After analysing how<\/a> trends work in commodities, we continue to dissect our recent research paper, \u2018Commodity Investing 3.0: The Rise of Factor and Curve-Aware Strategies.<\/a>\u2019 In this blog, we focus on two of the most important signals in commodity markets: carry and slope momentum. Again, we tested the three metrics both cross-sectionally and in time series. Cross-sectionally, we rank commodities by signal strength to identify relative winners and losers, forming terciles that represent high-, medium-, and low-signal commodities. In a time-series setup, each commodity is evaluated against its own historical performance rather than against its peers, producing a directional signal: long if the factor is positive, and short if the factor is negative.<\/p>\n\n\n\n <\/p>\n\n\n\n Carry, defined as the implied roll yield derived from the futures curve, is one of the most empirically robust sources of return in commodities. It balances inventories, storage costs, hedging pressure and convenience yield. The main theoretical explanation behind this factor1<\/sup> states that when inventories are tight, the convenience yield rises, pushing the curve into backwardation2<\/sup> and rewarding long future contract holders with a positive expected roll return. When inventories are abundant, the opposite occurs, creating negative carry. In this case, investors experience what is in jargon called \u2018roll drag\u2019, which is the performance loss that occurs when a futures curve is in contango3<\/sup>: as the position is rolled forward, the investor repeatedly buys a more expensive contract and sells a cheaper one, creating a negative yield that steadily erodes returns.<\/p>\n\n\n\n We test carry using two definitions:<\/p>\n\n\n\n Our results confirm the robustness of carry as a return driver over the last 25 years, primarily in cross-sectional analysis. In contrast to previous research papers, the time-series results are underwhelming, but the cross-sectional results are strong for both carry definitions. Ranking commodities by their relative roll yield consistently identifies where scarcity is most acute. Performance and significance are similar in both definitions. The main difference is that:<\/p>\n\n\n\n In our tests, both versions delivered more than 5% annualised long\u2013short excess returns, but front-month carry produced significantly higher weekly turnover (15%) than the 12-month version (9%). The chart below displays the three commodity terciles for the 12-month roll yield and the long-short performance (in yellow) of a strategy that goes long on the first tercile (high relative carry) and short on the third one (low relative carry).<\/p>\n\n\n\n From 06\/01\/2001 to 21\/10\/2025. Source: WisdomTree, Bloomberg, Factset. Excess returns in USD. Includes backtested data for illustration purposes. Historical performance is not an indication of future performance, and any investments may go down in value.<\/strong><\/p>\n\n\n\n One of the most famous investors to successfully use second-derivative thinking is Stanley Druckenmiller. Put simply, the second derivative shows how fast something is changing. It helps spot whether a trend is speeding up or slowing down, often before it’s visible in prices. In this context, carry measures the slope of the futures curve. It shows how steeply the curve is in backwardation or contango. Slope momentum measures how that slope is changing over time. It shows whether the curve is becoming steeper or flatter. This information is meaningful: while carry benefits from the persistence of a backwardated curve, slope momentum gains from a curve moving toward backwardation, which often reflects a tightening market.