Corn Richard James

High Bias in Early USDA Corn Production Forecasts

The One Minute Summary

  • The USDA’s corn production and yield forecasts of May 10, 2018 US are biased high, because the USDA WASE calculates a “weather-adjusted” trend that excludes the possibility of a very dry June.
  • As the June outcome is not yet known, a more realistic estimate of the trend-line yield would be closer to 172 bpa, rather than the 174 bpa estimated in the USDA WASDE report.
  • CropProphet’s 2018 trend yield estimate is 171.7 bpa, and the daily-updating forecast based on weather and satellite inputs is currently below 170 bpa.

USDA Early Corn Yield Estimate

USDA Corn Production (and soybeans) forecasts were issued on May 10. These are the first US corn and soybean yield and production estimates for 2018. According to the USDA’s WASDE report, US corn yield is expected to be 174 bushels per acre (bpa) and corn production is estimated at 14.04 billion bushels. If a corn yield of 174 bpa is realized, this would be the third-highest on record and only slightly below the excellent results of the past two years.

The CropProphet corn yield forecast currently expects US corn yield to be more than 5 bpa lower than the USDA’s estimate, and CropProphet indicates a greater than 75% chance that the yield will be lower than 174 bpa. Based on CropProphet information, the USDA’s early estimate is too high.

USDA Corn Yield Estimate Methodology

The USDA’s method for estimating yield at the beginning of the growing season is detailed here:

https://www.ers.usda.gov/publications/pub-details/?pubid=36652

[See a description of the CropProphet corn yield forecast methodology ]

The USDA WASDE forecast is essentially an estimate of the long-term linear trend with an adjustment made for mid-May planting progress, which is slightly delayed this year (a slight negative impact on yield). However, in deriving the corn trend yield estimate, the USDA makes an assumption that June will not be very dry; in other words, their trend is only valid when June is not dry. According to USDA’s yield equation, if June is very dry (in the driest 10% of years), then the US yield will be reduced by nearly 10 bpa for every inch of the June rainfall deficit compared to normal.

Given that the June outcome is not yet known, the current USDA WASDE estimate is biased high. We cannot yet exclude the possibility that June will be very dry, so a truly unbiased estimate would allow for that outcome (as unlikely as it may be). If the USDA allowed the possibility of exceptional June dryness to impact their model, then their trend would be approximately 2 bpa lower and their 2018 estimate would be closer to 172 bpa. CropProphet’s estimate of the 2018 corn yield trend is 171.7 bpa, and this represents a better estimate of likely US yield without any consideration of weather impacts this year.

Summary – USDA Corn Production Estimates are High

In summary, the USDA’s trend-line estimates for corn and soybean yield exclude the possibility of a very dry June, and therefore the USDA’s early-season yield and production forecasts are biased high. If the potential impact of a possible (albeit unlikely) dry June is included, their US corn yield estimate would be closer to 172 bpa. CropProphet makes no assumptions about future weather and provides unbiased estimates for possible outcomes, including realistic confidence intervals as the season evolves.

Avoid the Hype. Get the facts. Use CropProphet for Grains Market Intelligence.

Try a risk-free trial of CropProphet.

Back