Download this Paper Open PDF in Browser

Testing Heterogeneous-Agent Models: An Alternative Aggregation Approach

54 Pages Posted: 1 Nov 2006  

Pierluigi Balduzzi

Boston College - Carroll School of Management

Tong Yao

University of Iowa - Henry B. Tippie College of Business

Abstract

Starting from the cross-sectional aggregation of marginal utilities, rather than intertemporal marginal rates of substitution, this paper develops a new heterogeneous-agent pricing kernel. A closed-form version of the new kernel depends on changes in the cross-sectional variance of log consumption, rather than the variance of changes in log consumption, as in Constantinides and Duffie (1996). We implement the new kernel on household consumption data from the Consumer Expenditure Survey (CEX). At monthly and quarterly horizons, the new kernel reconciles the premium on U.S. equities with the consumption of asset holders, for reasonable values of relative risk aversion. The new kernel also fares better than kernels based on the aggregation of intertemporal marginal rates of substitution in explaining the cross-sectional variation of risk premia on stocks and bonds. Portfolios mimicking changes in the cross-sectional variance of log consumption for asset holders command significant negative Sharpe ratios.

Keywords: pricing kernel, heterogeneity

JEL Classification: G12

Suggested Citation

Balduzzi, Pierluigi and Yao, Tong, Testing Heterogeneous-Agent Models: An Alternative Aggregation Approach. Journal of Monetary Economics, Forthcoming. Available at SSRN: https://ssrn.com/abstract=941134

Pierluigi Balduzzi (Contact Author)

Boston College - Carroll School of Management ( email )

Department of Finance
140 Commonwealth Avenue - Fulton Hall 438
Chestnut Hill, MA 02467
United States
617-552-3976 (Phone)
617-552-0431 (Fax)

HOME PAGE: http://www.bc.edu/bc_org/avp/csom/faculty/

Tong Yao

University of Iowa - Henry B. Tippie College of Business ( email )

Acquisitions
5020 Main Library
Iowa City, IA 52242-1000
United States

Paper statistics

Downloads
360
Rank
68,613
Abstract Views
2,551