We show that decomposing macroeconomic risks across horizon is key to uncover a tight link between risk premia and the real economy. Exposure in four-year returns to fluctuations in macroeconomic growth (volatility) with a half-life larger than four years captures a large
and significant price of risk of 2.5% (-1.6%) per cross-sectional standard deviation in historical, stock-level exposure. Shorter-term risk is not priced. Our single long-term macro risk factor is
robustly priced in a wide variety of stock and bond portfolios, and obtains a cross-sectional fit similar to the three-factor model of Fama and French (1993). What separates our work from
the long-run risk model of Bansal and Yaron (2004) is the focus on exposure in long-term rather than short-term returns. This separation is important, as our long-term macro risk factor drives out standard measures of long-run risks in cross-sectional tests. Finally, long-term growth and volatility share a large common component of risk.