Commodities for the Long Run

Working Paper: NBER ID: w22793

Authors: Ari Levine; Yao Hua Ooi; Matthew Richardson

Abstract: This paper analyzes a novel data set of commodity futures prices over a long sample period starting in 1877, which allows us to shed new light on several important and controversial questions. We document that commodity futures returns (1) have been positive on average; (2) vary significantly across business cycles, inflation episodes, and periods of backwardation versus contango, (3) are driven mostly by variation of spot returns and therefore closely linked to the underlying commodity spot market; (4) perform well during inflation cycles and provide more return in backwardated states; and (5) display low correlation with stocks and bonds. These long-run stylized facts imply that commodity futures can add value to a diversified portfolio from an asset allocation perspective.

Keywords: Commodity Futures; Asset Allocation; Inflation; Business Cycles

JEL Codes: G10; G11; G13; N2; N21; N22


Causal Claims Network Graph

Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.


Causal Claims

CauseEffect
Commodity futures returns (G13)Inflation episodes (E31)
Inflation episodes (E31)Commodity futures returns (G13)
Economic conditions (E66)Commodity futures returns (G13)
Backwardation (G13)Commodity futures returns (G13)
Contango (G13)Commodity futures returns (G13)
Spot prices (G13)Commodity futures returns (G13)
Macroeconomic factors (E66)Commodity futures returns (G13)
Asset allocation decisions (G11)Commodity futures performance (G13)

Back to index