JSE Adopts Multiple Reference Points Model After Pilot for Deliverable Soybean Futures

The JSE outlined its decision in a recent Market Notice detailing the outcome of the Multiple Reference Points Model pilot.

Model development

The decision follows years of calls by the JSE for industry stakeholders to propose alternatives to the single reference point system, which does not reflect geographic variations in soybean demand.

In response, Grain SA, alongside independent scientists and technical specialists, developed and tested an alternative methodology aimed at better matching supply and demand through scientific research and market data.

The MRP model was subsequently piloted on the deliverable soybean futures contract over two marketing seasons. The JSE established five predetermined evaluation criteria for the pilot, with an industry-nominated technical committee tasked with reviewing:

• Trading activity, including volumes and open interest;• The number of active market participants;• The management and accumulation of stock in zero-differential areas;• The redelivery of JSE silo receipts; and• Stakeholder feedback and market experience.

Grain SA has raised concerns that the technical committee’s recommendations were omitted from the JSE’s final decision-making process.

Evaluation flaws

The industry body criticised the JSE for failing to provide detailed, criterion-by-criterion reasoning against the agreed evaluation framework.

Although the JSE acknowledged improvements in trading volumes, open interest, and physical deliveries during the pilot, the exchange attributed these gains to external market conditions rather than the MRP model.

Grain SA contends that the quantitative evidence was not fully evaluated and should not have been dismissed due to isolated causality factors or recent contract size changes.

Instead, the JSE’s final notice focused on challenges regarding soybean crushing data availability, disparities between processing facilities, historical data reliance, and divergent stakeholder views. Grain SA maintains that these factors do not provide a factual explanation of how the model performed against the five original success criteria.

The group also highlighted that concentrated soybean processing capacity creates market power imbalances and information asymmetry, leaving producers at a disadvantage and compromising market competitiveness.

Information asymmetry

Data availability remains a primary concern for the agricultural body. While sensitive supply and demand data is already collected under statutory measures via the Marketing of Agricultural Products Act, the JSE cited commercial sensitivity and POPIA constraints for using older data sets.

Grain SA pointed out that established mechanisms exist in other regulated markets to submit independently verified data confidentially without exposing company-level details.

The organisation questioned whether data limitations were applied consistently when comparing the two methodologies, noting that the single reference point model also relies on historical data.

With producers already facing limited visibility into cash-market basis information, concentrated market data prevents farmers from assessing whether location differentials reflect actual commercial conditions.

Additionally, the organisation noted that the JSE’s final notice failed to address concerns raised by its own expert consultant, Prof Roberts, regarding zero-differential delivery points and spatial arbitrage limitations.

Qualitative bias

Grain SA expressed further concern that qualitative feedback—such as claims that the MRP model was overly complex—carried disproportionate weight in the final assessment, given that stakeholder perception was only one of five evaluation criteria.

The organisation argued that a methodology should be judged on whether it is scientifically sound, measurable, implementable, and capable of producing equitable market outcomes, rather than its simplicity relative to older systems. Factual data submitted during the pilot demonstrated that the MRP model operated efficiently and reduced market distortions.

Producer pressure

The decision comes at a time when grain and oilseed producers face narrow margins, rising input costs, and significant production risks. Grain SA warned that farmers cannot absorb further inefficiencies caused by pricing models that misrepresent geographic demand and physical stock movements.

A single reference point risks creating artificial transport assumptions, exposing producers in specific regions to arbitrary deductions and creating opportunities for dominant market players to leverage information imbalances.

The JSE intends to return to a single reference point and has proposed shifting the location from Randfontein to Driefontein starting 1 March 2027. Industry stakeholders have until 14 August 2026 to submit comments on the proposal.

Grain SA confirmed it will not support the return to a single reference point or its relocation without a full assessment of the financial impact on producers, and confirmed it will submit formal comments advocating for an evidence-based location differential system.