The Cross-Section of Asset Synchronicity by Elastic-Net
83 Pages Posted: 1 Jan 2019
Date Written: December 13, 2018
How does one hedge factor risks without knowing the identities of the factors? We first prove a general theoretical result: even if the exact set of factors cannot be identified by the econometrician, any risky asset can use some portfolio of other similar risky assets to insure against its own factor exposures. A long position of a given risky asset and a short position of this portfolio represents this asset’s residual factor risks. We coin the expected return of this long-short position as an asset insurance premium. To empirically construct this portfolio, we regress a given stock’s return onto the returns of thousands of all other stocks using the elastic-net estimator, a machine learning method. We coin the regression R-squared as asset synchronicity. Unique stocks earn a higher return than ubiquitous stocks: in the cross-section, value-weighted stocks that are least (most) synchronized with all other stocks earn an asset insurance premium of 0.976% (0.305%) per month. The unconditional value-weighted asset insurance premium is positive and economically large at 0.575% per month. Asset synchronicity is countercyclical, where a 1% monthly change in macroeconomic consumption shocks is associated with a -1.725% change in the cross-sectional asset insurance premium. The unconditional and cross-sectional existence of the asset insurance premium is robust to equal and value portfolio weighting schemes, and to the effects of value, size, idiosyncratic volatility and illiquidity measures.
Keywords: Cross-section, elastic-net, machine learning, asset synchronicity, asset insurance premium, macrofinance, portfolio construction
JEL Classification: G11, G12, C55
Suggested Citation: Suggested Citation