The USDA’s U.S. corn and soybean balance sheet revisions failed to provide any new and meaningfully-supportive developments, allowing for at least a temporary pull-back from the recent constructive price action. All eyes remain on the weather. A larger-than-expected cut in the U.S. wheat crop provided more fund-fodder, allowing CBOT wheat to move to the highest levels since late April.
A summary of key report numbers relative to market expectations is attached to this post as a PDF file.
Corn
USDA didn’t provide much in the way of surprises in this month’s U.S. corn balance sheet revisions with feed/residual usage and corn for ethanol usage lowered, albeit both by less than ultimately will likely be needed, while exports were held steady for the time being, but will likely need an upward revision at some point. The new crop balance sheet revisions essentially fell in line, as the considerable reduction in this year’s “working†crop estimate was known to be coming following the June 30 Acreage report, while feed/residual usage was lowered given the reduction in old crop. The USDA’s old and new crop ending stocks estimates were both slightly below the average trade estimates, but well within the realm of expectations and were largely non-events from a psychological perspective. Fundamentally, the implications of the lower-than-expected acreage this year does put an increased onus in achieving respectable yields, but the balance sheet is still far from anything considered concerning at this point.
USDA lowered 2019/20 U.S. corn feed/residual usage by 100 million bushels to 5.600 billion, following suit with previous years in which June 1 corn stocks deviated considerably from expectations in reflecting a more modest demand revision than the stocks report would imply outright, as we discussed was likely in our pre-report commentary. This is a similar situation to 2010/11 when June 1 stocks were 346 million bushels larger than expected while feed/residual usage was lowered 150 million bushels in the July WASDE report and 2015/16 with June 1 stocks 194 million bushels higher than expected, followed by a 50 million bushel feed/residual usage cut in July. Both of those years saw final feed/residual usage prove lower than the July estimate – 2010/11 by 222 million bushels and 2015/16 by 90 million bushels. However, we would note USDA did not revise feed/residual usage in the August or September WASDE reports in both of those years, instead choosing to wait until after the end of September Grain Stocks reports and making revisions in the October/November WASDEs. Accordingly, we’re maintaining our 2019/20 feed/residual usage estimate of 5.450 billion bushels, 150 million below the USDA, but not expecting it to be reflected in visible balance sheets through the end of the marketing year unfortunately.
USDA took another step in the right direction in further lowering their 2019/20 corn for ethanol usage estimate by 50 million bushels to 4.850 billion bushels, now reflecting an estimated 528 million bushel decline from last year’s 5.378 billion. Based on the new estimate, we see ethanol production needing to average roughly 6% below last year’s level through the end of August vs the latest week’s production reflecting a 12.7% year-over-year decline. With year-to-date corn for ethanol usage estimated down 475 million bushels, there’s only room for an additional 53 million bushel loss over the last eight full weeks of the year, or roughly 6-7 million bushels/week. The recovery in ethanol production continues and margins are solid at the moment, but we still feel the USDA’s estimate is a bit optimistic. We’re maintaining our 2019/20 corn for ethanol usage estimate at 5.825 billion bushels, another 25 million below the USDA.
USDA left their 2019/20 U.S. corn export estimate unchanged this month at 1.775 billion bushels, but is increasingly appearing in need of upward revision, particularly following this morning’s USDA announced sale of 765k tonnes (30 million bushels) in old crop corn to China, as part of the 1.365 MMT daily announcement (600k tonnes new crop). Based on the differences between official Census Bureau data and Export Sales data, we feel there are already enough sales on the books to reach the USDA’s current projection, leaving us to maintain our 2019/20 export estimate 50 million bushels above USDA at 1.825 billion bushels.
All together, USDA raised 2019/20 U.S. corn ending stocks by 145 million bushels today to 2.248 billion, which was 39 million bushels less than the average trade estimate of 2.287 billion bushels reflected. Based on our lower feed/residual and corn for ethanol usage ideas and modestly higher exports, we see old crop corn ending stocks at 2.368 billion bushels, 120 million above USDA but feel we are unlikely to see the USDA’s ending stocks estimate above 2.300 billion bushels “when it matters†as lower feed/residual usage likely won’t be reflected until after the September Grain Stocks report when we’re moving into harvest. Higher USDA export revisions could be offset by lower corn for ethanol usage numbers in the coming months, keeping their stocks estimate near today’s estimate in WASDE reports through the end of the actual marketing year.
