Description: Amid the growing adoption of AI/ML models for underwriting and pricing, Life and Annuity carriers are facing increasing regulation in North America on Model Risk Management (MRM) and the use of AI. Colorado has set a precedent in the US by requiring reporting in June 2024, with other states looking to follow suit. To ensure compliance, Chief Risk Officers and Heads of MRM, working with their business partners, will need to evolve their MRM and AI risk frameworks, enhance model validation capabilities, and strengthen governance. This session will present: - Key trends and challenges faced by carriers amid increasing model landscape complexity and regulatory expectations - Regulatory outlook and implications for carriers' MRM and AI risk programs - An approach to enhance MRM and AI risk frameworks, including specific model validation program updates to address the unique risks of AI/ML models - Opportunities to enhance both MRM validation effectiveness and efficiency through the adoption of more risk-based approaches, smart sourcing, and automation - Case studies of how leading carriers are evolving their MRM and AI risk programs Upon attending the session, the attendees will develop a good understanding of increasing regulatory expectations, the required MRM and AI risk framework enhancements, and approaches to design and execute an effective MRM validation program in a cost-efficient manner.

Description: An overview of guaranteed lifetime withdrawal benefits (GLWB) on registered index-linkedannuity (RILA) products and a stochastic analysis of market risks.

Description: Insurers have long relied on scenario analysis to understand the long-term impacts of potential business decisions, but the advent of rapidly changing market environments and a growing desire to optimize capital has spurred business leaders to leverage new tools in order to explore potential options. Companies are beginning to explore new analytical solutions that trade-off accuracy for speed for a more robust approach to impact quantification. In particular, they are looking at the use of additional what-if scenario models, built from the top down, to support decision-making processes. Actuaries can help take charge in the development of these top-down analyses by leveraging data and actuarial techniques to supplement their existing models. In this way, business leaders can simultaneously rely on the actuarial rigor of typical standalone cash flow projection models, while also receiving support for real-time business decision-making. The potential of what-if scenario modeling can help actuaries unlock a powerful new approach to scenario analysis to enable more rapid decision-making. Learning objectives: In this session we will discuss the fundamental concepts of what-if scenario modeling in an actuarial context, including: - Core principles and components of what-if scenario modeling - Common challenges and considerations around the implementation of what-if scenario modeling - Potential solutions to challenges and their implications

Description: This session will discuss basic and advanced hedge modeling concepts. It will explain common hedge strategies used to support guarantees and options embedded within life and annuity products and the materiality of these products within the US. Attendees will learn how Machine Learning acts to reduce the computer processing needed to simulate the dynamic hedging of product liabilities projected in the future across many stochastic scenarios. Actuaries working in consulting, software development, and a large life insurance company will provide a wide range of hedging viewpoints ranging from a broad industry perspective to a hedging practitioner. Attendees will obtain a comprehensive understanding of a novel ML approach to modeling dynamic hedges and the hedge optimization opportunity it presents.

Description: Even in published scientific research, we find fraudulent data. What is an actuary to do? Let's consider our actuarial principles and standards of practice to question our data quality and truthfulness. In specific, the article addresses ASOP 56 on Modeling and considerations of model risk as well as ASOP 23 on Data Quality. Recent examples are given from a Harvard research fraud controversy as well as questionable sumo statistics.

Description: Article describing content and key takeaways from experience completing the SOA PA Certificate

Description: This project was developed by the Predictive Analytics Subgroup of the Society of Actuaries (SOA) Research Institute’s Individual Life Experience Committee (ILEC).

Description: Model runtime is a common challenge for actuaries in the life and annuity space. In particular, applications such as pricing and forecasting may require the use of nested-stochastic calculations, which may require extensive computing resources or use of rough approximations. There are also other areas where runtime may be a challenge, such as projecting certain reserves or layering solving algorithms within actuarial calculations. Recently, runtime challenges have been exacerbated by the development of increasingly complex insurance products and changes in reporting frameworks, such as IFRS 17, LDTI, PBR, and Solvency II. Many approaches have been explored by actuaries in the past to address model runtime challenges. To further this research, the SOA has recently published the paper Predictive Analytics and Machine Learning - Practical Applications for Actuarial Modeling. The authors of the paper will share their insights on innovative techniques using the implementation of artificial intelligence and machine learning (AIML). Come examine an introduction into AIML, review how AIML can be used to circumvent nested stochastic and other runtime challenges, explore case studies, review how costs and benefits can be measured and discuss model risk management considerations. Key learning objectives include: - Understanding the fundamental concepts of AIM - Understanding applications of AIML for actuarial modeling - Learning about the costs and benefits - Learning how AIML can be applied in practice to improve model runtime