<\/p>\n\n\n\n We construct Slope Momentum as the smoothed, volatility-adjusted change in daily 12-month roll yield (calculated in the same way as we did for the Carry factor). The signal captures whether a commodity\u2019s curve is moving towards backwardation (tightening) or towards contango (loosening). A positive signal indicates the curve is tilting into backwardation, while a negative signal indicates a tilt into contango.<\/p>\n\n\n\n Evidence shows that slope momentum performs well in both cross-sectional and time-series strategies. In the time-series approach shown below, commodities with positive slope momentum are held long, while those with negative slope momentum are held short.<\/p>\n\n\n\n From 02\/04\/2001 to 21\/10\/2025. Source: WisdomTree, Bloomberg, Factset. Excess returns in USD. Includes backtested data for illustration purposes. Historical performance is not an indication of future performance, and any investments may go down in value.<\/strong><\/p>\n\n\n\n In summary, the chart table shows the results obtained for the two factors:<\/p>\n\n\n\n Overall Carry and Slope Momentum results and statistical significance<\/strong><\/p>\n\n\n\n Source: WisdomTree, Bloomberg, Factset. The table reports annualised returns, t-statistics, and p-values for each model and sub-model tested across both cross-sectional and time-series frameworks. The t-statistic measures how statistically different the observed returns are from zero\u2014higher values indicate greater confidence that the factor\u2019s performance is not due to random chance. The p-value represents the probability that the observed result occurred by chance; lower values imply stronger statistical significance. For comparability, the annualised returns for cross-sectional models are divided by two. Historical performance is not an indication of future performance, and any investments may go down in value.<\/strong><\/p>\n\n\n\n <\/p>\n\n\n\n When building a commodity strategy, choosing which commodities to overweight is only part of the process. The next key decision is which futures contracts to use, with many options available along the curve. Traditional indices such as BCOM and GSCI are positioned at the front of the curve. This provides strong liquidity and close exposure to spot price movements. However, it also comes with a cost. Sitting at the front of the curve typically means greater exposure to roll drag, which can steadily reduce returns over time. Our research shows that contract selection plays an important role. Choosing contracts further along the curve can deliver meaningful long-term improvements.<\/p>\n\n\n\n It is important to make a key distinction. Commodity futures curves generally fall into two categories: seasonal and non-seasonal. Seasonal commodities follow predictable cycles in supply or demand. These cycles are often linked to planting and harvest patterns or weather-related usage, such as heating and cooling. Their futures curves tend to show clear, calendar-based peaks and troughs that repeat over time. Non-seasonal commodities, by contrast, do not display these patterns. Due to their non-perishable nature or because they are easily storable (e.g., industrial metals), they do not exhibit any seasonality in their futures curve.<\/p>\n\n\n\n From 25\/01\/2002 to 20\/10\/2025. Source: WisdomTree, Bloomberg. Excess returns in USD. Historical performance is not an indication of future performance, and any investments may go down in value.<\/strong><\/p>\n\n\n\n For markets with predictable seasonal demand or supply patterns (natural gas, grains, softs), high roll yield often reflects seasonality rather than true tightness. Anchoring exposure to the seasonal contract, such as December for natural gas, systematically captures the seasonal hedging premium.<\/p>\n\n\n\n This approach improved returns across every seasonal market tested, with gains that go from 0.7% (on an annualised basis) to the tune of more than +10% (for natural gas), while reducing volatility.<\/p>\n\n\n\nCarry<\/strong><\/h2>\n\n\n\n
