Another Look at USDA August Reports

USDA gave fundamental analysts a lot of fresh grist for the mill on Friday, as NASS released its first-of-the-season official harvested area, average yield and production forecasts for corn by state as well as nationally. Previous estimates for May through July were generated via historical quantitative methods but had no farmer surveys or satellite data in them.

The August NASS estimates include 14,672 farmer surveys taken between July 29 and Aug. 7 as well as MODIS satellite data through July 27. They do not include the USDA objective yield plots, where USDA personnel visit the random field plots. That will be done for the September and later reports.

NASS also made no changes to planted acres in this report, using the June 30 numbers. This has created some possible disconnects, which we will explore.

The NASS average yield estimate is down 2.4 bushels per acre (bpa) from last month and below their version of 1994-2003 trend yield. As we have shown in the past, trend yield varies greatly with the number of years of data you choose to incorporate, and also the end point. There are some fairly dramatic yield swings from last year. No state is currently seen with a record state average yield, although Indiana would tie its record if it hits 195 bpa. The yield shifts are intuitive, with the Plains states recovering from last year's drought, and an increase in D2-D4 drought area in pockets of Missouri, Illinois, Iowa, Wisconsin, Minnesota and North Dakota. Our Brugler500 index for crop condition ratings also agrees with the national yield estimate based on Aug. 6 conditions. The correlations are loose this early in August, plus or minus 8 bpa. Said another way, final yield is likely to be between 167 and 183 bpa using our conditions tools. NASS' 20-year reliability data shows a root mean square error for the August report versus final production of 4%, with an average shift of 384 million bushels (mb) and the biggest swing ever at 1.167 billion bushels (bb). It is way too early to put all your eggs in one basket when it comes to the final yield forecast.

We've talked a lot about the yield number, but recall final production is yield times acres harvested for grain. For corn, the spread between planted and harvested acres can be 8 million to 10 million acres. Little of that is actually abandoned or zeroed out for crop insurance purposes, although we can show you some dryland fields and hailed-out fields where that will be the case. Most of the gap is due to silage production for cattle. Most of that is planned, but within nitrate limitations stressed, corn for grain can become silage as producers make the best of a bad situation. Failing to make it to maturity can also result in more silage production, although this year's crop is ahead of average maturity at the moment.

There is a possible disconnect between NASS' planted and harvested acres numbers. Farmers are reporting 86.322 million acres will be harvested. That would be 91.74% of the planted acreage USDA is currently using. An improvement from 2022 is expected, just like percent harvested went up from 89.8% in 2012 to 91.71% in 2013. That 91.74% would be similar to 2013, but also the largest since 2017.

We think this discrepancy suggests a possible increase in planted acres in the September Crop Production report. Remember, they are using June 1 planting numbers right now. In September, administrative (FSA) data will be tied into the NASS frame samples. FSA data released Friday suggests either a big improvement in USDA's data-handling speed or that final planted acreage will be larger than 94.1 million based on historical ratios. Given the ratio above, it seems likely to be the latter.

Now, before you get too "beared-up" on corn, let me remind you harvested acres times yield is what drives production numbers. Finding more planted acres without changing harvested would fix the distortion in percent harvested without having to change production. FSA data was incorporated in the wheat production numbers in this August report and increased the estimate 200,000 acres from July. It did not increase U.S. production, however, due to a downward average yield adjustment.

The biggest wild card for corn production is still likely to be yield changes between now and January. Crop tours are underway right now to figure out conditions on the ground and correlate them to the satellites and surveys. NASS will do the same thing with their objective yield plots beginning in September, counting ears per acre, dried-down grain weight per ear, and estimating harvest losses. The production debate is far from over.

Alan Brugler can be contacted at alanb@bruglermktg.com

(c) Copyright 2023 DTN, LLC. All rights reserved.