why sales of annuities lag sales of life insurance products. The author theorizes that life insurance is easier to sell because customers perceive themselves to be in the domain of losses, while potential annuity customers perceive themselves to be in the domain of gains. Kahneman and Tversky showed us that people are mostly risk averse in the domain of gains and risk seeking in the domain of losses. Purchase of an annuity or life product is a risky investment. Customers will find it easier to take the risk in the life insurance domain of losses than the annuity domain of gains. The author then proposes a hybrid life annuity product that might address both mortality and longevity risk in one product that would allow both products to be more competitive than stand-alone products.

Description: Explore similarities between actuaries and data scientists as speakers share personal experiences on the importance of organizational structure. Attend the Actuary, Data Scientist, or Actuarial Data Scientist webcast to discover the effectiveness of partnering both of these career paths in a life insurance atmosphere. In addition to personal experiences, speakers will discuss several success stories where collaboration and connection became the key to driving innovation. Come learn about the progress these professions have caused in pricing and risk management.

Description: As AI is incorporated into more life insurance functions, take a look at the broader use of AI in insurance and how it may negatively impact the diversity, equity and inclusion efforts that need to be at the forefront of products introduced to the market. By attending the session, you will:Be able to recognize how AI may have bias coded into specific functions and how to use bias recognition and tools to minimize that impact.