We investigate drivers of commodity futures returns using mixed-frequency vector au-toregression. Slowing real economic activity and increasing macroeconomic uncertainty precede negative monthly returns. Stock markets predict daily but not long-term commodity returns. Information from these drivers yields significant trading profits, though predictability changes with financialization periods. In recent years, futures prices show less sensitivity to financial variables. Our findings indicate that financial hedgers (com-modity index investors) enhance the predictive power of financial variables but reduce the informativeness of fundamental information for future commodity returns.