Digging deeper in the commodity toolbox 

The global commodity market has grown into a multi-trillion-dollar market that, over the past decade, has expanded dramatically in both size and complexity. A market historically defined by precious metals, oil and coal, industrial metals and agricultural commodities such as soybeans and hog bellies now includes rare earths and – thanks to last year’s CLARITY Act – digital assets. 

Mutual fund and ETFs providers have responded to this growing diversification. The universe of equity and alternative commodities ETFs and mutual funds tracked by EPFR now encompasses 1,800 funds with a collective AuM totaling $1.137 trillion. In this universe, there are funds offering exposure to the new digital assets and funds dedicated to single stocks. 

What remains consistent is the fact that, from digital tokens to physical barrels and bushels, commodity markets provide plenty of investable corners. In this Quant’s Corner, we will look for new ways to unlock alpha-generating signals from EPFR’s growing coverage of commodities-mandated funds. 

To do so, we will build on the methodologies and analysis outlined in prior Quants Corners, which explored momentum-based rotation strategies for single-stock Cryptocurrency Funds, assessed whether the growing investors interest in Copper Funds can be translated into additional alpha, and set-the-scene for EPFR’s Commodity Model that uses aggregated 20-day flows to pick between gold, silver, energy and agriculture. [See links at the end of the piece] 

Looking beyond the usual suspects 

EPFR has been tracking dedicated Cryptocurrency Funds since mid-2018. It was not, however, until early last year that interest in these funds really took off. This was largely driven by the long-anticipated approval of spot Bitcoin ETFs by the SEC, which marked a turning point for the market. These ETFs allowed investors to gain direct exposure to Bitcoin without the complexities of managing digital wallets or navigating crypto exchanges, significantly lowering the barrier to entry.   

Within the universe of over 156,000 traditional and alternative fund shares classes domiciled globally and managing a collective $71 trillion that are tracked by EPFR, Cryptocurrency Funds currently sit in the Alternative Funds database. The AuM for this sub-group stood at just $12 billion four years ago, $20 billion in December of 2022, and $50 billion in Jan 2024 at which point it more than tripled in size to reach an all-time high of $175 billion in January of this year.  

Together, Cryptocurrency and Physical Gold Funds often overshadow smaller but increasingly important fund types, something reinforced by their ‘share of voice’ in the broader market narrative. The remaining segments – funds with energy, agriculture, other precious metals, rare earths, industrial or base metals mandates – have been flying under the radar in recent months. That presents an opportunity that this Quant’s Corner will delve into. 

Checking the selected tools are fit for purpose 

While these smaller Commodities Fund groups collectively represent a smaller share of total assets, they offer an additional, more granular perspective on the commodities market. In total, there are more than 20 distinct subgroups across energy, agriculture and metals with a combined AuM exceeding $26 billion that we can work with. 

To reduce noise, this analysis includes only those subgroups represented by at least five funds with assets greater than $10 million. This filtering provides a more robust foundation for quantitative techniques like the principal components analysis. The resulting dataset represents a combined AuM of nearly $21 billion. This includes: 

• The energy universe within Alternatives encompasses $9.2 billion in assets, primarily comprised of funds with long-only or short-only strategies tracking crude oil (Brent or WTI) and natural gas.  

• Agriculture Funds have over $2 billion in assets and are almost always crop-specific. The largest subgroups focus on soybeans, wheat & grains, and corn – markets where seasonality, natural disasters and global trade play a role.  

• Beyond gold and silver, the precious metals universe also includes funds tracking the price of platinum, palladium and rhodium. Other funds focus on base or industrial metals such as copper, nickel, aluminum and zinc. Platinum and Palladium Funds together hold nearly $7 billion in assets, and Copper Funds $2 billion in assets.  

With the groups of funds now selected, we can take a deeper dive into their behavior. To that end, we employ principal components analysis (PCA) to identify common patterns in fund returns.  

For PCA to produce meaningful results, an adequate number of funds must report consistently over a reasonable time horizon. Accordingly, we analyze monthly returns for each fund group over the past four years from 2020 through 2025. Each group is categorized by 48 observations, representing monthly returns over this period. From this data, we derive 48 linear combinations, or risk factors, capturing dimensions such as geographic exposure, return volatility, investment style and other key risk attributes.   

The combinations explaining the greatest share of variance becomes the first principal component, with the next strongest becoming the second. The remaining, weaker risk factors are then discarded. The two components then serve as the axes for a two-dimensional representation of the data, by way of a simple dot-plot chart.   

Encouragingly, PCA of EPFR’s custom asset groupings — such as those focused on energy, metals, or agriculture — show a tendency for those groups to cluster. This clustering indicates that the funds within each of these groups have similar return profiles, which not only validates our classification framework but also suggests that these assets are driving distinct and coherent performance patterns within the fund universe. 

Getting to Alpha 

Having validated our inputs, we utilize an established strategy framework to unlock potential excess returns. This strategy uses EPFR’s daily fund flows, ranking each of the commodity custom groups into four equal baskets based on compounded daily percentage flow. The groups with the strongest inflows are placed in the top quintile, while those with the weakest flows are placed in the bottom quintile. Using these, we go long the top quintile and short the bottom quintile.  

The quartile model based on the trailing five-day flow percentage works well. The table below shows the annualized returns to each quartile in excess of an equal-weight basket of commodities, providing details on the quartile spread, annualized return difference, and the associated Sharpe ratio.  

The results show that shorter holding periods lead to stronger performance. The strategy works best at a weekly horizon, where it achieves its highest risk-adjust return (Sharpe = 0.35) and return difference of about +9.1% above the benchmark. As holding periods increase, both returns and consistency in performance weaken.  

With the pace of change we are seeing in global commodities markets, we expect that the inputs to existing models and strategies will need frequent review as fund data expands and becomes ever more granular. 

 What have we learned? 

Our analysis suggests (a) that there is value in looking beyond the biggest fund groups, (b) the universe of fund groups tied to niche or new commodities offers untapped value and (c) because commodity prices adjust rapidly, often on an hourly or daily basis, responding quickly to changes in supply, demand and investor positioning increases the odds of capturing excess returns. 

 Building on the methodologies and analysis outlined in prior Quants Corners: 

• “Signals & Strategy in a Saturated Single-Coin Cryptocurrency Market” – 4Q25 

Introduces a momentum-based rotation strategy for single-stock currency funds (ie. Bitcoin, Ether, Ripple) identified within the Cryptocurrency Fund universe 

• “Copper, a silver bullet for a commodities strategy?” – 4Q23 

Explores whether the broad interest and growing coverage of Copper Funds can be translated, through quantitative means, into additional alpha? 

• “Getting the right balance on commodities exposure” – 1Q20 

Provides an overview of an existing quantitative strategy, EPFR’s Commodity Model, which uses aggregated 20-day fund flows to select between gold, silver, energy and agriculture.  

Did you find this useful? Get our EPFR Insights delivered to your inbox.