USDA Report Review

By Darin Newsom
DTN Senior Analyst

Let's start with the conclusion: USDA's quarterly stocks number for corn came in at 2.295 billion bushels (bb), below the average pre-report estimate of 2.349 bb but still the largest ending stocks figure (for 2016-2017) over the last 30 years. And this with record demand of approximately 14.645 bb. Imagine what the situation would've been if Brazil had any type of crop at all in 2016. Would we be talking domestic corn ending stocks about the 3.0 bb mark? Higher still?

Following the release of the Sept. 29 reports, corn tried to follow soybeans higher (more on that later). However, as one market commentator noted, "Corn trying to rally is like riding a bike uphill." I added, "With 2.3 bb on its back." That's the situation corn is facing -- burdensome stocks -- but it's nothing new. It's just that now we get to move on to new crop and all the arguments over production that entails until the release of USDA's January "final" numbers.

Yes, corn could see a rally. Seasonally, the markets (both futures and cash) tend to post a low the end of the first week of October and rally through early June. However, this year's seasonal move could be held in check by bearish futures spreads out through July, with the forward curve covering approximately 70% of calculated full commercial carry. If harvest is as large as projected, those spreads will get even more bearish over time.

As is usually the case with the September Quarterly Stocks report, the soybean number turned out to be the most entertaining. Historically, this report has shown just how "difficult" a time USDA has in estimating domestic ending stocks, starting with the "initial" May guess 15 months prior. DTN Analyst Todd Hultman has been tracking this phenomenon, telling me that 22 out of the last 27 years, the September quarterly stocks figure has been below the initial May estimate. I run a different study, one that shows USDA's September quarterly stocks number came in at 37% of its high estimate over the previous 15 months. This year, the percentages were flipped with the final 301 million bushels down 37% from USDA's peak estimates of 485 mb in both November and December 2016.

Given this track record, why did traders seem so surprised when USDA's figure came in under its September Supply and Demand estimate of 345 mb and below the pre-report average guess of 339 mb? The latter likely drew more head-scratching, with those polled ahead of the report more interested in playing pin-the-tail-on-the-donkey as they lined up USDA's September estimate rather than allowing for the normal percent change. While the subsequent 16-cent rally was impressive (Nov beans closed 8 3/4 cents higher), 301 mb is still the largest soybean ending stocks figure in 10 years. And if USDA is close to right (cough, cough) on its 2017 production estimate of 4.431 bb, the market is still looking at total supplies of roughly 4.7 bb. And that's a lot of soybeans to chew through over the next 11 months. Still, deferred futures spreads are holding a neutral level of carry, indicating commercial traders have doubts over the now-projected new-crop ending stocks estimate of 431 mb. The general idea seems to be this number will be ratcheted down over time. I wonder what ever gave them that idea.

One of the more eagerly anticipated numbers in either the Quarterly Stocks or Small Grains summary reports was the spring wheat production number. The U.S. Northern Plains didn't see rain for months, covering most of this past spring and summer. Meanwhile, temperatures rose to Phoenix-like levels, all expected to take a toll on the final production number. The average pre-report guess came in at 389 mb, well below last year's 493 mb, but still oddly above USDA's September Supply and Demand estimate of 364 mb. Sure enough, USDA hit wheat with another body blow by releasing a spring wheat figure of 416 mb, just shy of the high side of pre-report guesses at 420 mb.

As for quarterly stocks, all-wheat came in at 2.253 bb, above the average pre-report guess of 2.220 bb but below last year's Sept. 1 number of 2.545 bb. This is where wheat gets interesting: Q1 stocks were down year-to-year by 292 mb despite all-wheat production for 2017 coming in 569 mb less than the previous year. And, supposedly, demand had improved through the first quarter. For example: Total export shipments had been running ahead of last year's pace until late this month, after quarterly stocks had been calculated. Given exports make up nearly half of all demand for U.S. supplies, this would seemingly imply that Q1 stocks should have been lower. Instead, wheat on hand remains at bearish levels as farmers head to the field to seed the next crop.

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

Follow Darin Newsom on Twitter @DarinNewsom

(AG)

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