Is cross-sectional uncertainty acyclical?
In this paper, Dew-Becker and Giglio construct a novel option-implied measure of cross-sectional uncertainty. The measure is defined as the average firm-level option-implied conditional variance minus market implied conditional variance. Under general conditions, that gap measures the average variance of the residual from a regression of each stock’s return on that of the market. Because it is constructed from market prices, the measure is forward-looking, and is available continuously, in real time, making it particularly useful for policymakers. In addition, the measure is available for a long span of time (40 years), including six recessions. Their findings include:
- Cross-sectional uncertainty is approximately acyclical overall and has little unconditional ability to forecast changes in future real activity. Overall, sometimes the data is consistent with models in which uncertainty shocks have causal negative effects on the economy, while in other periods it is consistent with models in which cross-sectional uncertainty is high following good shocks;
- When comparing cross-sectional uncertainty in Switzerland, Germany, France, Great Britain, the Netherlands, and the constituents of the Euro Stoxx 50 index to the one in the U.S., they find a strong correlation even though the path of aggregate output in Europe over the period of 01/2002 to 5/2020 was very different from the US (a number of these countries went into recessions around 2012).
Cross-Sectional Uncertainty and the Business Cycle: Evidence from 40 Years of Options Data
Authors: Ian Dew-Becker, Stefano Giglio
From: Northwestern University, Yale University