\n
\n
Historical performance of roll yield (cross-sectionally using 12-m roll yield)<\/strong><\/h3>\n\n\n\n
![\"Historical](\"https:\/\/www.interactivebrokers.com\/campus\/wp-content\/uploads\/sites\/2\/2026\/01\/13-d-01-curve-1-1100x561.png\")
<\/figure>\n\n\n\nSlope Momentum<\/strong><\/h2>\n\n\n\n
Historical performance of Slope Momentum (time series)<\/strong><\/h3>\n\n\n\n
![\"Historical](\"https:\/\/www.interactivebrokers.com\/campus\/wp-content\/uploads\/sites\/2\/2026\/01\/13-d-01-curve-2-1100x616.png\")
<\/figure>\n\n\n\nModel type<\/th> Model<\/th> Submodel<\/th> Periods<\/th> Annualized return<\/th> t-stat<\/th> p-value<\/th><\/tr><\/thead> Cross-sectional<\/td> Roll Yield<\/td> 12 months<\/td> 6727<\/td> 5.21%<\/td> 2.75<\/td> 0.60%<\/td><\/tr> Cross-sectional<\/td> Roll Yield<\/td> Front<\/td> 6727<\/td> 5.46%<\/td> 3.13<\/td> 0.18%<\/td><\/tr> Cross-sectional<\/td> Slope Momentum<\/td> –<\/td> 6652<\/td> 4.28%<\/td> 2.51<\/td> 1.19%<\/td><\/tr> Time-series<\/td> Roll Yield<\/td> 12 months<\/td> 6035<\/td> -0.62%<\/td> -0.28<\/td> 78.03%<\/td><\/tr> Time-series<\/td> Roll Yield<\/td> Front<\/td> 6098<\/td> 3.26%<\/td> 1.88<\/td> 5.99%<\/td><\/tr> Time-series<\/td> Slope Momentum<\/td> –<\/td> 6035<\/td> 3.78%<\/td> 2.03<\/td> 4.24%<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n Bonus track: contract selection<\/strong><\/h2>\n\n\n\n
Seasonal commodities<\/strong><\/h3>\n\n\n\n
Name<\/th> BCOM contract selection<\/th> Seasonal contract selection<\/th> Relative return<\/th> Relative volatility<\/th><\/tr><\/thead> RBOB Gasoline<\/td> 5.90%<\/td> 7.70%<\/td> 1.80%<\/td> -4.30%<\/td><\/tr> ULS Diesel<\/td> 4.70%<\/td> 5.40%<\/td> 0.70%<\/td> -3.00%<\/td><\/tr> HRW Wheat<\/td> -7.10%<\/td> -4.10%<\/td> 3.00%<\/td> -1.20%<\/td><\/tr> SRW Wheat<\/td> -10.00%<\/td> -5.90%<\/td> 4.10%<\/td> -2.00%<\/td><\/tr> Soybean<\/td> 4.80%<\/td> 6.50%<\/td> 1.70%<\/td> -1.00%<\/td><\/tr> Corn<\/td> -6.50%<\/td> -4.80%<\/td> 1.70%<\/td> -1.60%<\/td><\/tr> Cotton<\/td> -5.20%<\/td> -3.00%<\/td> 2.20%<\/td> -1.80%<\/td><\/tr> Sugar<\/td> -0.30%<\/td> 2.20%<\/td> 2.50%<\/td> -3.40%<\/td><\/tr> Lean Hogs<\/td> -9.30%<\/td> -1.20%<\/td> 8.10%<\/td> -3.60%<\/td><\/tr> Live Cattle<\/td> -0.20%<\/td> 1.90%<\/td> 2.10%<\/td> -1.50%<\/td><\/tr> Natural Gas<\/td> -24.50%<\/td> -13.90%<\/td> 10.60%<\/td> -8.20%<\/td><\/tr> Soybean Oil<\/td> 0.90%<\/td> 2.30%<\/td> 1.40%<\/td> -0.80%<\/td><\/tr> Soybean Meal<\/td> 8.10%<\/td> 9.90%<\/td> 1.80%<\/td> -1.20%<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n Non-seasonal commodities<\/strong><\/h3>\n\n\n\n
Name<\/th> BCOM contract selection<\/th> Max roll yield contract selection<\/th> Relative return<\/th> Relative volatility<\/th><\/tr><\/thead> WTI Crude<\/td> -1.90%<\/td> 5.30%<\/td> 7.20%<\/td> -5.50%<\/td><\/tr> Brent Crude<\/td> 4.80%<\/td> 6.40%<\/td> 1.60%<\/td> -3.00%<\/td><\/tr> Gasoil<\/td> 5.10%<\/td> 5.70%<\/td> 0.60%<\/td> -2.50%<\/td><\/tr> Aluminium<\/td> -2.00%<\/td> -0.70%<\/td> 1.30%<\/td> -1.00%<\/td><\/tr> Zinc<\/td> 0.90%<\/td> 2.70%<\/td> 1.90%<\/td> -0.90%<\/td><\/tr> Lead<\/td> 5.30%<\/td> 6.00%<\/td> 0.70%<\/td> -0.90%<\/td><\/tr> Nickel<\/td> 2.40%<\/td> 4.20%<\/td> 1.80%<\/td> -0.80%<\/td><\/tr> Coffee<\/td> -5.70%<\/td> -4.40%<\/td> 1.30%<\/td> -1.60%<\/td><\/tr> Copper<\/td> 5.50%<\/td> 7.00%<\/td> 1.50%<\/td> -0.70%<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n