FORT COLLINS, Colo. (Reuters) - With most of the U.S. Crop Watch corn and soybeans now in the bins rather than in the fields, it is worth taking a moment to examine the results more closely and make comparisons with both the previous year and U.S. government expectations.
For the most part, the Crop Watch data is consistent with recent industry estimates for the U.S. corn and soybean harvest. But it is important to understand the origin of those numbers and what they actually represent.
Crop Watch provides weekly updates on eight corn and eight soybean fields in major U.S. producing states from planting through harvest. This is its second year, and it features the same producers as the 2018 edition.
The purpose of Crop Watch is not to predict the national yields, but rather to offer consistent data points on the same locations as guidance for how recent conditions have affected fields in different parts of the country.
The producers provide both condition and expected yield scores weekly from the start, but yield becomes the focus during the second half and conditions are discontinued once the crops start naturally deteriorating. Yield is rated on a scale of 1 to 5. A score of 3 would be near average, while a 5 is well above average or record and a 2 is below average.
Given the sensitivity of yield forecasting this year, the producers were provided more specific instructions for the ratings to ensure consistency. Scores of 1 or 5 represent yields close to or exceeding 15% below or above average, while 2 and 4 are assigned to yields around 5% to 10% from the average.
The producers’ actual yields are not reported because it would encourage comparisons with the U.S. Department of Agriculture’s state-level yield estimates, which would be unfair to do because of the small sample. Also, all eight producers operate in highly productive regions with farm yields comfortably above those of their states.
The Crop Watch yield scale offers a quick snapshot of the expectations versus normal, saving the reader time, eliminating guesswork, and reducing the misuse of the data. It is also not realistic for producers to estimate exact yields prior to harvest, especially months in advance.
A weekly, eight-field average yield score (and condition score, when applicable) is provided solely to gauge how overall attitude has changed from the previous week. The averages are not weighted because the scores are not intended to predict a national number or correlate directly with a government outlook.
It is worth noting that the fields were selected prior to planting and that all of them had equal odds of any given hardship or weather issue. So if a field does not seemingly represent average fields in the state, that is only by random chance.
As of Sunday, the corn and soybean yield averages were both 3.28, which implies a slightly above-average yield for both, by at most 2%. Those figures include five completed corn and seven completed soybean fields.
That compares with final corn and soybean yield scores of 3.94 and 4.06, respectively, in 2018. So far, this year’s scores would suggest that farmers had relatively worse outcomes for soybeans versus last year than for corn.
USDA’s current national yield estimates suggest the same thing. At 46.9 bushels per acre, soybean yield is down 7% from last year while corn is down 4.5% at 168.4 bpa.
Another interesting observation is that last year’s Crop Watch corn score suggested solidly above-average results, but safely short of record. National yield came in at 176.4 bpa, just 0.2 bpa shy of the previous year’s record and down significantly from previous projections above 180.
But recently, Sept. 1 corn stocks suggested that the 2018 crop may have been overstated, and in hindsight, the Crop Watch scores likely reflected that.
For the most part, however, it is not best practice to use the scores in this way because of the small sample size. USDA’s 2019 yields are below average, not consistent with the Crop Watch averages. Therefore, Crop Watch focuses on the week-to-week movement in ratings and the difference from the previous year.
The 2019 yield scores suggest that Crop Watch growers were more pleasantly surprised with their corn results than their soybean results, though the difference is small.
Early on, corn yield was closer to 3, and then it built up starting in September. The current score of 3.28 is lower than in the previous three weeks, however. The soybean scores had slightly less variation throughout the season.
When the Crop Watch yield scores are weighted by five-year production averages, the results reinforce the upper limit of national production when the top-growing states are not at their best. Doing that brings the 2018 corn score to 4.1 from 3.94, and soybeans to 4.24 from 4.06. Last year, Iowa scored a 5 in both while Illinois corn went 4 and soybeans 4.5.
However, the 2019 scores barely change when weights are applied. Corn yield moves to 3.26 from 3.28, and soybean yield goes to 3.25 from 3.28. This year, the Iowa fields scored 3.75 apiece. Illinois corn and beans were 1.75 and 2.5, respectively, both below what the producer expected.
The opinions expressed here are those of the author, a market analyst for Reuters.
Field photos and more information on Crop Watch 2019 can be found on Twitter using the hashtag #CropWatch19 or by following the handle @kannbwx.
The following are the states and counties of the Crop Watch corn and soybean fields: Griggs, North Dakota; Freeborn, Minnesota; Burt, Nebraska; Rice, Kansas; Cedar, Iowa; Crawford, Illinois; Boone, Indiana; Fairfield, Ohio.
Editing by Matthew Lewis