Todd's Take

A Problem Estimating World Grain Stocks

Todd Hultman
By  Todd Hultman , DTN Lead Analyst
Connect with Todd:
USDA's Oilseeds: World Markets and Trade report from Feb. 8 revised ending soybean stocks for Brazil and Argentina as far back as their local years for 2000-01 to the present. Overall higher ending stocks for Brazil and lower ending stocks for Argentina remind me of a three-year old conversation with a USDA official and should remind us not to take these estimates too seriously. Data from USDA, Oilseeds: World Markets and Trade for December 2018 and February 2019. (DTN chart)

Much of last week in DTN's newsroom was spent preparing for USDA's reports on Friday. Ever since USDA closed the lockup room in Washington, D.C., last year, World Agricultural Supply and Demand Estimates (WASDE) report day has taken on an extra layer of anxiety as a whole village of staff plans and works behind the scenes to make sure we all hum in unison to get USDA's new estimates out as quickly and accurately as possible.

We've all seen hectic portrayals of newsrooms in the movies with people yelling and reporters running through the office to make deadlines. WASDE report day at DTN is actually much different. There is an eerie quiet when the clock strikes 11:00 a.m. Every second waiting for data to show up feels like a minute, and when the numbers do appear, a nervous chatter of cross-checking begins.

Coming off of a five-week government shutdown, Friday had plenty of new estimates to examine, ranging from U.S. and South American crop estimates to Dec. 1 grain stocks to winter wheat seedings with numerous details in between. As DTN's lead analyst with exactly one hour to prepare a public webinar, I was scrambling to highlight the most important findings and slap up the best charts I could get ready.

Now in my 35th year of watching WASDE reports, I have to admit the process has not gotten any easier. In my role at DTN, I no longer trade markets, but the tight stomach never goes away and I am constantly wondering how USDA might surprise us this time. Both in my older days as a broker and trader and my more recent days as an analyst, there has always been a concern of having overlooked something.

After Friday's duties were done and I had a little time to reflect on what just happened, it seemed to me that the reports in general had been fairly tame. There was even some slightly bullish news. USDA reported record-high corn demand for the first quarter of 2018-19. World ending soybean stocks had also fallen more than expected, thanks to modest revisions from previous years.

What I didn't realize until later Friday evening were the surprises in Friday's data, which many probably missed. It didn't come from the WASDE report or the grain stocks report, or even winter wheat seedings. It came from USDA's Oilseeds: World Markets and Trade, a less-talked-about publication that is full of helpful information (see it here:…).

The February issue began with a bullish headline: "Smaller Supplies to Limit Brazil's 2018-19 Soybean Exports." Opening remarks explained how USDA expects Brazil's soybean exports to drop from 84.2 million metric tons (mmt) in 2017-18 (the local season that just ended on Jan. 31, 2019) to 70.0 mmt, or 2.57 billion bushels (bb), in 2018-19.

WASDE estimates, based on the U.S. crop year, don't match with Brazil's local season, but the thought that Brazil just exported 84.2 mmt of soybeans and USDA estimated China's total import needs at 88.0 mmt is a bit breathtaking. A new-crop estimate of 70.0 mmt of exports offers some relief, but is still a lot of soybeans for the U.S. to compete against.

Now for the interesting surprise: Under a section titled "Brazil and Argentina Soybean Stock Adjustments," USDA explained how they had just gone back to 2000-01 and revised ending stocks estimates for the two countries. For Argentina's latest marketing year, which will end on March 31, 2019, December's ending stocks estimate of 16.85 mmt was roughly cut in half, to 8.44 mmt, or 310 mb.

For Brazil's latest marketing year, which ended on Jan. 31, 2019, December's estimate of ending soybean stocks was raised from 775,000 metric tons (mt), or 28 million bushels (mb), to 1.15 mmt, or 123 mb -- not a huge difference. However, looking back at prior years, USDA now says the local seasons from 2011-12 to 2016-17 had far higher ending stocks than previously estimated. Four of the years had ending stocks four times higher than originally estimated.

As the report admits, "Determining realistic stock numbers for both Argentina and Brazil has been an issue for some time." This reminded me of an email exchange I had with a USDA official in early 2016 (written about here:…).

At the time, I asked the official if he could offer any insight as to how Brazil's soybean prices could be so low at a time when USDA was estimating ending stocks at an "unbelievably low 17 mb for the current marketing year?"

As I explained in the article, USDA was not willing to make a public comment, but the anonymous official offered some helpful possibilities, which included unreported production and unreported imports. He did not specifically mention Argentina, but it was not difficult to imagine Argentine farmers finding creative ways to get around punitive export taxes by crossing the border.

Within four months, USDA revised 11 years of corn data for China, adding nearly 6 billion bushels to its count of ending stocks, and now has revised 18 years of South American ending soybean stocks, showing significant changes through the years.

Understandably, some will cuss USDA for being off by so much all these years, and I understand the frustration. But as I see it, the real world is messy. You will not hear me bash USDA for attempting to make their estimates more accurate. I would also not want the unenviable task of trying to figure out how much grain is in any country, especially those that are politically closed or have reputations for political corruption.

On a more practical note, however, our customers need to know that DTN's Six Factor Market Strategies do not depend on USDA's estimates of world grain inventories, and I would not recommend any producer bet the farm based on these numbers. Let these two recent examples remind us yet again why USDA's world estimates should be taken lightly. We have other data and market clues that offer more help.

After watching WASDE reports for so long, it's no wonder why my stomach knots up on report day.

Todd Hultman can be reached at

Follow him on Twitter @ToddHultman


Todd Hultman