Waiting for USDA's World Agricultural Supply and Demand Estimates (WASDE) report to pop up on my computer screen Tuesday, Nov. 10, I saw the new ending stocks estimate flash up 1.702 billion bushels (bb) and also saw corn prices quickly trading higher, up 10 cents within seconds of the report release. The 1.702 bb number was significantly less than the trade estimate of 2.048 bb, so it didn't take long to see this had the makings of a bullish day. Or at least that's what most of us are inclined to think in that sort of situation -- we're prone to make snap judgements when the numbers look as obvious as they did on Tuesday. The problem, however, is that when it comes to WASDE reports, it's generally not a good idea to just take USDA estimates at face value -- not all estimates are equal.
I was recently invited to take part in a virtual workshop put on by Virginia Tech University's Department of Agricultural and Applied Economics. The purpose was to give feedback to USDA's Economic Research Service regarding some of the products they publish, including the monthly commodity outlooks, reports typically put out a few days after the WASDE report, which offer further explanation (see the latest reports at
I was glad to be asked and hope they didn't regret the invitation when I explained that, yes, I use the data and information in the reports USDA provides to help explain markets to our customers; but, no, I don't use the price outlooks as a part of DTN's Six Factors Market Strategies for our customers.
The problem with USDA's price predictions, as I see it, is that no effort is made to identify confidence levels in the many numbers USDA publishes. There are some things USDA does very well and provides a good service to help inform the market.
As I've said often said, quarterly grain stocks reports, even with flaws, contain the most helpful market information USDA provides and keeps all the WASDE estimates anchored to reality. This year's acreage report, which showed corn plantings much lower than widely expected, was also timely and helpful information that would not have been revealed until August had we not had that report.
In contrast, USDA's 2019 acreage report was an example of what not to do as USDA tossed out a guess in the middle of unprecedented planting conditions. USDA began the report estimating corn planted area at 91.7 million acres, only to later reveal that only 83% of the intended corn crop had been planted at the time of the survey. USDA touted the low-confidence estimate and buried the more useful information on page 40.
Unfortunately, many USDA estimates fall into the low-confidence category. For example, the initial May WASDE report at the start of each new crop year is the poster child of low-confidence estimates. In May, some planting has commenced, but we typically have a vague sense about what is being planted and can't yet know how many acres are involved or what the weather will be like for the growing season ahead.
The demand estimates in May are typically a leap of faith, trying to predict a new marketing year that is still 15 months from completion. Statistically, it should be no surprise that the ending stocks-to-use ratios in the May WASDE report have very low correlation (R-squared value = 0.30) to the average farm price estimates the marketing year ends with.
We can analyze the May WASDE report all we want, but we're not going to learn anything useful about future prices. Even so, many still treat USDA's reports as if they're all equal.
The November WASDE report, which USDA released Tuesday, Nov. 10, is further into the season than May and I consider USDA's new-crop estimate of 14.51 bb to be a high-confidence estimate this year. Harvest weather has been nearly ideal in 2020, quick progress has been made and USDA has field data to accompany the producer surveys. It's not perfect, but as estimates go, this one should hold up.
The 325 million bushel (mb) increase in the corn-export estimate, however, raises questions and I can't yet say this is a high-confidence estimate. Supporting USDA's increase, 1.345 bb of corn export sales are already on the books for 2020-21.
January corn on the Dalian exchange is trading near $9.80 a bushel, suggesting China has reason to buy more. The U.S. is also well positioned for more business while South American supplies are low and Ukraine's corn crop was limited by dry weather this year.
The problem is, we don't have a good idea of what China's corn needs are or even how much they have. In Tuesday's report, USDA increased its estimate of China's ending corn stocks from 188.90 million metric tons (mmt) to 191.51 mmt (7.54 bb) -- over four times the ending supply estimated for the U.S. The evidence for China's corn stocks is as real as the invisible thread in the Emperor's new clothes.
I don't make the above points to bash USDA and I mean no disrespect for the efforts of carrying out an impossible task on a limited budget. My focus, however, is in trying to help grain producers make good marketing decisions and understand the markets they compete in. In that regard, it is absolutely necessary to separate out the things we reasonably think we know from the many things we don't.
We (meaning all of us) didn't know what level of demand China would have for U.S. corn earlier this year and we still don't know now. Frankly, it would be more helpful to put "I don't know" responses in the WASDE reports than to make up some number. At least we would have a more realistic view of the risk we're actually facing -- an improvement that would be good for all of our marketing decisions.
Todd Hultman can be reached at Todd.Hultman@dtn.com
Follow him on Twitter @ToddHultman
(c) Copyright 2020 DTN, LLC. All rights reserved.