The role of aggregate risk aversion in the pricing of economic uncertainty

ABSTRACT

We find that time-varying aggregate risk aversion is a vital state variable to the pricing of economic uncertainty documented by Bali et al. (2017). Economic uncertainty is priced only during periods with a high aggregate risk aversion level. Risk aversion-related component in economic uncertainty contributes the major proportion of uncertainty premium. This proportion fluctuates with real economy, whereas the remaining varies with market sentiment. Uncertainty premium is most prominent following periods simultaneously having high aggregate risk aversion and high economic uncertainty.

KEYWORDS:

• Risk aversion
• economic uncertainty
• asset pricing
• anomalies
• sentiment

JEL CLASSIFICATION:

• G12
• G14

Acknowledgments

We thank Qian Li, Jingdong Liu, Yong Li, Yuxi He, Chaohua He, Xianling Jiang, Xuyang Zhang and seminar participants at University of International Business and Economics and Beijing Foreign Studies University for their insightful comments. We also acknowledge the support of the Fundamental Research Funds for the Central Universities (No. 2020QD002) and the National Social Science Foundation of China (No. 21CJY070).

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

¹ Stocks whose returns covary positively with economic uncertainty underperform stocks whose returns covary negatively with economic uncertainty. Bali, Brown, and Tang (2017) find that stocks in the highest uncertainty beta decile underperform stocks in the lowest uncertainty beta decile by 0.51% per month.

² See equation 1(1) $\begin{array}{r}\\ R_{i,t}={\alpha }_{i,t}+{\beta }_{i,t}^{UNC}UN{C}_t+{\beta }_{i,t}^{MKT}MK{T}_t+{\beta }_{i,t}^{SMB}SM{B}_t\\ +{\beta }_{i,t}^{HML}HM{L}_t+{\beta }_{i,t}^{UMD}UM{D}_t+{\beta }_{i,t}^{LIQ}LI{Q}_t\\ +{\beta }_{i,t}^{R_{I/A}}{R}_{I/A,t}+{\beta }_{i,t}^{R_{ROE}}{R}_{ROE,t}+{\epsilon }_{i,m}\end{array}$(1) in Bali, Brown, and Tang (2017). Bali, Brown, and Tang (2017) do not control for aggregate risk aversion while estimating uncertainty betas.

³ http://mba.tuck.dartmouth.edu/pages/faculty/ken.french/data_library.html

⁴ http://global-q.org/index.html

⁵ http://faculty.chicagobooth.edu/lubos.pastor

⁶ https://www.sydneyludvigson.com/

⁷ https://www.nancyxu.net/risk-aversion-index

⁸ Economic uncertainty is mainly driven by news about the real economy. Aggregate risk aversion is mainly driven by news affecting investor preferences. Please refer to Bekaert, Engstrom, and Xu (2019) for a detailed discussion about the correlation between economic uncertainty and aggregate risk aversion.

⁹ Our sample starts from July 1986, because the risk aversion index is available since June 1986.

¹⁰ The full sample value-weighted return spread is −0.6% with a Newey-West t-statistic of −2.46.

¹¹ CFNAI is short for Chicago FED National Activity Index. It is widely used, also in Bali, Brown, and Tang (2017), to measure the states of the economy.

¹² Baker-Wurgler sentiment index is from https://pages.stern.nyu.edu/~jwurgler/

Additional information

Funding

This work was supported by the Fundamental Research Funds for the Central Universities [2020QD002]; National Social Science Foundation of China [21CJY070].