Inflation dans l'estimation des provisions techniques en soins de santé particuliers

Details

Supervisors

Courtois, Cindy ; Lohest, Alizee

Faculty

Faculté des sciences

Degree label

Master [120] en sciences actuarielles, à finalité spécialisée

Abstract

enThe assessment of inflation is a key component in the calculation of technical provisions for supplementary health insurance, as it directly affects insurers’ ability to meet long-term liabilities in the face of rising medical costs. In a context marked by persistent inflationary pressure, driven by demographic trends and technological advancements in healthcare, rigorous treatment of this factor is essential to ensuring the financial soundness of insurance undertakings. The Solvency II regulatory framework strictly governs these requirements, notably by mandating that Best Estimate (BE) provisions be based on prudent and well-justified economic assumptions, which explicitly account for risks associated with medical inflation. Accordingly, the adoption of modelling approaches that can capture the uncertainty and volatility of this variable has become a methodological imperative. This thesis contributes to that effort by proposing an empirical approach to the projection and calibration of medical inflation for the computation of Best Estimate premiums, within the context of a Belgian supplementary health insurance portfolio. It explores several complementary actuarial techniques designed to capture the multifaceted nature of inflationary dynamics. Particular attention is paid to the Smith-Wilson method, which is employed to extrapolate long-term interest rate curves in line with regulatory requirements concerning extended projection horizons. Three inflation projection approaches are developed: two based on standard hypotheses corresponding to the horizontal analysis (AH) and the Solvency II framework (SII); and a third, more stochastic, based on drift-based time series models specifically, a random walk with drift and an ARIMA(1,1,1) model with drift calibrated on Belgian medical inflation series disaggregated by hospital room type. These approaches are applied to Belgian market data, with a distinction made between policy benefits according to pricing structure and duration of coverage. The resulting Best Estimate provisions under each modelling scenario are compared, enabling an assessment of sensitivity to various inflationary dynamics over both short- and long-term horizons. This is further enhanced by a comprehensive sensitivity analysis, which quantifies the impact of inflation parameters on provision estimates and identifies relevant prudential levers for the forward-looking management of medical inflation risk.