Gro’s China Pork Consumption Model forecasts are now available to our Enterprise users on the Gro Platform. The model is able to predict pork demand nine months before the USDA PS&D’s initial estimate and a full two years before the USDA’s estimate tends to stabilize.
Gro’s model creates predictions with an error of only 2.7% versus the final USDA PS&D number. The USDA’s own initial estimate has an error of 5.3%. Gro’s model also updates the forecast daily, as opposed to USDA which updates its forecasts quarterly.
China is the world’s largest consumer and producer of pork. The scale of China’s hog herd, the largest in the world, makes it a key driver of global soybean demand, and a major influence on agricultural markets worldwide. Understanding how much pork is being consumed in China is paramount for decision-making.
The multi-stage machine-learning model was built on 7 categories of signals that govern protein demand, selected after testing over 400 features. It is a meaningful step forward in the industry’s ability to forecast meat demand. As usual, Gro Enterprise clients have access to the full methodology, underlying data, and backtests, making the model fully transparent, reproducible, and modifiable to suit their needs.
We are excited to be adding this valuable resource to our already extensive library of Gro models. Please email email@example.com if you would like access, or if you would like a recording of a detailed webinar introducing this model.