Updated on: October 28, 2024
- Planted area of corn vs soybeans per state
- The U.S. national level soybean conditions
- The U.S. state level soybeans percentage rated as "Good and Excellent"
- National yield estimate and ZS=F closing price
- Yield forecasts by states
- Soybean yield predicitions based on crop conditions: Model performance metrics
- The U.S. soybeans stage progress
- 5-Minute Takeaways:
- Unlike for corn, where crop conditions in Week 29 are already a good predictor for final yields, soybean yields remain uncertain until the late stages.
- Early-season soybean conditions (until at least Week 29) are not strong yield predictors and explain at most 60% of yield deviations from trends.
- States with high variability in historical trend yields and large shares in total production require robust predictive models. Notably, crop conditions provide little predictive information for North Dakota.
- Crop conditions offer moderate yield prediction power for Illinois and Minnesota; however, these states show relatively low historical yield volatility. On average, measured crop conditions account for less than 60% of yield variation around the trend in these states.
In the early season, crop condition models tend to rely more on trend yields unless there is strong historical evidence to suggest otherwise. Therefore, the early season national estimates will generally be biased upwards during poor seasons and biased downwards during good seasons. The plot below shows inverted relationship between estimated yields and ZS=F closing price.
- When looking at the yield model performance, the broad context questions to consider are:
- Which regions have historically the most volatile yields?
- How important are these region in the "total pie" of production?
- In this context we examine: How much volatility do models capture?