The granddaddy of them all and at a time grain markets are in the midst of the most significant price gains and fundamental concerns in years…the USDA’s January 12 set of reports loom large. With the markets on edge as U.S. soybean and corn ending stocks ideas are already at concerning levels, demand showing little sign of slowing and South American weather less than ideal, the table appears set for potential fireworks in either direction depending on whether any surprises are revealed in the USDA’s numbers next Tuesday
Wire services have not yet published trade estimates for next week’s reports. We will post a summary of the estimates on Market Insights as soon as available.
Corn
With so much to cover heading into the January 12 reports, let’s just get into it…
The multi-faceted nature of the mid-January reports leaves multiple ways for surprises to occur, making this easily the most risky day of the year for the grain markets in our opinion. Combine that with already extremely heightened emotions, rampant demand and notably tightened U.S. balance sheet ideas, next Tuesday’s report day is ripe with explosive price potential in either direction.
For corn, the series of reports starts with the Annual Crop Production report as any changes/surprises there would feed into the quarterly Grain Stocks reports and eventually potentially/likely into the U.S. corn ending stocks estimate in the WASDE report.
Throughout the crop estimating season this fall, the USDA lowered the U.S. average corn yield in each monthly report, with a 3.3 bushel/acre reduction in September followed by a 0.2 bu/acre decline in October a 2.6 bu/acre further cut in November. Since 1970, only six other years also saw the same pattern of yield revisions, with the last four (2010, 2000, 1995, 1993) all seeing another yield cut in the January annual report by 0.2-2.4 bushels/acre. Going back further nearly 30 years ago to 1983, the final yield was ticked back higher by 0.6 bu/acre, while a 1.6 bu/acre increase was reported in the annual report in 1970.
Also supporting prospects for a yield reduction from the USDA’s 175.8 bushel/acre estimate in November is the rather elevated nature of that estimate relative to final crop conditions, as shown in the chart of the relationship between the two factors in the following chart. The USDA’s monthly yield estimates have consistently run above the historical relationship with monthly crop conditions throughout this year’s crop estimates so that, in and of itself alone, is not a strong enough factor to imply the final yield is below the USDA’s current estimate as upward yield shifts relative to conditions do occur at times as evidenced in the following chart. However, when combined with the pattern of yield revisions this year, we feel there is added credence for at least a modest yield cut.
We also feel the full impact of this summer’s derecho wind event across Iowa has not yet been accounted for as the state’s harvested acreage estimate was lowered by a total of 850k acres from the initial 13.550 million acre estimate in August to 12.700 million where is currently stands – and which was not revised in the November report. We feel total harvested corn area could still easily be lowered by another 300-500k acres from the USDA’s 82.5 million acre estimate in November.
Based on these ideas, we see the U.S. corn crop at 14.385 billion bushels with an average yield of 175.0 bushels/acres, down 122 million bushels and 0.8 bu/acre from the USDA’s November estimate of 14.507 billion, 175.8 bu/acre. In recent years, the bias for the USDA’s January corn production estimate has clearly been to the downside, with five of the last seven years seeing reductions of 53-206 million bushels from November, with only two years increasing by a modest 26-31 million bushels.
There has also been a bias in recent years for the January corn crop estimate to come in below market expectations, having done so in five of the last seven years, as well, though last year’s 179 million bushel higher-than-expected estimate was the largest “bearish surprise†in 10 years and 2nd largest of the last 30 years.
Moving onto the quarterly Grain Stocks report, recent-year reports have actually been fairly uneventful in context of the overall balance sheet construct with the last two years’ December 1 corn stocks estimates coming in 122-140 million bushels below the average trade estimate (modestly supportive, but far from significant given the large stocks situations at the time), while the prior four years’ reports were all within 80 million bushels of the average estimate. This was a welcome break after the massive surprises which regularly occurred from 2007 through 2013 of 250+ million bushels in both directions.
First quarter 20/21 (Sep-Nov) U.S. corn demand has been decent, but not necessarily spectacular. Exports have been solid, estimated at 450 million bushels vs last year’s historically poor 270 million, while corn for ethanol usage in the quarter of 1.267 billion bushels was down modestly from last year’s 1.302 billion and the lowest in seven years. As the post-COVID shock recovery continues, so should corn for ethanol usage, but a flattening out of the recovery is clearly evident of late.
