Market Forecasting

Forecasting commodity prices with advanced geospatial insights
Sugar price modelsSoybean meal price models
Quantitative price forecasting and systematic strategies

The imperative has never been greater for traditional market participants to keep up with alternative data and algorithmic trading. Discover how proprietary geospatial data can be used to enhance machine and human intelligence for data-driven trading and procurement. We make our forecasts operational, driving customer ROI by:

 

  • Providing daily trade strategy actions to merchants, optimized for optimal profit, drawdown or various other performance metrics
  • Designing smart procurement or hedging programs for firms procuring raw commodities, mitigating the impact of price volatility and achieving optimal savings
  • Customizations specific to client business needs and risk profile
  • Full model and strategy transparency; no black boxes
Get in touch to learn more
Market Forecasting Performance Reports
Strengthen your hedging or trading strategies with agricultural commodity market forecasts. Evaluate our quarterly performance and read our market commentary.
Sign Up for Performance Reports
Proven performance

2.5%

reduction in sugar purchasing costs

67.5%

directional accuracy on 10 day price model

7.5%

accuracy boost using geospatial data

23.5%

annualized return on
sugar trading
The technology
Powered by machine learning and remote sensing, our market forecasting solutions provide improved price prediction for futures and cash markets by fusing fundamental and macro data with proprietary geospatial data from our geospatial processing platform.

Core datasets

Quantify market forces by combining macro, fundamental, and geospatial datasets

Base price models

Increase market visibility with blend of macro and fundamental features

Enhanced models

Adjust base model parameters to better reflect each customer’s specific requirements

Case Study

Delivering ROI with an improved sugar forecasting model

Descartes Labs recently worked with a large CPG client to reduce purchasing costs across $300 million of annual sugar spend. Our goal was to create a systematic approach to sugar purchasing that would enable reliable long-term savings over a heuristic purchasing strategy.

 

We leveraged publicly available macro data to create a systematic approach to predicting price movements along with fundamental remote sensing and geospatial data to fill gaps in market information around physical events that impact sugar price.

 

By creating a robust ICE Sugar futures price model that quantified anomalies from weather, supply, demand, price, volume, energy, and more, we were able to deliver more than 2.5% annual savings using a more accurate sugar price forecast.

Our Sugar #11 price model produced a strong purchasing signal and reliable average annual procurement savings of 2.5%.
Customized for your unique needs
Modifications are made to existing price model packages to reflect each customer’s specific business needs and market insight. Strategies are tailored to specific customer requirements and every solution is unique to the customer. A package of daily model inputs, factors, forecasts and strategy recommendations are delivered in the format that works for you.
Art of the possible

We analyze multimodal data streams to quantify market forces and systematically recommend profitable purchasing or trading actions.

We employ a robust model development process that limits the risk of overfit and data leakage

Learn how proprietary geospatial data can be used to enhance machine and human intelligence for data-driven trading and procurement.
Watch the webinar
Ready to get started?
Whether your goal is to develop purchasing strategies that ensure more stable price, quality, and supply, or the generation of trading profits, or hedging with futures contracts, our market forecasting models can be delivered directly into your technology or trading environment in as little as 12 weeks.
Contact us today