From macro to micro: Sparse macroeconomic risks and the cross-section of stock returns

Abstract

Our study presents novel evidence on the pricing effectiveness of macroeconomic risks at the firm level. We employ a sparse PCA approach to aggregate macroeconomic variables from the FRED-MD database and obtain the first eight components, namely inflation, housing, spreads, production, employment, personal income, yields, and credit. We then construct the corresponding firm-level macroeconomic risk exposures as the sensitivity of the individual stock returns to the sparse economic components. Our research yields three main findings: (i) macro betas cannot be fully captured by common firm fundamentals, indicating the unique information in macro betas; (ii) the betas of inflation, production, personal income, yields, and credit are strong predictors of future stock returns beyond other micro risks; and (iii) behavioral mispricing theory, especially arbitrage frictions and investor sentiment, can help explain the macro beta premium.