Working Paper: NBER ID: w13220
Authors: John Donaldson; Rajnish Mehra
Abstract: This essay reviews the family of models that seek to provide aggregate risk based explanations for the empirically observed equity premium. Theories based on non-expected utility preference structures, limited financial market participation, model uncertainty and the small probability of enormous losses are detailed. We impose the additional requirements that candidate models yield consistent inter temporal portfolio choice and that a representative agent can be constructed which is independent of the underlying heterogeneous economy's initial wealth distribution. While many models are able to replicate a wide variety of financial statistics including the premium, few satisfy these latter criteria as well.
Keywords: equity premium; risk-based explanations; asset pricing; preference structures
JEL Codes: D10; D11; D50; D52; D90; D91; E30; G00; G11; G12
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| non-expected utility preferences (D81) | equity premium (G12) |
| limited financial market participation (G19) | equity premium (G12) |
| increasing the coefficient of relative risk aversion (D11) | equity premium (G12) |
| preference-based theories (D11) | equity premium (G12) |
| trading frictions (F12) | equity premium (G12) |
| market incompleteness (D52) | equity premium (G12) |
| modifications to standard neoclassical representative agent paradigm (E19) | equity premium (G12) |