AI for commodity trading: Permutable AI’s one-year results review

Although at Permutable, our core offering is market intelligence, one of our key differentiators is that we run a live systematic commodity trading strategy built on those insights which has delivered clear, repeatable results over the past year. 

Clients asked whether our sentiment signals only back-test well. We took this one step further – using our AI for commodity trading, leveraging our AI-driven signals  in real time. The result is a disciplined commodity trading strategy that has not only navigated tariff headlines and geopolitics with conviction, but taken the right side of regime shifts, and delivered consistent risk-adjusted returns. Against other benchmarks we have outperformed 90%+ of CTAs, proving that sentiment is not only actionable but has clear alpha. 

From 1 Oct 2024 to 1 Nov 2025 our live book outperformed major benchmarks with low equity beta and tight drawdown control. The track record stands at 20.6% with 7.3% volatility, a 4.4% maximum drawdown and a Sharpe of 2.85. Over the same period the S&P 500 returned 18.1% with 18.2% volatility, while the GSCI delivered 4.1% with 17.2% volatility. We run a balanced long and short structure, trade highly liquid front-month futures, and spread risk evenly across energy, agriculture and metals. 

AI for commodity trading 2
Strategy: A simple balance mixed 50/50 long-short commodity book of six highly liquid contracts across energy, agriculture and precious metals. These results highlight the effectiveness of using AI for commodity trading in delivering steady, risk-adjusted performance even amid volatile global markets.

Performance Snapshot (live, 1 Oct 2024 to 1 Nov 2025)

  • Return: 20.6%
  • Volatility: 7.3%
  • Max drawdown: 4.4%
  • Sharpe ratio: 2.85
  • Correlation to S&P 500: ~12.2%
  • Risk Allocation: Evenly distributed VaR across global energy, agriculture, and metals markets.

October strategy review

October Review

Brent Crude (+13.6%)
Brent delivered a firm 13.6% return over October, with our strategy capturing both the early-month softness and the subsequent recovery. A series of well-timed long entries into the late-month rebound drove gains, while short exposure earlier in the period cushioned the drawdown. The performance reflects a measured approach to positioning in a market that remained volatile but directional.

Our AI-driven signals turned higher as sanctions tightened and shipping/insurance frictions pushed a premium into prompt barrels. With logistics risk lifting the front of the curve, conviction rose and we seized longs while keeping discipline as spreads cooled. This is reflected in the graph below, where our system detected the convergence of bullish factors for Brent crude oil in late October and made a timely long call that captured the upside.  

Brent rally

Natural Gas (+21.8%)
Natural gas was the standout performer, returning 21.8% across a turbulent month. The model navigated sharp reversals effectively, closing shorts before the mid-October rally and maintaining disciplined long exposure as prices surged. This result underlines the strategy’s capacity to handle high volatility and extract value from rapid shifts in market sentiment.

The signal pivot came as weather risk and record US export prints outweighed supply additions. Alerts around project restarts and cargo cancellations kept risk light into early November, preserving gains as momentum faded.

Silver (-1.3%)
Silver finished marginally lower, hindered by premature long entries ahead of a short-lived correction. Losses were contained, however, with risk controls limiting downside. The strategy adapted well in the first half of the month, realigning with renewed downside pressure. Overall, performance was steady given the reversal of conditions in precious metals.

Gold (+9.0%)
Gold posted a formidable 9% gain, benefitting from well-placed long exposure during the mid-month rally. The strategy avoided early noise and tracked the broader risk-off theme that spurred precious metals. Its patience in holding positions through the mid-October climb captured the bulk of the upward move, reinforcing gold’s value as a defensive trade.

Conviction eased later in the month as the cooling of tension and risk appetite re-emerged, cooling the heat out of the yellow metals run. We followed the signal back to neutral stance and banked the long position, treating following movements as profit-taking within a still supportive structural backdrop of central-bank and Asian demand.