For new crop, USDA made the expected supply side revisions as a result of the June 30 Acreage report, with the 2020/21 assumed U.S. corn crop revised down 995 million bushels to 15.000 billion, amid the reflection of planted and harvested acreage being lowered 5.0 million and 5.6 million acres, respectively, from previous ideas. The average yield was left unchanged, as expected, at 178.5 bushels/acre.
For demand, USDA made a corresponding reduction in 2020/21 feed/residual usage of 200 million bushels to 5.850 billion, maintaining expectations for a modest/decent 250 million bushel increase from this year, but left their new crop corn for ethanol usage and export estimates unchanged at 5.200 billion and 2.150 billion bushels, respectively. We’re a bit more optimistic on ideas of both, but lower on feed/residual usage given our lower old crop ideas than USDA. The USDA lowered 2020/21 U.S. corn ending stocks to 2.648 billion bushels, a 675 million bushel cut from last month’s 3.323 billion, which was a bit larger than the average trade estimate of 2.715 billion bushels reflected, but well within overall ideas. For the most part, it appears the new crop balance sheet revisions were “as expected†but ending stocks ended up a bit below the average trade estimate due to old crop stocks not being revised up as much as anticipated.
As we head into the heart of the growing season, with the weather watch on high alert, our first deep dive into 2020/21 crop ideas based on current state-level crop conditions resulted in a production estimate of 14.910 billion bushels with an average yield of 176.5 bushels/acre, 90 million bushels and 1.5 bu/acre below the USDA’s current assumption. We feel the USDA’s harvested acreage estimate is a bit too low as their 91.3% harvesting percentage is below any of the years from 2013/14-2018/19 (91.5-92.3% range) and the average of the six years of 91.8%. We’re assuming a 91.8% harvesting percentage, resulting in harvested area 440k acres larger than USDA.
Based on our new crop production and demand ideas, we’re currently estimating 2020/21 U.S. corn ending stocks at 2.718 billion bushels vs USDA at 2.648 billion and this year’s 2.368 billion (USDA 2.248 billion). Our 18.6% stocks/usage ratio estimate would be up from this year’s 17.5% and the highest since 2004/05. As long as a “dome of doom†doesn’t develop across the heart of the corn belt, it appears difficult to get overly excited about the new crop balance sheet. Demand, of course, is uncertain. The recent notable uptick in Chinese buying interest of U.S. corn adds a risk factor. However, we’re already assuming a 350 million bushel (8.9 MMT) increase in exports in 2020/21. Even pushing that up to a record export program of 2.450 billion bushels (625 mil bu/15.9 MMT increase from this year) still leaves ending stocks around 2.450 billion bushels with our 176.5 bu/acre yield estimate. If a true crop problem does develop, and yields fall to around 170 bu/acre, ending stocks ideas would drop to around 2.150 billion bushels under our current new crop demand ideas and around 1.900 billion bushels if a record export program is tacked on top of that, as well. Under the right situation, the 2020/21 U.S. corn balance sheet could get “interesting,†but in our opinion, it’s going to take the combination of several debatable developments to all occur.
World revisions were mostly uneventful with USDA raising Argentine corn exports by 2 MMT, while lowering Brazilian exports by 1 MMT and leaving both crop estimates unchanged. Old crop Chinese corn ending stocks were lowered 3 MMT due to higher feed usage, with new crop stocks lowered 4 MMT on the carry-through of the old crop revision and a 1 MMT increase in feed usage, as well. World ending stocks were revised down more than expected, but due mostly to the U.S. situation.
Soybeans
USDA’s U.S. soybean balance sheet revisions were quite limited this month, with old crop and new crop exports left unchanged at 1.650 billion and 2.050 billion bushels, respectively, and modest 15 million bushel upward revisions in crush for this year and next year to 2.155 billion and 2.160 billion bushels, respectively. With the record crush in May of 179.6 million bushels easily beating the previous record for the month of 172.5 million in 2018, while being 8.5% above last year’s May crush of 165.4 million, it easy to understand their motivation for the increase. Based on the new estimate, though, June-August crush will need to total 527 million bushels, 2.3% above last year’s 515 million and setting a new record in beating 2017/18’s 518 million bushels. The crush pace through May certainly justifies the USDA’s upward revision today, but crush margins running below each of the last six years’ levels since early June raises a few questions. NOPA’s monthly crush report of June data, out next Wednesday, will be quite interesting to see if it supports today’s USDA revision.
