Working Paper: NBER ID: w27864
Authors: Ian Dewbecker; Stefano Giglio
Abstract: This paper presents a novel and unique measure of cross-sectional uncertainty constructed from stock options on individual firms. Cross-sectional uncertainty varied little between 1980 and 1995, and subsequently had three distinct peaks -- during the tech boom, the financial crisis, and the coronavirus epidemic. Cross-sectional uncertainty has had a mixed relationship with overall economic activity, and aggregate uncertainty is much more powerful for forecasting aggregate growth. The data and moments can be used to calibrate and test structural models of the effects of uncertainty shocks. In international data, we find similar dynamics and a strong common factor in cross-sectional uncertainty. The data is available on our websites. A companion paper [Dew-Becker and Giglio, "Real-time forward-looking skewness over the business cycle"] finds firm-level skewness is significantly procyclical.
Keywords: cross-sectional uncertainty; business cycle; options data; forecasting; aggregate uncertainty
JEL Codes: C58; D81; D84; E22; E30; E32; E37; G13
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| cross-sectional uncertainty (D89) | economic activity (E20) |
| aggregate uncertainty (E10) | economic activity (E20) |
| aggregate uncertainty (E10) | declines in industrial production (L16) |
| aggregate uncertainty (E10) | declines in employment (J63) |
| realized dispersion (D39) | real activity (E23) |
| cross-sectional uncertainty (D89) | aggregate uncertainty (E10) |