Habit Formation in an Overlapping Generations Model with Borrowing Constraints

Abstract

We introduce habit-formation in the three-period OLG borrowing-constrained framework of Constantinides et al. (2002) by allowing the utility of the middleaged (old) to depend on consumption when young (middle-aged). This specification enables us to separate the effect of the two habit parameters (middle-aged and old) since each representative age-group can face different levels of habit persistence. The two-habit setup underlines some important issues with regards to savings and security returns which do not always conform to the standard findings in the literature. In addition, the model produces equity premium consistent with US data for relatively small levels of risk aversion.We are grateful to John Donaldson for numerous comments and suggestions during the course of this research. We are especially thankful to the editor, John Doukas, and George Constantinides, the referee, for their various helpful comments and suggestions. Finally we also want to thank Richard Clarida, Dan O'Flaherty, Rajnish Mehra, Bruce Preston, Stijn Van Nieuwerburgh, and conference participants of the 2005 Financial Management Association.

Age‐Dependent Increasing Risk Aversion and the Equity Premium Puzzle

“…There is, however, an interesting overlap between our results and the habit framework of DaSilva, Farka and Giannikos (). In their work, they examine the impact of two habit parameters (middle‐aged, and old) in the same OLG framework of CDM () and report complex interactions between the two‐habit setup and asset price dynamics.…”

“…In particular, a number of studies assume heterogeneity in risk aversion when agents have expected utility (Dumas, ; Wang, ; Basak and Cuoco, ; Kogan, Markov and Uppal, ; Weinbaum, ; Abbot, ). Others assume heterogeneity in risk aversion when agents exhibit habit‐forming preferences (Chan and Kogan, ; Xiouros and Zapatero, ; DaSilva, Farka and Giannikos, ; Bhamra and Uppal, ). A few other studies employ Epstein Zin preferences and extend preference heterogeneity along two dimensions: heterogeneity in risk aversion and elasticity of intertemporal substitution (Gomes and Michaelides, ; Guvenen, ; Gârleanu and Panageas, ).…”

“…The IRMS in the model of DaSilva, Farka and Giannikos () is given by , where is the habit parameter for middle‐aged and for old agents. Setting , this simplifies to: .…”

“…See, for example, Rietz (), Weil (), Constantinides (), Epstein and Zin (), Telmer (), Heaton and Lucas (), Basak and Cuoco (), Campbell and Cochrane (), Constantinides, Donaldson and Mehra (), McGrattan and Prescott (), Bansal and Yaron (), Barro (), Guvenen (), DaSilva, Farka and Giannikos (), Gârleanu and Panageas (), and Abbot ().…”

Asset pricing puzzles in an OLG economy with generalized preference

“…See, for example, Abel (), Constantinides (), Benartzi and Thaler (), Lauterbach and Reisman (), and DaSilva, Farka, and Giannikos ().…”