Yesterday, the USDA cut their estimate of this year’s US corn crop by 1.35 billion bushels. Citing planting delays, they trimmed planted area by 3 million acres and slashed yield by a historic 10 bushels per acre. Since corn yield estimates were included with the June WASDE in 1993, the largest previous yield reduction in the month was 5.9 bushels per acre in 1995. While the USDA chose not to adjust soybean acreage this month, the impact was not limited to the corn balance sheet. Reduced corn production drove the USDA to increase the amount of wheat used to feed livestock by almost 60%.
The USDA also released its first crop condition estimates for the corn crop this week, and the low 59% good-to-excellent level bolsters the USDA’s pessimistic outlook. Looking at the historical relationship between crop rating at this point in the season and final yield, the 59% rating alone implies a yield of 164 bushels per acre.
Condition ratings can be used as a simple way to model yield during the growing season. Historically, final crop conditions have proven to be a strong indicator of yield. This year’s initial rating at 59% good-to-excellent is the weakest level since 2002. As the corn crop develops over the season, ratings typically decline. Looking at the past 15 years of conditions data, this week’s 59% level implies a final rating at 53% good-to-excellent. That implied final rating can then be converted into a yield estimate using historical final yield and condition levels. As corn yields continue to trend higher with improved genetics and agricultural practices, how far the final yield deviates from trend is the best metric to compare to crop ratings. Final corn crop ratings and yield deviation from trend have a 91% correlation over the previous 15 years, and a final 53% good-to-excellent rating historically implies a national yield 7% below trend. That would put the crop at 164 bushels per acre using the USDA’s 176 bushel per acre trend yield estimate.
Gro Intelligence’s corn yield model takes a much more sophisticated approach, estimating yield at the county level and aggregating up to a national yield. The model uses crop condition ratings as well as precipitation, NDVI, and other variables like soil type. It currently estimates the crop at 167 bushels per acre.
In May, the USDA estimated that the US would end the 2019/20 marketing year with the highest corn stocks since the 1980s. With this month’s revision, those stocks are projected to be at their lowest level in 6 years. With market chatter indicating the USDA still has another 4 million acres to cut from their corn planted area estimate, the situation could get even uglier. Yield is almost assured to be weak as planting delays lead to increased susceptibility to summer heat and higher prevalence of disease.
In the chart on the left below, 2019’s first corn condition rating (in blue) at 59% is shown in the context of historical ratings. On the right, the deviation of final yield from trend from 2004 to 2018 is shown along with the 7% below trend level for 2019 that current condition ratings imply.