Optimal Plan Design and Dynamic Asset Allocation of Defined Contribution Pension Plans: Lessons from Behavioural Finance and Non-expected Utility Theories

The question of optimal asset allocation strategy for defined contribution (DC) pension plans is addressed. A primary motivation for this study is provided by the recent literature on behavioural finance and intertemporal life-cycle investment theory. In this thesis two alternative utility forms are considered: loss aversion and Epstein-Zin recursive utility. We develop a dynamic-programming-based numerical model with uninsurable stochastic labour income and borrowing constraints. In the loss aversion case, members are assumed to be loss averse with a target replacement ratio at retirement and a series of suitably defined interim target prior to retirement. We also extend the intertemporal life-cycle saving and investment theory to the dynamic asset allocation problem of DC pension schemes. A new approach to model contribution and investment decisions with focus on the member’s desired pattern of consumption over the lifetime (based on Epstein-Zin utility preference) is proposed. The thesis draws on empirical evidence of salary scales and loss aversion parameters from UK households, with labour income progress estimated from the New Earnings Survey and loss aversion parameters estimated on the basis of face-to-face interviews with 966 randomly selected UK residents.

Name dc:type.qualificationname

phd

Level dc:type.qualificationlevel

doctoral

Grantor dc:publisher.institution

City University London

Year dc:date.issued

2009

Author dc:creator

• Zhang, Y.

• HB Economic Theory