Commodity Futures Term Structure and Single-Commodity ETP Returns : Evidence from 2015–2025
Year: 2026
Type: dissertation
Abstract: This thesis examines whether the term structure of commodity futures helps explain the returns of futuresbased single-commodity exchange-traded products (ETPs) and whether this relationship changed during the COVID-19 period. The empirical analysis is based on weekly panel data for 16 non-leveraged singlecommodity ETPs from the energy, metals, and agriculture sectors over the period from 2015 to 2025. The data is obtained from LSEG Workspace. The main explanatory variable is a commodity-specific measure of futures term structure constructed from front-month and second-month futures prices. Product returns are analyzed using fixed-effects panel regressions with product-level clustered standard errors, complemented by robustness checks using alternative covariance estimators and week fixed effects. The results show that, in the full sample baseline setting, a steeper contango is associated with lower ETP returns. This indicates that futures curve conditions are relevant for explaining the returns of futures-based commodity ETPs beyond movements in the underlying commodity price alone. However, the evidence does not support the view that the negative contango effect became uniformly stronger during the COVID-19 period. Instead, the relationship varied across pandemic subperiods: it weakened after the initial shock, while the recovery-period estimate in the full sample became positive but remained only weakly significant. Subsample and robustness analyses further show that the stronger time variation is driven mainly by energy products, whereas the baseline negative relationship is not limited to the energy sector. The findings imply that investors in futures-based commodity ETPs should consider not only expected commodity price movements but also the slope of the futures curve, especially in energy-related products. Future research could extend the analysis by using larger product samples, more product-specific roll measures, and alternative approaches to modeling common time effects.
Source: UTUPub (University of Turku)
Author: Nieminen, Niki
Language: English
Subfield: Finance
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W7220749365