Evidence of S-Shaped Consumption Utility

60 Pages Posted: 13 Jan 2021 Last revised: 7 Sep 2021

See all articles by Gaosheng Ju

Gaosheng Ju

China Center for Economic Studies, School of Economics, Fudan University

Qi Li

Texas A&M University

Date Written: November 14, 2020


Using a macro quantile factor model, we examine cross-state (i.e., cross-quantile) heterogeneity in consumption behaviors. We find that common macro factors generate a “big bang/crunch” effect on micro consumption. Generally speaking, when the aggregate effect of the common factors on the consumption in low consumption-growth states is negative (resp. positive), this effect in high consumption-growth states is positive (resp. negative). The big bang/crunch suggests that consumption utility functions are not concave-shaped but S-shaped. We justify the S shape using consumption-based asset pricing. We show that various consumption-based asset pricing puzzles arise from the global concavity of the utility functions, and the S-shaped consumption utility offers a solution. Also, the S-shaped utility helps explain the low elasticity of intertemporal substitution reported in the literature. In addition, we propose a simple estimation method that allows one to estimate the macro quantile factor model with big data.

Keywords: S-Shaped Consumption Utility, Common Factors, Consumption-Based Asset Pricing, Elasticity of Intertemporal Substitution, Quantile Factor Model, Big Data.

JEL Classification: D91, G12, D15, C21, C38, E21

Suggested Citation

Ju, Gaosheng and Li, Qi, Evidence of S-Shaped Consumption Utility (November 14, 2020). Available at SSRN: https://ssrn.com/abstract=3730422 or http://dx.doi.org/10.2139/ssrn.3730422

Gaosheng Ju (Contact Author)

China Center for Economic Studies, School of Economics, Fudan University ( email )

600 GuoQuan Road
Shanghai, 200433

Qi Li

Texas A&M University ( email )

7101 University Avenue
STEM 318 H
Texarkana, TX 75503
United States

Do you have a job opening that you would like to promote on SSRN?

Paper statistics

Abstract Views
PlumX Metrics