Labor Market Networks and Asset Returns

Abstract

This paper proposes a measure of labor market connectivity based on the similarity in the composition of occupational knowledge characteristics across industries and provides evidence of return predictability in the cross-section of industries that are connected through the labor market. In long-short portfolios, an industry's return is strongly predicted by the past return of its labor-market-connected neighboring industries with an annualized return of up to 9%, which is not explained by established asset pricing models. The return predictability remains significant after controlling for the supply chain momentum as well as the industry lead-lag effect, and is concentrated in stocks with higher arbitrage costs and higher ownership of uninformed investors. We find similar predictive relations for the labor productivity, wages, employment, and profitability of labor connected industries. Our findings are consistent with positive spillover of productivity shocks among industries that are connected through the labor market. Informational frictions, costly arbitrage, and investors' limited attention magnify the delayed response of stock prices to the spillover of labor productivity shocks, which results in the observed return predictability.