Modelling Regime Shifts in Insurance Dynamics: A Latent Markov Approach with Macroeconomic Covariates

The European Life insurance market has recently faced challenges mainly derived from macroeconomic shocks, notably the 2022-2023 inflationary spike and rapid interest rate hikes. While traditional literature often examines these shocks using static panel regressions, this thesis introduces a dynamic Latent Markov Model (LMM) approach to capture unobserved heterogeneity and regime shifts within the sector. Utilizing a comprehensive quarterly panel dataset of 29 European countries from 2017 to 2025, the empirical analysis identifies a fundamental dual-state market structure (\mathbit{k}=\mathbf{2}): a "Resilient" core and a "Volatile/Liquidity Stress" regime. The baseline model reveals a significant structural break in 2023-2024, characterized by a mass migration of historically stable markets (e.g., Italy, Germany) into the volatile state due to severe liquidity pressures. By incorporating macroeconomic and socio-demographic covariates directly into the measurement model, this study isolates the exogenous drivers of this shift, demonstrating that the observed volatility was not a symptom of chronic operational inefficiency, but a mechanical reaction to sudden surrender risks driven by rising 10-Year Government Bond yields. Ultimately, the estimated transition probabilities highlight the sector's robust internal resilience, showing a higher probability of recovery (19.22%) than of falling into distress (9.07%). The findings provide critical policy implications for regulatory bodies and insurance management, emphasizing the need for dynamic liquidity risk monitoring.