| Titre : | Stochastic Maximum Principle Applied to an Insurance Firm |
| Auteurs : | Aya CHEKALBI, Auteur ; Nabil khelfallah, Directeur de thèse |
| Type de document : | Mémoire magistere |
| Editeur : | Biskra [Algérie] : Faculté des Sciences Exactes et des Sciences de la Nature et de la Vie, Université Mohamed Khider, 2026 |
| Format : | 1vol.(44.p) / 30 CM |
| Langues: | Anglais |
| Mots-clés: | Stochastic differential equations, backward SDEs, stochastic maximum principle, premium policy, partial information, Kalman filter, Riccati equation, insurance control. |
| Résumé : |
This thesis investigates the optimal premium policy of an insurance firm under both full
and partial information, using the tools of stochastic analysis and stochastic optimal control. The first chapter provides a self-contained introduction to the essential pillars of stochastic calculus, Brownian motion, Itˆo integration, Itˆo’s formula, stochastic differential equations (SDEs), backward SDEs (BSDEs). All fundamental results are stated with precise hypotheses and proved or referenced in detail. The second chapter applies these tools to a linear-quadratic control problem with a terminal expectation constraint, modeling the firm’s cash reserve as a controlled Ornstein- Uhlenbeck process and forward-backward systems (FBSDEs), and the stochastic maximum principle. Under full information (direct observation of the reserve), the stochastic maximum principle yields a coupled FBSDE. A Riccati decoupling leads to an explicit feedback control u∗t = −N−1 t e?t(tXt + t), where t solves a Riccati ODE and t a linear ODE. The Lagrange multiplier enforcing E[XT ] = c0 is determined explicitly, and the optimal cost is given in closed form. Under partial information, the reserve is observed through a noisy signal. A decomposition of the state and observation processes, together with Kalman filtering, reduces the problem to an equivalent full-information problem for the conditional expectation ˆXt. The candidate control ˆ?t = −N−1 t e?t(t ˆXt + t) is shown to be optimal via a verification argument based on Itˆo’s formula. The optimal cost includes an additional term representing the cost of imperfect observation. Explicit solutions and closed-form cost expressions are provided for both information structures. The results extend the existing literature on insurance control by incorporating premium adjustment, terminal solvency constraints, and partial observability. |
| Sommaire : |
Dedicace i Acknowledgements ii Abstract (Arabic) iii Abstract iv R´esum´e v Table of Contents v Introduction 1 1 Foundations of Stochastic Analysis 6 1.1 Measure-Theoretic Probability . . . . . . . . . . . . . . . . . . . . . . . . . 6 1.2 Filtrations and Stochastic Processes . . . . . . . . . . . . . . . . . . . . . . 8 1.3 Brownian Motion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 1.4 Martingales . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 1.5 Itˆo Integral . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 1.6 Quadratic Variation and Itˆo’s Formula . . . . . . . . . . . . . . . . . . . . 12 1.7 Stochastic Differential Equations . . . . . . . . . . . . . . . . . . . . . . . 13 1.8 Ornstein-Uhlenbeck Process . . . . . . . . . . . . . . . . . . . . . . . . . . 14 1.9 Backward Stochastic Differential Equations . . . . . . . . . . . . . . . . . . 15 2 Full Information Optimal Premium Policy 17 2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 2.2 Forward-Backward Stochastic Differential Equations . . . . . . . . . . . . . 17 2.3 Stochastic Optimal Control . . . . . . . . . . . . . . . . . . . . . . . . . . 18 2.4 Stochastic Maximum Principle . . . . . . . . . . . . . . . . . . . . . . . . . 19 2.5 Problem Formulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 2.6 Reformulation as a Recursive Optimal Control Problem . . . . . . . . . . . 22 2.7 Stochastic Maximum Principle . . . . . . . . . . . . . . . . . . . . . . . . . 22 2.8 Decoupling via Riccati Equations . . . . . . . . . . . . . . . . . . . . . . . 23 2.9 Determination of the Lagrange Multiplier ? . . . . . . . . . . . . . . . . . . 25 2.10 Optimal Cost Functional . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 vi 3 Partial Information Optimal Premium Policy 32 3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 3.2 Problem Formulation and Decoupling . . . . . . . . . . . . . . . . . . . . . 33 3.3 Filtering of the State . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 3.4 Candidate Optimal Policy and Verification . . . . . . . . . . . . . . . . . . 36 General Conclusion 40 Bibliographie 41 Abbreviations and Notations 44 vii |
Disponibilité (1)
| Cote | Support | Localisation | Statut |
|---|---|---|---|
| MM/1431 | Mémoire master | bibliothèque sciences exactes | Consultable |




