Sponsors
Contact
Address:
-
Institute of Applied Mathematics, METU
Çankaya / Ankara, Türkiye
Invited Talks
Click for the book of abstracts.
Ozan Evkaya - University of Edinburgh, School of Mathematics
Importance of LLMs and their potential usage for Actuarial Science
Under the impact of fast-evolving AI technologies, recently, Large Language Models (LLMs) have emerged, and spurred public interest since November 2022. Regarding its ongoing development in the last couple of years, various research fields including actuarial science show various potential with certain implementation risks using LLMs. Specifically, the concept of multi-modality and multi-agents have immense potential how practitioners from different fields can communicate with machines. Under the light of these recent advancements, this talk aims to explore the significance of LLMs today, emphasizing their transformative potential in actuarial practices. Based on some pioneering LLM based tools applied to finance, one can argue that these advanced models enhance traditional actuarial functions such as risk assessment, policy design, and claims processing by leveraging vast datasets to predict outcomes and automate complex decision-making processes. With the ethical and responsible integration of LLMs, actuarial science (finance in a broader setting) may expose to new upcoming innovations and potential risks. Practically, this talk will include some background information regarding the history and enhancement of LLMs in the last couple of years. By referring to some recent papers and implementations, the discussion will be enriched by case studies that illustrate the practical applications and benefits of LLMs in real-world actuarial tasks.
Oytun Haçarız - Karabuk University, Actuarial Sciences
Recent developments in prediction power of GWAS and insurance
Since DNA-based genetic testing has become available in the 1990s, insurance has become one of the most contentious fields on the usage of the results of the genetic testing. Individuals and patient groups have often worried that they will be declined or not even afford on access of insurance due the fact that their genetic test results would be so highly predictive of greatly increased risk related to their morbidity and mortality. Insurers have often worried that, if genetic test results are not disclosed to them, adverse selection, i.e., information asymmetry between applicants and insurers in underwriting, will occur. At early stage, the insurance industry’s focus was on economically relevant monogenic disorders (such as Huntington’s Disease, Hypertrophic Cardiomyopathy, etc.), caused by rare but highly penetrant genetic variants. Necessary premium increases (due to unknown test results of these disorders) were estimated to be low under ‘realistic’ adverse selection scenarios. Recently, (a) the cost of testing whole human genome has dramatically decreased, which led insurers to expect genetic testing to become widespread and (b) genome-wide association studies (GWAS) using large biobank data have promised a potential to identify genetic substrate of many common disorders, caused by a combination of slight variants in many genes and environmental factors. In this talk, we will be discussing recent developments in prediction power of genetic material alone in predicting common disorders, and the future course of the disorders, and its impact, with the expansion in genetic testing and technological innovations in future, on insurance under adverse selection.
Hamza Hanbali - Monash University, Econometrics & Business Statistics
Mean-variance longevity risk-sharing for annuity contracts
This paper investigates longevity risk-sharing as a solution to the sustainability and affordability problems in the annuity market, and in particular how much longevity risk could be transferred back to policyholders assuming mean-variance preference functions. First, it provides dynamic risk-sharing rules for annuities. Second, it studies the contract properties from the perspectives of both the provider and individual policyholders. Third, it highlights and accounts for two levels of uncertainty and two levels of correlation induced by systematic longevity risk. Fourth, it provides necessary and sufficient conditions on the premium loading and the share of transferred risk, such that both parties prefer risk-sharing. The analytical and numerical results of the paper offer a deeper understanding of the effects of systematic and diversifiable risks on those preferences, and show that the products presented in this paper are suitable retirement solutions.
Uğur Karabey - Hacettepe University, Actuarial Sciences
Analyzing Mortality Rate Jumps within the IFRS 17 Framework
The financial performance of insurance companies hinges on the balance between incomes and costs. This balance becomes more complex compounded by uncertainties inherent in insurance contracts regarding the timing and amount of benefits. Traditional accounting practices often fail to capture these nuances accurately, leading to misleading representations of profit and loss. To address this, the International Accounting Standards Board (IASB) introduced the IFRS 17 Insurance Contracts standard in 2017. While the actuarial community is still in the nascent stages of grappling with the implications of IFRS 17, recent studies have made significant improvements. This paper investigates the financial ramifications of unexpected mortality rate jumps within the IFRS 17 framework, coupled with an analysis of cash flows over time, a pivotal aspect of the standard. The study integrates both a permanent and a temporary mortality jump model, to capture the dynamic nature of mortality rates. Utilizing data from the United States, parameter estimations are conducted, and simulations are generated to examine the impact of mortality rate jumps, particularly in light of the COVID-19 pandemic.
Sevtap Kestel - Middle East Technical University, Actuarial Sciences
Premium share between insurer and reinsurer in Stop-loss under stochastic loss behavior
The pricing in the stop-loss contracts is an important consideration of insurer and reinsurer. Based on historical loss amounts a stochastic model with the timevarying parameters to capture the time-dependent structure is developed. The analytical derivations of costs associated with reinsurance contract for reinsurer and insurer with constraints on time, loss amount, retention, and both retention and cap levels are made in the course of the claim payments. Along with these, the analytical forms of exposure curves are derived for determining the premium share between reinsurer and insurer under prescribed constraints. An illustrative case study is given at which the calibration of timevarying parameters is made using dynamic maximum likelihood estimator The findings depict that implementation of a stochastic model with time-varying parameters improves the prediction power and ascertains a fair risk share between insurer and reinsurer.
Selin Özen - Ankara University, Actuarial Sciences
Mortality Models in the Wake of COVID-19: Jumps, Forecasts, and Insurance Pricing
Population events such as natural disasters, pandemics, extreme weather, and wars might cause jumps that have an immediate impact on mortality rates. The recent COVID-19 pandemic has demonstrated that these events should not be treated as nonrepetitive exogenous interventions. Therefore, mortality models incorporating jump effects are particularly important to capture the adverse mortality shocks. The mortality models with jumps, which we consider in this study, differ in terms of the duration of the jumps–transitory or permanent–the frequency of the jumps, and the size of the jumps. To illustrate the effect of the jumps, we also consider benchmark mortality models without jump effects, such as the Lee-Carter model, Renshaw and Haberman model and Cairns-Blake-Dowd model. We discuss the performance of all the models by analysing their ability to capture the mortality deterioration caused by COVID-19. We use data from different countries to simulate the mortality rates for the pandemic and post-pandemic years and examine their accuracy in forecasting the mortality jumps due to the pandemic. Moreover, we also examine the jump-free and jump models in terms of their impact on insurance pricing, specifically term annuity and life insurance present values calibrated for both pre- and post-COVID data.