Todd's Take

USDA Soybean Stocks Estimate Raises Questions

Todd Hultman
By  Todd Hultman , DTN Lead Analyst
Connect with Todd:
On Sept. 30, USDA announced Sept. 1 soybean stocks in the U.S. totaled 256 million bushels. The number was 84 million bushels above the average trade guess among analysts in Dow Jones' survey and 44 million bushels higher than analysts' highest guess. (DTN ProphetX chart)

On Sept. 30, USDA surprised traders with the declaration "Old-crop soybeans stored in all positions on Sept. 1, 2021 totaled 256 million bushels (mb), down 51% from Sept. 1, 2020."

I remember my mind going blank for an instant, trying to compute the gravity of USDA's words; 256 mb was 84 mb more than Dow Jones' pre-report survey expected and was 44 mb higher than the highest guess among analysts in the survey. Frankly, it seemed a little unreal.

Even more, it was far out of line from what we had told ourselves all year long. In the first seven months of 2021, USDA's ending stocks estimates fell between 120 mb and 140 mb. During the year, many private analysts had estimated stocks below 100 mb, and it was difficult to disagree.

DTN's National Soybean Index hit a high of $16.47 in early May, its highest level in eight years. Cash basis was also its strongest in eight years and, even today, the soybean basis is at its strongest level in at least nine years.

The other bullish kicker was how soybeans in delivery would go off the board at fat premiums to the next distant month. On Aug. 10, just 21 days before the end of the season, the August soybean contract was priced 99 cents above the September contract, an extremely bullish sign of demand from commercials willing to pay up rather than wait another 30 days to secure their soybeans.

P[L1] D[0x0] M[300x250] OOP[F] ADUNIT[] T[]

On the other side of the argument, I have long contended USDA's Grain Stocks reports are some of the best, most reliably helpful forms of information USDA provides. They are not infallible, as we saw from the revision in corn stocks USDA posted one year ago. By and large, however, I'll take USDA's inventory estimates over WASDE report estimates any day.

The one market clue that bothered me throughout the summer was the long, downhill slide in cash soybean prices. Shortly after that dizzying May peak, DTN's National Soybean Index posted a larger-than-expected drop of $2.43 a bushel in the final quarter of the year -- a drop that was also difficult to explain, if supplies were truly tight.

It seemed astonishing that USDA estimated 135 mb of ending soybean stocks in the July World Agricultural Supply and Demand Estimates (WASDE) report and was then left with this 256 mb bombshell for the October WASDE report. I decided to check previous years to see if anything like this had happened before.

The short answer is no. In the brief July to October period, USDA's estimate of old-crop ending stocks went down in nine of the past 12 years. The largest gain of the three years when stocks went higher was 25%. This year's increase from 135 mb in July to 256 mb in October was 90% -- a big jump without recent precedent.

I cannot say with 100% confidence that USDA's 256 mb estimate is a mistake. But, USDA, if you are out there, please double check your figures.

If there was an error, China just got an early Christmas gift and should be actively buying soon.

If 256 mb is the correct amount of ending stocks for 2020-21, I would expect to see more relaxation in the price of cash soybeans versus the futures board and will have to accept the fact that a lot of good questions will go unanswered. Just as cash corn prices told the truth about the 2019 corn crop, I suspect there is still some truth in soybeans' bullish basis. Who knows? Maybe China's demand will eventually provide the answer.

**

Comments above are for educational purposes only and are not meant as specific trade recommendations. The buying and selling of grain or grain futures or options involve substantial risk and are not suitable for everyone.

Todd Hultman can be reached at Todd.Hultman@dtn.com

Follow him on Twitter @ToddHultman1

P[L2] D[728x90] M[320x50] OOP[F] ADUNIT[] T[]
P[R1] D[300x250] M[300x250] OOP[F] ADUNIT[] T[]
P[R2] D[300x250] M[320x50] OOP[F] ADUNIT[] T[]
DIM[1x3] LBL[] SEL[] IDX[] TMPL[standalone] T[]
P[R3] D[300x250] M[0x0] OOP[F] ADUNIT[] T[]

Todd Hultman