Newsom on the Market

By Darin Newsom
DTN Senior Analyst

Before you roll your eyes, "Here he goes again, bashing USDA ...," let me tell you, this isn't that kind of column. In fact, I'm going to spend some time pointing out where USDA has it right, almost to the point of being ridiculous.

In fact, if I could foretell any market aspect near as well as USDA does in one particular area, I'd have a different job trading my own account, drinking bourbon on a beach, and saying, "Feeling good, Louis!"

But I can't, so I write about my view of markets from the press box.

As you may have heard, this past Wednesday saw the release of USDA's May Supply and Demand and Crop Production reports, including its annual "initial" look at new-crop expectations. As most of you know, I write initial that way because it is actually the third look at new-crop estimates, following USDA's annual Outlook Forum in February and the March 31 Prospective Plantings report.

If you're looking to make a strong bet on where USDA is going to come in on production in next year's May report, simply multiply its Outlook Forum trend-line yield by its prospective plantings harvested area. You will amaze your friends and family with how close your guess is.

This column isn't about production estimates though, but rather USDA's insane ability to peg what new-crop corn ending stocks will be in the following year's (18 months down the road) September 30 Quarterly Stocks report, showing stocks on hand as of September 1. I keep a table of such things, and the last two marketing years have been nothing short of phenomenal.

In May 2014, USDA estimated 2014-15 U.S. corn ending stocks at 1.726 billion bushels. When the September 30, 2015, Quarterly Stocks report rolled around, USDA pegged corn stocks at the end of the marketing year (August 31, 2015) at 1.731 bb. That's a difference of only 5 million bushels. A difference so small it has to be an outlier, right?

Wrong. In May 2015, USDA opened the bidding on 2015-16 corn ending stocks at 1.746 bb. And, while its estimates over the next 18 months ranged from 1.599 bb (July 2015) to a whopping 1.862 bb (April 2016), quarterly stocks for September 1, 2016, settled in at 1.738 bb. A difference from May 2015 of 8 mb. Again, stunningly close despite its wild ride of outlandish guesses over the 18-month timespan.

Do you recall where USDA put 2016-2017 corn ending stocks a year ago? If not, here's a reminder -- 2.153 bb. Since then, we've seen the number vacillate between 2.008 bb (June 2016) and 2.409 bb (August 2016), before settling into a sideways pattern near this May's estimate of 2.295 bb. But, if the last two years are a true indicator, USDA could be currently overestimating 2016 U.S. corn ending stocks by roughly 144 mb.

Similarly, USDA's May soybean ending stocks estimate also makes the final quarterly stocks figure easy to estimate, but in a far different way than corn. When corn has been nearly spot-on to the bushel, one can quickly calculate U.S. soybean ending stocks by dividing USDA's May projection in half.

Sounds crazy, I know, but it tends to work. Back in May 2013, USDA projected 2013-14 soybean ending stocks at 336 mb only to have the September 30, 2014 Quarterly Stocks report come in at 130 mb. The 2014-15 marketing year saw an opening salvo from USDA of 336 mb and closing calculation of 210 mb, off "only" 37.5%. 2015-16 brought the average back to roughly 50% with its May 2015 number of 500 mb and September 1, 2016 final of 195 mb, or off by 61%.

This marketing year, 2016-17, USDA opened with 305 mb, quickly dropped it to 260 mb (June 2016) before racing to a peak of 480 mb in both November and December 2016. Wednesday's report has seen it trimmed slightly, as USDA is wont to do this time of year, to 435 mb. Once again, if history repeats itself, could we be looking at continued slicing over the coming months down to approximately 150 mb?

What does all this mean? New-crop 2017-18 ending stocks came in at 2.110 bb (corn) and 480 mb (soybeans), and if I see the patterns between now (May 2017 (for those of you reading this in the future) and September 2018, then so do computer algorithms. Therein lies USDA's conundrum.

I know many of you are undying fans of all things USDA; some of you are not. But, all of us in agriculture need to ask ourselves if it is actually in the best interest of the U.S. ag industry for the department charged with "helping improve the economy and quality of life in rural America" (part of the mission statement pulled from USDA's website) to be giving global investment traders such strong predictions about the future? Where is the incentive for markets to figure it out on their own when all they have to do is take long-term positions based on USDA's May guesses?

Granted, using history this short discounts the weather wildcard. But, that means we have to hope for severe weather causing production shortages somewhere, every year. While it happened from 2010 through 2012, as the last four years have shown, it's hard to count on it always occurring.

**

Editor's Note: We're sharing this week's Newsom on the Market commentary from DTN Senior Analyst Darin Newsom in our Top Stories segment across all DTN/The Progressive Farmer platforms. Newsom on the Market regularly appears Friday mornings on our DTN subscription products such as MyDTN. To find out more, visit http://www.dtn.com/…

Darin Newsom can be reached at darin.newsom@dtn.com

Follow him on Twitter @DarinNewsom

(CZ/BAS)

Copyright 2017 DTN/The Progressive Farmer. All rights reserved.