As always, feed/residual usage is, by far, the greatest unknown in the December 1 stocks equation, accounting for more than half of total 1st quarter usage. With the higher prices during the 1st quarter than last year and with animal numbers pulling back from historically high levels, we see the potential for Sept-Nov implied feed/residual usage to be down from last year’s 2.634 billion bushels. While the explosion in corn prices is a more-recent situation, the gains started to heat up in October, resulting in average cash prices in many/most locations during Oct-Nov 20-35 cents/bushel higher than a year ago. We also saw U.S. hog numbers begin to decline for the first time in years with a nearly 1% nationwide decline reported for the quarter, while broiler hatch during the Sept-Nov period averaged 1.2% below year ago levels, as well. Cattle on feed numbers were still solid during Sept-Oct, though, but began pulling back in November and were unchanged year-over-over year as of December 1.
Any attempted measure/adjustment of implied feed/residual usage using prices/animal numbers as guidance is only as accurate as the validity of the previous year’s feed/residual usage, as well, which we all know has been far from comforting of late. Unfortunately, though, it is what we have to go on when trying to assess quarterly stocks figures. Accordingly, we’re estimating Sept-Nov implied feed/residual usage at 2.500 billion bushels vs 2.634 billion last year, but would still be the 2nd highest on record. With 1st quarter food usage estimated at 335 million bushels, we see 1st quarter total corn usage at 4.552 billion bushels, nearly identical to the last two years’ 4.531 billion and 4.550 billion bushels.
This puts our December 1 U.S. corn stocks estimate at 11.835 billion bushels vs 11.327 billion last year, but below 2016/17-2018/19 Dec 1 stocks of 11.9-12.6 billion bushels. While U.S. corn stocks through the first quarter of 2020/21 are expected to maintain a decent year-over-year gain, that is likely to be short-lived as export demand and corn for ethanol usage during the remainder of the marketing year should notably outpace year ago levels.
Shifting to the WASDE report, obviously the USDA’s U.S. corn production revision and December 1 corn stocks estimates will have an impact on 2020/21 ending stocks as USDA adjusts the supply side of the balance sheet and potentially annual feed/residual usage ideas, accordingly, but we don’t feel export and corn for ethanol usage revisions are immediately necessary.
By all means, export demand for U.S. corn remains impressive, with total commitments of 1.701 billion bushels as of 12/24/20 up 135% from last year’s 723 million and easily record high. Moreover, sales on the books currently account for 64% of the USDA’s annual export projection, the highest level of late December sales to annual exports in eight years. On the other hand, the USDA’s 2.650 billion bushel export projection already reflects expectations for record exports in 2020/21 and there have been several past years where late December sales accounted for 65-70% of annual exports. Additionally, 2018/19 and 2016/17 saw 61% of eventual annual exports already on the books by now so the current level of sales relative to estimated annual exports is certainly higher than average, but not egregious. This, in our opinion, provides USDA with the ability to wait and see how sales progress further before making another revision if they so choose. Based on current sales on the books, we estimate corn sales will need to average roughly 26 million bushels/week through the end of August in order to reach the USDA’s 2.650 billion bushel export projection vs last year’s 29.5 million/week and the most-recent 5-year average of 29.0 million. We do feel export business is likely to remain quite stout, with additional Chinese buying likely at some point, leaving our 2020/21 export estimate of 2.700 billion bushels above USDA and leaning higher. However, we do feel we’re at a point where we can pause and await better perspective on South American crop prospects and additional Chinese buying interest before pushing our estimate higher yet.
We do not see a need for USDA to revise their 2020/21 corn for ethanol usage estimate of 5.050 billion bushels at this time as ethanol production has mostly been in line with “needed†pace over the last 6-7 weeks. We are slightly below USDA at 5.025 billion bushels but largely immaterial at this point. Slumping ethanol margins and continued weak gasoline demand amid the stagnant COVID recovery has raised warning flags of potential lower estimates to come, but for now we’re holding steady. Based on our 1st quarter feed/residual usage ideas discussed above, we’re estimating 2020/21 annual usage at 5.650 billion bushels vs USDA last at 5.700 billion. This is obviously subject to revision once we see the numbers next week.
All told, we’re estimating 2020/21 U.S. corn ending stocks at 1.605 billion bushels vs USDA last at 1.702 billion and 1.995 billion last year. Our stocks/usage ratio estimate of 10.8% compares to USDA at 11.5%, last year’s 14.4% and would be the lowest since 9.2% in 2013/14. The ability to maintain current price levels or potentially continue to move higher largely lies with the January 12 reports in our opinion. Balance sheet ideas are in a position were any bullish surprises could easily propel corn notably higher should ending stocks prospects move below 1.5 billion bushels. On the other hand, if somehow a notable amount of corn is “found†in the December 1 stocks report, downside price risk should not be underestimated given the considerable emotion and fund buying experienced on the way up.