The one interesting revelation in today’s U.S. soybean balance sheet was their revision in the 2019/20 residual to -46 million bushels from +4 million last month. Following the June 30 Grain Stocks report, we wrote about the unusual situation of reported June 1 stocks not falling in line with the typical relationship with “known†usage in the quarter, with USDA using the not-often-used negative residual as a “placeholder†until the September Grain Stocks report hopefully provides a bit more clarity. Remember, in the September Grain Stocks report, USDA can and does revise last year’s crop if known usage and September 1 stocks do not agree.
There have been three years in the past when USDA reflected a negative residual late in the old crop marketing year after the June stocks report, 2013/14, 2007/08 and 2003/04. In 2013/14, USDA assumed a -69 million bushel residual in July and further lowered it to -94 million in August and held it there for the September WASDE, as well. In 2007/08, USDA had a -35 million bushel residual in the July, August and September WASDEs. The only other negative residual was -15 million bushels in September for the 2003/04 marketing year.
In all three of those years, USDA revised the crop higher in the September Grain Stocks report, with 2013/14 raised 68 million bushels, 2007/08 raised 92 million bushels and 2003/04 raised 36 million bushels. This clearly points to the likelihood that last year’s current 3.552 billion bushel crop estimate is too low and will be revised higher at the end of September. Each one of those years ended up with a historically low residual (+8 mil bu 2013/14, zero 2007/08, 18 mil bu 2003/04). Over the next month or so, as likely/potential 2019/20 total crush and exports become more clear, we will be able to hone in on the likely upward revision in last year’s crop. Based on the USDA’s current balance sheet ideas for demand and ending stocks, it appears last year’s crop would need to be revised higher by 50-55 million bushels to just over 3.600 billion bushels. How crush and exports play out through the end of year will affect that calculation. Based on our current old crop demand ideas, we would peg last year’s crop around 3.580 billion bushels.
With the negative residual, USDA raised 2019/20 U.S. soybean ending stocks to 620 million bushels from 585 million last month, reflecting a solidly larger increase than expected as reflected by the average trade estimate of 588 million bushels. Based on our re-assessed view of last year’s crop as it relates to our old crop demand ideas, though, our 2019/20 ending stocks estimate didn’t change much from our pre-report ideas of 613 million bushels, as we’re now at 616 million.
For 2020/21, the minor revisions to acreage as a result of the June 30 Acreage report prompted a mere 10 million bushel increase in their production assumption to 4.135 billion bushels, as expected, with the yield assumption of 49.8 bushels/acre left unchanged. With the increase in crush of 15 million bushels partially offsetting the 45 million bushel increase in total supplies (10 mil production, 35 mil beginning stocks), USDA raised 2020/21 U.S. soybean ending stocks by 30 million bushels to 425 million bushels, which was in line with the average trade estimate of 424 million bushels.
As with corn, we’ve started taking a closer look at 2020/21 production prospects based on current crop conditions. Unlike corn, though, our initial view of this year’s soybean crop is higher than USDA, as our state-level yield ideas worked to a U.S. average yield of 51.0 bushels/acre for a crop of 4.237 billion bushels vs USDA’s current 4.135 billion bushel working estimate. Again, this is based on current crop conditions, with the assumption those conditions hold steady throughout the growing season. If conditions begin to worsen, obviously crop ideas would decline, but vice versa, as well. As we’ve shown in our Monday afternoon Crop Progress updates on Market Insights, the overall rating of the U.S. soybean crop is just above 2018’s for the best since 2014 and the 2nd best since 1999. Accordingly, early yield ideas/prospects are encouraging.
We are much more optimistic on new crop export demand than USDA, as we’re at 2.125 billion bushels vs their 2.050 billion bushel estimate, while we are in line with their ideas on crush. The larger export estimate largely offsets our higher crop ideas at this time, but not entirely, leaving our 2020/21 U.S. soybean ending stocks estimate at 453 million bushels vs USDA at 425 million. Fundamentally, the soybean balance sheet has improved markedly from two years ago. However, stocks remain more than comfortable from a historical perspective, leaving the development of a crop problem or much stronger demand being needed to push things towards uncomfortable territory.
For the products, USDA made offsetting old crop soybean oil demand revisions, with a much-needed 150 million pound increase in exports being offset by a 150 million pound reduction in domestic usage, which was not biodiesel related at it was left unchanged. The higher crush resulted in 2019/20 production being raised 105 million pounds and ending stocks being raised by 80 million pounds to 2.020 billion, slightly offset by imports lowered 25 million pounds. For new crop, total supplies rose 255 million pounds due to the higher beginning stocks, along with production raised 175 million pounds. This was largely offset by a 200 million pound increase in new crop exports, as well, to 2.400 billion pounds, leaving 2020/21 ending stocks being revised up 55 million pounds to 2.050 billion.
