A quantitative model for the asset liability management of a Pension Fund

The key objective of pension plans is the delivery of retirement benefits, typically payable for life or a set period of time, to the specified group of recipients. The management of such funds entails therefore a constant monitoring of the risks exposure and a regular rebalancing of assets. This thesis is directly related to these topics and proposes a quantitative method (mainly based on stochastic optimal control theory) to determine the optimal investment policy of a pension fund’s wealth, under financial and actuarial risks. The thesis unfolds as follows: Chapter 1 includes a basic introduction to pension systems. The topics addressed here are: how to classify pension systems, the main characteristics of each kind of system, examples of major systems and the important reforms that have been implemented in Italy; the chapter ends with a description of Italian professional order pension funds. Chapter 2 describes asset liability management techniques in pension schemes; it contains a review on major literature on asset liability management and a discussion on interested parties in this topic and on policies and instruments which can be adopted. Chapter 3 contains an original model to determine the optimal financial investment policy in a pension fund, considering both financialand actuarial risk. Moreover, the model takes care of the pension plan’s sustainability, i.e. of the balance between the active and retired members. Chapter 4 is a numerical application of the model described in the previous chapter to a real Italian pension fund. Finally, in Chapter 5 conclusions are drawn related to the question asked.

Grantor dc:publisher

Università degli Studi di Cagliari

Year dc:date

2011

Author dc:creator

• CANNAS, GIUSEPPINA

• Dynamic asset allocation
• Pension fund
• Sthocastic optimal control
• Settore SECS-S/06 - Metodi mat. dell'economia e Scienze Attuariali e Finanziarie

Statement dc:rights

• info:eu-repo/semantics/openAccess
• license:Non specificato

Language dc:language

eng

Handle dc:identifier

http://hdl.handle.net/11584/265933

OAI identifier oai:identifier

oai:iris.unica.it:11584/265933