Soybeans
Starting with the Crop Production report, we also see the potential for a modest downward revision from the USDA’s November estimate of 4.170 billion bushels (50.7 bu/acre yield). While the pattern of USDA monthly yield revisions this year doesn’t really fit any other years to provide some guidance, estimated yields so far have been a bit higher than the historical relationship with crop conditions would imply.
For reference, the following chart shows years since 1980 when USDA lowered the U.S. average soybean yield estimate in the September and November reports, regardless of the direction of revision in October. This year, USDA lowered soybean yields 1.4 bu/acre in September and 1.2 bu/acre in November, but left yields completely unchanged in October. As seen, yield revisions in the January Annual report have been a mixed bag in these with minor 0.1-0.4 bushel/acre increases in the most-recent year occurrences, with 0.1-0.4 bu/acre declines in earlier-year occurrences. Perhaps the one factor that stands out the most is that all January yield revisions in these years were quite limited.
Influencing our ideas of the U.S. average soybean yield being a bit lower than the USDA’s 50.7 bu/acre estimate is the extremely strong relationship between final crop conditions and final yields in recent years as shown in the following chart as a nearly perfect straight-line relationship has existed for the last three years’ crops. While we’re not saying we expect the U.S. average soybean yield to decline fully to 50.0 bu/acre as reflected as a possibility in the chart below, we do feel the risk bias in the revision is to the downside. Based on the limited yield revisions in January in the past, as discussed above, we’re looking for a modest yield cut to 50.4 bushels/acre, prompting a production decline to 4.147 billion bushels, down 23 million bushels from the USDA’s current estimate.
While last year’s Annual Crop Production report only saw a minor 8 million bushel revision in the U.S. soybean crop, to the upside, recent history has held a solid bias for moderate downward revisions being made in January, with the prior four years seeing 33-56 million bushel reductions.
Relative to trade expectations, last year’s minor upward revision in the January report resulted in the 2nd largest “bearish surprise†on record for the Annual Crop Production report for soybeans, but each of the previous four years’ reports were lower than expected. Over the last 10 years, there has been an even 50/50 split between higher or lower than expected January soybean production estimates.
Moving to the December 1 stocks report, needless to say, U.S. soybean demand during the first quarter of 2020/21 has been nothing short of incredible. Sept-Nov exports are estimated at 1.089 billion bushels (official Nov trade data will be published tomorrow (1/07/21) vs 611 million bushels last year and the previous record 1st quarter exports of 929 million bushels in 2016/17. Sept-Nov crush of 559 million bushels was also easily record high in surpassing 2018/19’s 531 million and was up from last year’s 524 million bushels. Combined, estimated total known 1st quarter usage of 1.648 billion bushels was up a massive 513 million bushels (45%) from last year’s 1.135 billion and 234 million bushels (16.5%) above the previous record of 1.414 billion bushels in 2016/17.
As for the ever-elusive quarterly “residual,†the record-high demand and significant increase in production (i.e. supplies) from last year notably increase the uncertainty heading into next Tuesday’s reports. The relationship between the 1st quarter “residual†and various fundamental factors is sketchy at best even in “normal†situations. Add in unprecedented demand (i.e. moving supplies) and potentially the largest year-over-year percentage increase in production in 15 years and the assessment risks of the 1st quarter residual increase substantially.
The previous two year’s 1st quarter residuals were quite small at 77 million bushels and 90 million bushels – not all that surprising in the context of the notably reduced demand in recent years. However, those residuals were still well below those of 2015/16 and 2010/11 which experienced similar demand structures. This raises the question of the possibility of an improvement in stocks accountability of late as not as many “missing bushels†were reflected in the last two years vs previous years with similar demand structures. On the other hand, recent unexpected substantial revisions to past quarterly Grain Stocks report numbers would counter that argument.
In an attempt to put some context on the 1st quarter residual possibilities, the following charts show the relationship (used loosely) between total 1st quarter demand and 1st quarter exports with the 1st quarter residual. The record high levels of both obviously push the charts into uncharted territory, but one can see a historically high residual certainly appears to be a possibility.
As seen, both would imply the possibility/likelihood of a 1st quarter residual of at least 200 million bushels, with past years of historically large 1st quarter demand seeing residuals of 186-248 million bushels. Accordingly, we’re heading into the quarterly Grain Stocks report with the assumption of the 1st quarter residual at 200 million bushels.
