Events Library

Sustainability is a multidimensional concept requiring systemic thinking across environmental, economic, and social domains. While actuaries have traditionally focused on financial and environmental resilience, we will argue that human diversity—particularly neurodiversity—is equally vital. Neurodiversity recognises that cognitive variations such as autism, ADHD, and dyslexia are natural differences, offering significant potential for innovation and adaptability. In this presentation, we will examine the challenges neurodivergent individuals face, including exclusionary recruitment processes and inadequate workplace accommodations. We will propose actionable strategies to address these barriers, including skill-based hiring, individualised workplace supports, and neurodiversity-focused leadership training. Further, we will advocate for developing metrics to measure the impact of these initiatives, leveraging actuarial expertise to assess long-term value. We will draw on global case studies, such as SAP’s Autism at Work program and the International Actuarial Association’s growing emphasis on diversity, to highlight best practices. By integrating neuro-inclusivity into actuarial workplaces, professional societies, and global standards, we will demonstrate how actuaries can drive systemic change. Ultimately, we aim to show that neuro-inclusivity is not just an ethical imperative but a strategic opportunity. By embedding these principles into the profession, actuaries will enhance innovation, strengthen decision-making, and position themselves as leaders in shaping a more inclusive and sustainable future.

With the outbreak of Covid-19 pandemic, the Brazilian supplementary healthcare sector became a conducive environment for using complex data analysis and modeling tools. In this study, we apply different Machine and Deep Learning techniques (SVM, XGBoost and RNN) to predict healthcare expenses and evaluate if these techniques would present better performance in comparison to traditional ones, such as time series and regressions. Prediction scenarios were generated upon expense official databases between 2015-2022, considering two panoramas: (i) real, and; (ii) counterfactual, in which we assume the non-existence of the pandemic data for 2020. Using RMSE as the performance indicator, we find out that XGBoost model presented the best performance for the real panorama, with better fit in 32.2% of the scenarios. For the counterfactual panorama, we observe that RNN and SVM models obtained better fit in 22.3% of the cases. It is noteworthy that, until now, no studies were identified that address the use of predictive Machine and Deep Learning models into the Brazilian healthcare expenses. We also expect that this study offers insights for decisions made by the several players in this sector, such as operators and regulators, especially when it comes to pricing and development of healthcare products.

In recent years, the Brazilian agricultural insurance market has faced catastrophic losses exceeding USD 1 billion, primarily due to climate change and insufficient data for risk selection and pricing. The predominant use of public data, such as national statistics, has proven inadequate. To address this, an innovative productivity estimation model was developed, integrating both public and private data sources, including information on soil, climate, altimetry, and satellite imagery. This model, utilizing deep learning through Artificial Neural Networks, enhances the accuracy of productivity assessments and provides more precise coverage by considering regional agricultural conditions. Variables such as geographic location, soil type, climate data, and vegetation indices were used to improve the productivity estimates. The information generated by the model was used to create a pricing methodology that calibrated the final price using classical statistical and actuarial methods. A pilot project involving approximately 300 crop areas demonstrated an 89% accuracy in estimating past productivity, improving pricing models and reducing the loss ratio by about 75%. Ultimately, the proposed methodology ensured the financial health of the insurer by optimizing risk pricing and mitigating exposure.

Accurate assessment of spatial risks, particularly those influenced by climate, weather and demographic factors, is crucial for the insurance industry to enhance underwriting precision and risk management. In this talk, we present a novel approach to constructing spatial embeddings of geographic locations by integrating satellite imagery and OpenStreetMap (OSM) data aggregated into hexagonal grids. Leveraging contrastive learning techniques, we generate embeddings that capture complex spatial features related to insurance risks. We propose and train a model that leverages these embeddings to analyze the distribution of spatial risks with specific case studies in home and automobile insurance. The results demonstrate that our method effectively identifies risk patterns and offers superior predictive capabilities compared to traditional models. This study highlights the potential of advanced spatial embeddings and novel geographic data sources to improve risk assessment and decision-making processes within the insurance sector.

The use of data to provide better decisions has been paramount in modern business world. Artificial Intelligence (AI), Predictive Analytics, Big Data and Digital Transformation are topics that gained relevance in the recent decades, and insurance and reinsurance companies can leverage these new technologies to guarantee their sustainability as providers of risk coverages. Although the importance of the topic is clear, there is a gap to be filled regarding how these technologies are built in the companies, beginning with the issue of how the vast amount of data is treated. This work aims to describe a practical tool that can be used by actuaries, data scientists and data engineers to help their companies in their Digital Transformation journey. Departing from the common sense that good models are as good as the data that feeds them, this text addresses this issue by demonstrating an application of a Data Lineage process to a P&C insurance company. The Data Lineage framework is a systematic approach that traces data from its origin to its destination, ensuring proper treatment and consistency across different business functions. This process brings significant gains to the company’s data life cycle, setting the fundamentals for more sophisticated models and analysis.