“Annotated price chart showing gold prices alongside machine-readable fundamental, sector, and macroeconomic sentiment signals, illustrating a sustained bullish regime identified by Permutable AI’s Trading Co-Pilot during a period of rising gold prices

Soybeans (+5.1%)
Soybeans advanced 5.1% as the model successfully rotated from short to long in the second half of the month. The strategy captured the late-October recovery in agricultural prices, with improved timing following a choppy start. The gain reflects stronger trend recognition and tighter discipline in execution.

Policy tone improved with Washington–Beijing détente and re-emergence of Chinese booking late October, we rotated long and trimmed as headlines faded and momentum cooled into month-end.

Wheat (+0.8%)
Wheat recovered from early losses to close modestly higher at 0.8%. Initial exposure proved to miss the mark, but later long entries boded well in the late-month rebound. The system’s adjustments helped preserve capital through volatility, as momentum pivoted.

In late October the easing trade tensions lifted confidence in forward purchasing, yet a comfortable supply backdrop (Russia, Argentina, Western Australia) capped bullish enthusiasm. Signals supported a tactical long bias late in the month, with risk kept modest given soft US inspections and slower EU exports.

Each commodity’s performance further validates our thesis: that AI for commodity trading can recognise shifting market narratives earlier than price and guide positioning with measurable confidence. Our systematic trade performance from late September to early November reflects the effectiveness of signals and strategy.   

systematic trade

How this fits into client strategy and workflow: The case for AI for commodity trading

Our approach offers institutional desks a practical way to integrate AI for commodity trading directly into macro, CTA, and multi-asset strategies – providing low-correlation returns and explainable signals.

Where it fits
  • Diversifier with low equity beta: Adds a differentiated return stream through commodities market due to the current stock rally, correlation ~0.12 to the S&P 500.
  • Macro overlay: Signals highlight regime shifts around tariffs, supply dislocations and geopolitics, supporting tilt decisions in broader macro books.
  • CTA complement: Cleaner risk profile and shorter reaction time versus broad commodity indices.
Transparency
  • Explainable signals: Each change is accompanied by a short note on drivers, including from geo-political, supply, demand, broken down further through region, to which narratives are dominant.
  • Audit trail: Version controlled signal history, orders and fills, with a daily, weekly and monthly review.
Benefits to the desks and PMs
  • Earlier conviction: Signals often turn ahead of price, helping analysts and PMs to lean into market moves rather than chase them.
  • Clean event narratives: Clear asset agnostic event allocation capturing the drivers as they unfold in real-time, allowing them to keep on top of regimes and infection points.
  • Operational ease: Delivered as an API plus an intuitive UI dashboard, ready for modelling, market catch ups for morning meetings and intelligent insights briefing notes.

Why Permutable is the leader in AI for commodity trading

Running the strategy live demonstrates what matters for investors: sentiment is actionable. The same domestic and international lenses that power our research inform entries, exits and risk allocation, helping us spot regime shifts early, lean into market narratives that matter, capturing the alpha. Our systematic performance record makes the case clear, sentiment matters more than ever whilst AI for commodity trading can operationalise that insight at scale.

How clients work with us

We have often been asked why we haven’t launched a fund to showcase our systematic strategy. The answer is simple – our goal isn’t to compete with clients or run capital, but to provide technology, data, signals and workflow infrastructure directly to institutional partners in a variety of ways.

Through our API Licence, partners can access our asset- and macro-level APIs, embedding real-time signals and alerts into existing OMS and trading systems without the friction of new infrastructure. For those seeking a more visual experience, our Trading Co-Pilot dashboard offers secure access to live signals and alerts, market intelligence, and event-driven reporting – ready to plug into existing workflows.

We recognise that adopting new AI systems can come with integration and governance challenges, so our model is designed for transparency, explainability, and control – helping partners deploy adaptive AI with minimal disruption.

Explore our systematic trading solutions or read our case study to see how institutions are using our approach in practice. To stay up to date, sign up for our newsletter or contact us at enquiries@permutable.ai to find out more.