ASSET HOLDING AND CONSUMPTION VOLATILITY: Results 2

Posted by Kathryn Schwartz on June 17, 2014
ASSET HOLDING AND CONSUMPTION VOLATILITY

There are several ways in which the IMRS can be estimated or approximated. In the simplest version of the life cycle model, the marginal utility of consumption is a monotonic transformation of consumption. In more complex and possibly realistic versions of the model, the IMRS might depend in a flexible fashion on the composition of the household as well as on labor supply behavior. As a detailed characterization of preferences is not the main focus of this paper, we use a parsimonious specification. We assume that utility is an isoelastic function of non-durable consumption per adult equivalent.

We use the time series of consumption growth of the predicted group of shareholders to construct mean-variance pairs for the IMRS, assuming values for the coefficient of relative risk aversion between 0.5 and 5 and a discount rate of 2 per cent. These are shown by the ‘crosses’ in Figure 5a. Clearly — in comparison to using aggregate expenditure data — the plausible estimates of the IMRS mean-standard deviation pairs implied by the time series of consumption growth of shareholders lie much closer to the IMRS frontier implied by the observed asset returns. It is also the case that, given the smaller variance of consumption growth for the group of likely non-shareholders, we need a much larger value of у to get close to the HJ bounds for this group (see Figure 5b). These results suggest that distinguishing between shareholders and non-shareholders is crucial to resolving the equity premium puzzle.
w6567-17
w6567-18

Tags: , ,