Catastrophe models and climate models can be combined to enhance flood risk management by providing both short-term and long-term insights. Catastrophe models provide an assessment of today’s risk by simulating the financial and physical impacts of specific flood events based on current hazard, exposure, and vulnerability data, enabling insurers, policymakers, and emergency planners to prepare for and respond to immediate threats. Climate models, on the other hand, work with longer time horizons, projecting future changes in precipitation patterns, sea levels, and extreme weather events due to climate change. By integrating these models, stakeholders can assess how flood risks may evolve over time and develop adaptive strategies, such as improved infrastructure, land-use planning, and risk-based insurance pricing, to mitigate the increasing threat of floods. This session will present how a probabilistic Brazil flood model and a global climate risk tool can be used to improve the flood risk management in Brazil.

This study applies the HAND (Height Above the Nearest Drainage) model to assess flood susceptibility in the state of São Paulo, incorporating data on soil types and regional structural characteristics. The HAND model, widely used to identify areas prone to water accumulation during extreme precipitation events, estimates the relative height of the terrain in relation to the nearest drainage level. By integrating variables such as soil type and infrastructure, the risk analysis becomes more refined, as the water retention and drainage capacity of different soil types, along with the resilience of local infrastructure, are crucial factors in determining flood vulnerability. For the insurance sector, this approach offers a strategic advantage. Understanding the susceptibility of specific areas allows insurers to adjust policy pricing and coverage conditions based on detailed risk mapping, improving the accuracy of claims forecasting and optimizing portfolio management. Additionally, by identifying areas requiring investment in drainage infrastructure, insurers can collaborate with public and private sectors to reduce the risk of large-scale claims, thus supporting long-term financial sustainability.

In recent decades, the impact of climate change has intensified globally, with extreme weather events like floods, droughts, and storms becoming more frequent and severe. In Brazil, these events have led to significant social and economic repercussions, notably affecting the insurance sector due to increased risks to property, agriculture, and vehicles. Many of these occurrences are now classified as catastrophic, despite Brazil's historical reputation as a non-catastrophic region. Traditionally, insurance pricing relies on statistical and probabilistic models to estimate the frequency and severity of potential losses. However, the rise in climate-related events poses challenges to these models, especially in Brazil, with limited historical data on such extreme occurrences. This article examines how the Brazilian insurance market is adapting its pricing strategies to account for these catastrophic events and explores how other countries, with more extensive historical data (though sometimes less applicable), are considering alternative pricing models that could also be suitable for Brazil’s evolving climate risks.

What are the levers of attractiveness of long-term care insurance that we need to act on? In France, the creation of a 5th branch of Social Security dedicated to autonomy is a significant step forward. However public finances will not be able to assume the costs linked to the loss of autonomy of all our fellow citizens and families will not have all the financial resources to absorb the remaining expenses of their elders. These issues undeniably argue for the use of insurance solutions to support the public authorities. It is therefore essential that we, as actuaries, continue to work on making our long-term care products more desirable and durable. This workshop will be an opportunity to present an overview of the long-term care insurance market, highlighting in particular the issues, practices, market projects and prospects specific to this risk.

Taxation, regional regulations and certain exogenous factors might affect a plan sponsor’s interpretation, approaches and success in achieving sustainability of their defined benefit plan. Rules regarding the design, funding and taxation of most defined benefit plans are regional, typically by country, or perhaps by state or province. The rules usually focus on encouraging sponsorship and participation, and/or ensuring sufficient funding. There are often other goals, such as limiting tax deductions or preventing discrimination by age, gender, pay-level, etc. While they may be well-intended, the rules can often constrain a sponsor’s ability to implement effective, long-term policies that seek to optimize plan sustainability. Layered on top of the general rules are often tax laws that can influence or reward sponsor and participant actions. These incentives, or sometimes disincentives, can lead to sponsor and participant choices that might be counter to a plan’s optimal path to sustainability. In addition, an organization’s approach to plan sustainability could be constrained by exogenous factors, such as: prioritization of short-term financial results diverting from long-term funding; demographic aging if benefit costs (intentionally or unintentionally) rely on intergenerational cross-subsidies; long-term global trends affecting capital market returns, long-term return expectations or inflation experience; and mortality improvements. The paper will examine how these constraints can affect a plan’s sustainability, how a sponsoring organization might better achieve sustainability if unconstrained, and a case study of the United Nations Joint Staff Pension Fund, which operates free of certain constraints that exist for plans operating under regional regulations and/or taxation regimes.