Athens University Economics and Business (AUEB)
Stochastic linear quadratic control problems with applications to insurance
Abstract
dc:descriptionΤhis thesis aims to extend further the modelling and decision-making (control) methodstechniques for the deeper investigation of several important actuarial (practical) problems. The major drive in this task is the application of stochastic linear-quadratic control methodologies into several actuarial processes. In more details, a significant contribution of this PhD thesis is the further understanding of the role of (stochastic) control theory in the specification of conceptual actuarial models and subsequently the transition towards the development of more formal dynamic models. A quite general discrete and continuous-time (stochastic) control framework, suitable for discussing a number of important issues arising in the analysis and the design of actuarial processes, is introduced. It will be described as a mean of integrating diverse areas where mainly linear system concepts are used. A general approach towards the definition of linear stochastic systems will also be given. Analytically, a) A discrete-time stochastic control multi-dimensional model for a quasi PAYGO social security system is introduced to allow the potential accumulation of a special (contingency) fund, which can oscillate deliberately absorbing fluctuations in the different system parameters involved. b) A more realistic approach for the management of a defined contribution pension scheme is proposed in the distribution phase (post-retirement period) providing a whole life assurance benefit. In that direction, it is assumed a stochastic framework for both mortality and investment risk and additionally suggested a correlation effect between the two separate risks. c) A stochastic control model is introduced for a pension fund which provides a variable death benefit to its members during the post-retirement period. The main framework model is described by two correlated fractional Brownian motions which correspond to investment and mortality risks, accordingly. d) The classical problem of premium rating within a heterogeneous portfolio of risks is investigated by using a continuous stochastic framework. The portfolio is divided into several classes where each class interacts with other. The risks are modelled dynamically by the means of a standard Brownian motion. This dynamic approach is also transferred to the design of the premium process. e) Finally, in the last work, the various claim reserving methods are discussed for non-life insurance systems. Furthermore, a dynamic control model with the associated stochastic differential equations is analytically designed describing the mechanisms of the payments and the reserving process, as well.
Degree
thesis:*- Grantor dc:publisher
- Athens University Economics and Business (AUEB)
- Year dc:date
- 2008
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Παντελούς, Αθανάσιος
- Pantelous, Athanasios
Subjects
dc:subject × 12Rights
- Language dc:language
- eng
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/17561