Title: Claim risk and carbon uncertainty in electric versus combustion vehicles: frequency, severity, and the randomness of longevity

Abstract: The growing diffusion of electric vehicles has raised important questions for insurers and sustainability assessment. Existing discussions often concentrate on either insurance risk or environmental performance separately, whereas these two dimensions are less frequently examined together. This study develops an integrated perspective on electric and combustion vehicle types through the joint consideration of claim frequency, claim severity, and carbon footprint measurement. From an actuarial point of view, the study compares insured exposures and claims experience across vehicle types while controlling for other observable characteristics such as coverage, vehicle class, brand, and age-related factors. From an environmental point of view, the study examines the limitations of deterministic carbon footprint measures and argues that uncertainty in vehicle longevity should be explicitly incorporated into the assessment of total emissions over the vehicle life cycle. Insurance portfolio data, claims records, and repair-level information provide the basis for discussing how repair patterns, replacement decisions, and vehicle durability may shape both financial losses and environmental outcomes. The study intends to contribute for the discussion on how motor insurance may account more adequately for the transition towards more sustainable mobility.

Title: On the Estimation of Reinsurance Premiums for Heavy-Tailed Losses

Abstract: Reinsurance plays a fundamental role in risk management by allowing insurers to transfer their exposure to extreme losses. In many insurance portfolios, claim sizes exhibit heavy-tailed behavior, so that extreme events have a significant impact on the overall risk. Although such events occur with low probability, their realization can lead to substantial losses, potentially threatening the solvency of the insurer or requiring significant capital reserves to meet regulatory requirements.

Extreme value theory provides a natural framework to model such phenomena and to quantify tail-related risk measures. In this context, the estimation of reinsurance premiums, particularly for high-excess loss layers, is of primary importance. However, the presence of rare and severe events raises significant statistical challenges, especially in terms of accuracy and uncertainty quantification.

In this work, we investigate different approaches for estimating reinsurance premiums under heavy-tailed assumptions. We focus on the construction and comparison of several estimation procedures, with particular attention to their performance in finite samples. We also aim to provide insights into the selection of appropriate reinsurance contract parameters. Through both theoretical considerations and empirical analysis, we seek to better understand the strengths and limitations of these methods in the presence of extreme risks.

Title: A dynamic optimal reinsurance strategy with capital injections in the Cramér-Lundberg model

Abstract: We consider the surplus process of an insurance company within the Cramér–Lundberg framework. 

We study the optimal reinsurance strategy and dividend distribution of an insurance company under proportional reinsurance, in which capital injections are allowed. Our aim is to find a general dynamic reinsurance strategy that maximises the expected discounted cumulative dividends until the time of passage below a given level, called ruin. These policies consist in stopping at the first time when the size of the overshoot below 0 exceeds a certain limit a, and only pay dividends when the reserve reaches an upper barrier b. Using analytical methods, we identify the value function as a particular solution to the associated Hamilton Jacobi Bellman equation. This approach leads to an exhaustive and explicit characterisation of optimal policy. The proportional reinsurance is given via comprehensive structure equations. Furthermore we give some examples illustrating the applicability of this method for proportional reinsurance treaties.

Title: Synthetic Data Generation: A New Paradigm for Modeling – Application to Rare Events and Imbalanced Data in the Insurance Sector

Abstract:  Data plays a pivotal role in machine learning, statistical modeling, and, more broadly, artificial intelligence. It provides the essential raw material for model construction, ranging from supervised and unsupervised learning methods to traditional statistical techniques and deep neural networks. The richness, diversity, and representativeness of the data directly determine a model's precision, generalizability, and utility across various contexts. Consequently, the quality of predictive outcomes is intrinsically linked to the quality of the underlying data. Learning from rare values—whether extreme or not—and more generally from imbalanced datasets, remains a significant and relatively under-explored challenge in regression tasks. Furthermore, these rare values often represent critical events that practitioners aim to analyze or predict. This presentation seeks to address several key questions: How can this phenomenon be identified? What are its implications for modeling? And what solutions exist to effectively tackle this issue?

Panel 2: Modeling Challenges of Emerging Risks 

Abstract: The panel on “Modeling Challenges of Emerging Risks” will bring together academic and industry experts to examine the new modeling challenges facing the insurance and risk management sector. Discussions will focus on climate risks, cyber risks, pandemics, systemic risks, and the impact of artificial intelligence and the data revolution on actuarial and financial models. 

The panel will highlight the limitations of traditional approaches in the face of increasing uncertainty, the scarcity of historical data, and the abundance of real-time data from sensors, video, satellites, and extreme weather events. Speakers will also discuss recent advances in stochastic modeling, statistical learning, risk theory, and numerical methods to improve the assessment, pricing, and management of emerging risks. 

Through the intersection of academic and industry perspectives, this session will offer forward-looking insights into the innovative tools needed to strengthen the resilience of the insurance sector in a rapidly changing environment.