Powered by Gro
Brazil Soybean Crush Forecast Model
Get Gro insights daily or weekly
Subscribe to the Digest

Gro’s Brazil Soybean Crush Forecast Model provides a pre-season and in season estimate for soybean crush in Brazil, the world’s soybean export leader. Because Brazil is the world’s largest soybean producer and exporter, the processing of its soybeans into meal and soybean oil is closely watched globally. 

With Gro’s forecast for Brazil’s domestic soybean crush, users can get a snapshot of global soybean crush demand. They can also use our Brazil Soybean Crush Forecast Model to predict soybean ending stocks in Brazil. Supply and demand in Brazil is key to understanding the global balance sheet of soybeans and how that will impact trade flows and price.

Customers Use the Model to

  • Monitor Brazil’s soybean crush demand in season
  • Get a global picture of soybean crush demand
  • Forecast Brazil’s soybean ending stocks
  • Anticipate Brazil’s soybean available export capacity and flow to make better sourcing and shipping decisions

Why It Matters

Brazil is the global leader in soybean exports, and soybean crush is a major use for its crop. Having access to a Brazil soybean crush estimate can help users position themselves ahead of the USDA’s monthly supply and demand data releases (WASDE), and it can help users arrive at a more finely tuned soybean ending stocks estimate.


Before the soybean market year starts, our Brazil Soybean Crush Forecast Model uses a simple linear trend in the annual Brazil soybean crush from 2000 to the present to make a pre-season prediction. We continue to use this method after the market year starts in February. 

In August, we switch to an in-season forecasting method that takes the year-over-year difference for the accumulated monthly crush data for each previous month up to the month that we are setting a forecast for; it assumes that difference to be the same as the change in the actual annual crush compared to the previous year. 

Our Brazil Soybean Crush Forecast Model has been tested against more than 20 years of USDA PS&D ground truth Brazil soybean crush data that is equal to the sum of the monthly data (Source: ABIOVE). 

Pre-season through August of the market year, our Brazil Soybean Crush Forecast Model’s mean absolute percentage error (MAPE) is 2.75%; its MAPE then falls from 2.2% to 0.8% from September to the end of the market year.  These results indicate that, on average, our Brazil Soybean Crush Forecast Model is more accurate than USDA’s forecast. 

Related Resources

Available in these plans
Check mark
Check mark
View plans
Get a demo of Gro
Talk to our enterprise sales team or walk through our platform