The Academic Case: Gorton and Rouwenhorst
The modern institutional case for commodities as a strategic portfolio allocation was largely built on one paper: "Facts and Fantasies About Commodity Futures" by Gary Gorton (Wharton School, University of Pennsylvania) and K. Geert Rouwenhorst (Yale School of Management), published in 2006 by the Yale International Center for Finance.
Examining a diversified portfolio of commodity futures from 1959 to 2004, Gorton and Rouwenhorst found three properties that made commodities appear to be an unusually attractive diversifier:
Equity-like returns over the full period — commodity futures matched the total return of equities without requiring the same risk premium.
Negative correlation to stocks and bonds — particularly during inflationary recessions, the scenario most damaging to a traditional 60/40 portfolio. When equities fell and bonds were under pressure from rising rates, commodity futures tended to hold value or appreciate.
Positive correlation to inflation — both expected and unexpected. This was the critical property: commodities performed best in the precise economic environments where financial assets struggled most.
The paper's findings underpinned a significant shift in institutional portfolio construction in the mid-2000s, with pension funds, endowments, and sovereign wealth funds adding commodity index allocations as a strategic hedge against inflation and a diversifier away from correlated financial assets.
The Post-Publication Challenge
Photo by Anne Nygård on Unsplash
The Gorton-Rouwenhorst paper covers 1959 to 2004. What happened next proved to be a significant complication.
Following the paper's publication and the broader institutionalisation of commodity index investing, large capital flows poured into commodity index products. The value of assets tracking commodity indices grew from tens of billions to hundreds of billions of dollars between 2005 and 2012. Michael Masters, a hedge fund manager, testified before the US Senate in 2008 that this "index speculator" demand was itself a contributor to the 2007-2008 commodity price spike.
More fundamentally, subsequent academic research — most notably Claude Erb and Campbell Harvey's 2013 paper "The Golden Dilemma" and related work — found that the returns from commodity futures over the 2005-2013 period were substantially weaker than the historical record suggested. Commodity futures as a diversified asset class delivered poor nominal returns during a period when equities performed well, and the inflation-hedging properties — while still present in principle — were less consistent in practice.
The Financialisation Problem
The concept of financialisation captures the key mechanism by which the commodity opportunity may have deteriorated. The Gorton-Rouwenhorst paper documented that the positive returns to commodity futures over 1959-2004 came substantially from roll yield in backwardated markets — a return generated because commodity producers structurally needed to hedge (Keynes' normal backwardation theory).
As massive institutional capital entered commodity futures markets post-2005, the structure of commodity markets changed. Large sustained buying of commodity futures pushed near-month contracts up relative to deferred contracts, reducing and in many cases eliminating the backwardation that had historically generated positive roll yield. In energy markets, persistent contango became the norm for extended periods, imposing a negative roll yield on index investors. The structural return source documented by Gorton and Rouwenhorst was partially dissipated by the capital flows that their research had encouraged.
This is a classic example of a strategy that works at small scale (when few investors are pursuing it) but becomes self-defeating when pursued at large scale — a concept related to the limits of arbitrage in financial economics.
The Current Consensus
Photo by Scottsdale Mint on Unsplash
The contemporary investment consensus on commodities is more measured than the early 2000s enthusiasm suggested:
Inflation hedge: yes, but imperfect. Commodities remain one of the most direct inflation hedges available in a liquid investable form. When inflation is driven by commodity prices — as in 2021-2022 — commodity exposure can provide meaningful protection. However, not all inflation is commodity-driven, and the hedge is imperfect over shorter time horizons.
Return enhancer: probably not, at scale. The equity-like return documented over 1959-2004 has not been reliably replicated in subsequent decades. Current consensus treats commodity allocation as a risk management tool rather than a return source.
Strategic allocation size: 5-10%. Most institutional investors who retain commodity exposure treat it as a modest strategic allocation — large enough to provide meaningful inflation hedging and diversification in stress scenarios, small enough that its underperformance in non-inflationary environments does not materially drag portfolio returns.
Implementation matters enormously. Given the impact of roll yield and financialisation on index returns, how commodity exposure is implemented — which index, which roll methodology, whether through futures or commodity equities — can be as important as the allocation size decision itself.