Based on these ideas, we’re estimating December 1 U.S. soybean stocks at 2.827 billion bushels vs 3.253 billion last year and the lowest in five years. The 1.72 December 1 stocks/Sept-Nov usage ratio would be down sharply from 2.87 last year and the lowest on record, which is of no surprise given the massive level of 1st quarter demand. However, as we move deeper into the 2020/21 marketing year, the fact this year’s balance sheet is fundamentally tighter than the 2012/13 and 2013/14 situations at this stage clearly justifies the sharp price acceleration and prospects for continued price gains IF demand continues to run at elevated levels.
The December 1 soybean stocks report has come in below the average trade estimate in 9 of the last 11 years, although last year’s report was the 2nd largest “bearish surprise†on record at 65 million bushels higher than expected.
Turning to the balance sheets, there’s no doubt USDA is between a rock and a hard place as the continued very strong demand clearly supports higher balance sheet demand estimates, while there is little room to reflect higher demand with stocks estimates already pushing historically low levels – a clear demand-rationing need which currently shows little signs of actually taking place.
U.S. soybean export commitments currently stand at 2.016 billion bushels, already reflecting 91.6% of the USDA’s 2.050 billion bushel export projection with 8 months remaining in the 2020/21 marketing year. Both measures of export demand are record high, with the previous record of sales on the books at this time being 1.763 billion bushels in 2016/17, while 2013/14’s 91.0% of late December sales vs eventual annual exports is the record for now.
Based on the level of commitments already on the books, we estimate soybean sales would only need to average roughly 5.8 million bushels/week through the end of August in order for 2020/21 exports to reach the USDA’s 2.200 billion bushel annual projection vs average Jan-Aug sales over the previous three years of nearly 19 million bushels/week. In fact, over the last 40 years, only three years have experienced average weekly soybean sales of 6 million bushels or less during Jan-Aug: 2013/14, 2003/04 and 1988/89. We clearly feel exports are likely to exceed the USDA’s current estimate, but again there is little room for an upward revision in the balance sheet. Nonetheless, we’re estimating 2020/21 U.S. soybean exports at 2.250 billion bushels, 50 million above the USDA, at this time, up from 1.676 billion bushels last year and surpassing the current record of 2.166 billion bushels in 2016/17.
As previously discussed, 1st quarter soybean crush was 559 million bushels vs 524 million a year ago, with each of the first three months of 2020/21 reflecting records for the respective month. More importantly, the USDA’s 2.195 billion bushel annual crush projection reflects an estimated 30 million bushel increase from last year’s 2.165 billion, with the annual estimated increase already being surpassed. Again, a clear demand rationing need moving forward. While that may ultimately be seen later in the marketing year, an increase in the USDA’s estimate certainly is warranted, but again with little room to do so. We’re currently estimating 2020/21 soybean crush 20 million bushels above USDA at 2.215 billion bushels.
All together, we’re estimating 2020/21 U.S. soybean ending stocks at 126 million bushels vs USDA last at 175 million and 523 million bushels last year. Our stocks estimate is only marginally above 2013/14’s record low 92 million bushels, while our 2.7% stocks/usage ratio estimate is nearly identical to the 2.6% record. The need for demand rationing cannot be emphasized enough and the way to do that is through higher prices. Clearly the market is sending that signal, but only time will tell if/how effective it has been. If, somehow, soybeans are “found†in the December 1 stocks figure, re-evaluation will be needed, but for now, price risk remains to the upside with balance sheet projections comparable to 2013/14 when soybeans traded north of $15.00. Assessing upside potential in these situations is extremely difficult, if not impossible, but as long as South American weather remains less than ideal, fund money continues pouring in, and export demand remains average or better, pull backs could/should be quite limited with additional gains likely.
Wheat
With the supply side of the 2020/21 U.S. wheat balance sheet already set, we can jump straight to the December 1 Grain Stocks, which is much less eventful for wheat than it is for corn and soybeans. Sept-Nov (2nd quarter) U.S. wheat exports are estimated at 235 million bushels, nearly identical to last year’s 233 million, and the 2nd best of the last seven years, but hardly noteworthy. Since 2014/15, U.S. wheat exports during the 2nd quarter have been consistently mundane, ranging from 192-239 million bushels. A 47 million bushel (1.3 MMT) range of exports for the quarter over a 7-year period clearly highlights the U.S.’ lack of influence in global trade.
If the steadiness of 2nd quarter exports appears surprising, the consistency of 2nd quarter U.S. wheat food usage is simply numbing. Over the last 10 years, Sept-Nov wheat food usage has ranged from 244-251 million bushels, statistically flat-lined for the last decade. Accordingly, Sept-Nov food usage is estimated at 248 million bushels vs 247 million last year.
