Portfolio choice and mental accounts: A comparison with traditional approaches

This paper analyses the ability of a realistic mental accounting model for portfolio choice (presented in Hübner and Lejeune (2021)) to compete with traditional utility alternatives. Das, Markowitz, Scheid, and Statman (2010) and Hübner and Lejeune (2021) have shown that the mental accounting framework embeds optimal allocations derived from quadratic utility optimization and/or the Gaussian distribution assumption. We complement their work by demonstrating here the flexibility and numerical superiority of the optimization outputs that a non-Gaussian version of HAMA generates with respect to two traditional expected utility maximization rules that produce a wide and relevant spectrum of portfolio allocation rules for a variety of realistic investor types: the decay rate approach proposed by Stutzer (2003), which is analogous to maximizing expected utility, and the class of flexible three-parameter utility functions (FTP) introduced by Conniffe (2007) that encompasses a wide set of popular utility functions.