By Roger Farmer
The representative agent model (RA) has dominated macroeconomics for the last thirty years. This model does a reasonably good job of explaining the co-movements of consumption, investment, GDP and employment during normal times. But it cannot easily explain movements in asset prices. Two facts are hard to understand 1) The return to equity is highly volatile and 2) The premium for holding equity, over a safe government bond, is large. This paper constructs a lifecycle model in which agents of different generations have different savings rates and different attitudes to risk and I use this model to account for both a high equity premium and a volatile stochastic discount factor. The model is persuasive, precisely because it explains so much with so few parameters, each of which is pinned down by a few simple facts.
While I do not think the model is as simple and has as few degrees of freedom as Roger Farmer makes it appear, it is quite powerful in resolving the excessive volatility and equity premium puzzles. We have learned that the life cycle matters for many economic issues, and this seems to be another one. While the life cycle is modeled in a very crude way here, it would interesting to see whether the quantitative results would still hold in a model that tracks life events with more granularity.