Where some uncertainty does come into play is the statistical wiggle room of the balance sheet in the feed/residual. Almost always negative, the 2nd quarter residual partially offsets 1st quarter implied feed/residual usage, but the relationship between the two is lacking to say the least. Prior to last year’s +10 million bushel 2nd quarter residual anomaly, the Sept-Nov residual was negative for 33 consecutive years, ranging from -5 million bushels to -168 million bushels. The following chart shows 1st quarter vs 2nd quarter wheat feed/residual usage. While there is a bit of a direct relationship between the two (higher positive 1st quarter with larger negative 2nd quarter: smaller positive 1st quarter with smaller negative 2nd quarter), it is hardly statistically sound. Accordingly, with 1st quarter implied feed/residual usage of 215 million bushels vs 196 million last year and the highest in four years, we’re estimating the 2nd quarter residual at -50 million bushels. We would lean towards a bit larger negative residual, but last year’s unexpected positive 2nd quarter residual prompts a bit of uncertainty.
Based on our 2nd quarter demand views, we estimate December 1 U.S. wheat stocks at 1.713 billion bushels vs 1.841 billion last year and the lowest in six years, mirroring current annual balance sheet ideas for the lowest ending stocks since 2014/15. However, while more fundamentally constructive than Dec 1 stocks levels over the previous four years, our estimate hardly represents a fundamentally bullish scenario as seen in corn and soybeans. To the contrary, December 1 2020/21 U.S. wheat stocks are essentially middle-of-the road type numbers in the big picture.
Last year’s very unexpected positive 2nd quarter residual prompted the largest “bullish surprise†on record for December 1 wheat stocks, coming in 83 million bushels below the average trade estimate. However, outside of that, this report has held a notable bias for coming in higher than expected with 14 of the previous 17 years’ Dec 1 stocks figures being above the average estimate by some degree.
Turning to the balance sheet, U.S. wheat export demand remains respectable, but far from balance sheet changing. Total commitments of 756 million bushels are up decently from last year’s 693 million and are the highest for late December of the last four years, but again middle of the road in historical perspective. Similarly, sales on the books represent 77% of the USDA’s 985 million bushel annual export projection, better than in recent years, but below the 80-85% best levels over the last 15 years.
Additionally, based on the level of commitments already on the books, we estimate wheat sales will need to average roughly 11 million bushels/week during Jan-May in order for 2020/21 exports to reach the USDA’s 985 million bushel annual projection, which would be below the last two years’ 13.5 million bushels/week average, but right in line with the most-recent 5-year average from this point forward. We see the potential for exports to exceed the USDA’s current projection, but likely only slightly/modestly so as we’re maintaining our estimate marginally higher at 990 million bushels at this time.
Based on our 2nd quarter feed/residual usage ideas, we’re estimating 2020/21 annual usage at 95 million bushels, marginally below the USDA’s 100 million. If Dec 1 wheat stocks come in mostly in line with our expectations, the 3rd-4th quarter combined feed/residuals would need to be roughly -70 million bushels based on our annual estimate vs last year’s -104 million, which was the largest negative of the last 12 years.
We’re currently estimating 2020/21 U.S. wheat ending stocks at 857 million bushels, little different than USDA’s 862 million bushel estimate and down from last year’s 1.028 billion. Again, yes, U.S. wheat ending stocks are estimated to be the lowest in six years, but are considered historically average and hardly price-rally worthy as is happening in corn and soybeans. Wheat has been reluctantly pulled along by the strength in the other markets, but extended gains do not appear fundamentally justifiable in our opinion unless something dramatically alters demand for U.S. wheat, which is not anticipated at this time.
Perhaps the wheat market has the most to look forward to in next Tuesday’s numbers in the Winter Wheat Seedings report. While not likely to be game-changing in nature, it will provide the first true fundamental numbers for the 2021/22 picture. The modest strength in wheat prices last fall, and encouraging overall demand ideas at the time as export interest really started to gear up, mostly for corn and soybeans, but also for wheat amid the unknowns of China, is expected to have prompted the first increase in winter wheat area in eight years. We’re estimating total winter wheat planted area at 31.4 million acres vs 30.415 million last year, with hard red winter wheat area at 22.1 million acres vs 21.36 million last year, soft red winter wheat at 5.8 million vs 5.56 million last year and white at 3.5 million vs 3.49 million last year.
On a final note, after an unprecedented stretch for an individual USDA report to consistently come in the same way relative to market expectations, the Winter Wheat Seedings report has broken tradition in recent years. For more than 20 years, the Winter Wheat Seedings report failed to come in above the average trade estimate a single time from 1989 through 2011, before doing so in 2012, but again returning to lower than expected reports from 2013-2017. However, two of the last three years have been above the average estimate, although last year’s by a minimal 100k acres.