Soybean meal saw a 300k ton increase in old crop production offset by a corresponding increase in domestic usage, while a 450k ton increase in new crop production was offset by a 300k ton increase in domestic usage and 150k ton increase in exports.
Globally, USDA raised the Brazilian soybean crop by 2 MMT to 126.0 MMT, going against the average trade estimate of 123.3 MMT reflecting an expect small decline, while the Argentine crop was left unchanged at 50.0 MMT. USDA raised this year’s Brazilian soybean exports to 79.5 MMT (local marketing year basis) from 77.0 MMT last month (73.4 MMT last year), while the Argentine balance sheet was unchanged. USDA raised 2019/20 Chinese soybean imports by 2 MMT to 96.0 MMT, but left new crop imports unchanged at 96.0 MMT, as well.
Wheat
Possibly the most surprising part of today’s USDA report was the fact the winter wheat revision stole the focus from first objective estimate of this year’s U.S. spring wheat crop. The first estimate of the year always holds a bit of an added risk as USDA’s initial field-based assessments can provide unexpected surprises. That was not the case this time, though, as USDA estimated the “other spring†wheat crop at 550 million bushels, exactly in line with the average trade estimate of 549 million bushels and nearly unchanged from last year’s 562 million. The hard red spring wheat crop was estimated at 502 million bushels vs 522 million last year. The USDA’s durum crop estimate of 56 million bushels was also in line with the average trade estimate of 57 million and little-changed from last year’s 54 million. The “other spring†wheat average yield of 46.6 bushels/acre is down from last year’s 48.2 due largely to North Dakota’s estimated 4 bu/acre decline from last year to 45 bu/acre. With the first improvement in spring wheat crop conditions last week for the first time in four weeks, along with the improved rains of late, the crop may move a bit higher in updates to come.
The USDA’s lowering of the total winter wheat crop to 1.217 billion bushels from 1.266 billion last month caught the market off-guard, as the 49 million bushel cut was much more than expected as reflected by the average trade estimate of 1.246 billion bushels, down 20 million from last month, and was even below the lowest trade estimate of 1,223 billion bushels. Additionally, the decline from June matched the 2nd largest reduction in the winter wheat crop in the July report since 2004, with 2015’s 50 million bushel reduction being an inconsequential 1 million bushels greater. Hard red winter saw the brunt of the reduction, with a 33 million bushel cut from June to 710 million bushels, while the soft red winter wheat crop was lowered a solid 17 million bushels, as well, to 280 million. The white winter wheat crop was ticked up by 2 million bushels. As seen in the table below, nearly all major producing states saw yields lowered from last month.
With the decline in the winter wheat crop, USDA put the 2020/21 U.S. all wheat crop at 1.824 billion bushels vs the average trade estimate of 1.848 billion bushels last year’s 1.920 billion. Based on our slightly higher spring wheat crop ideas, we’re using 1.839 billion bushels in our balance sheet at this time, putting 2020/21 total supplies at 3.013 billion bushels vs 3.105 billion last year and the lowest since 2015/16. U.S. wheat supplies remain comfortable from a historical perspective, but clearly have tightened up from levels of the last few years.
The USDA cleaned up the 2019/20 balance sheet in expected fashion, with the June 30 Grain Stocks report reflection of June 1 stocks at 1.044 billion bushels, vs the USDA’s previous ending stocks estimate of 983 million bushels, being accommodated by a 61 million bushel cut in last year’s feed/residual usage to 74 million bushels from 135 million previously. This was the only revision to the old crop balance sheet as May official trade data verified their previous 965 million export projection.
For 2020/21, the 61 million bushel increase in beginning stocks was largely offset by the larger-than-expected 53 million bushel decline in production, leaving the USDA’s total supplies estimate to rise only 7 million bushels from June. The only demand revision for new crop was a minor 10 million bushel reduction in feed/residual usage to 90 million bushels. With the large cut in old crop feed/residual usage, each of the last three years’ feed/residual being 51-90 million bushels, this year’s wheat crop being down moderately from last year and U.S. corn supplies estimated at a record high, we’re surprised the reduction in 2020/21 feed/residual usage wasn’t larger. We’re using 70 million bushels in our balance sheet at this time.
The net result for the 2020/21 balance sheet was a 17 million bushel increase in ending stocks from last month to 942 million bushels, which was slightly below the average trade estimate of 954 million bushels, and reflects a 102 million bushel decline from 2019/20. We’re estimating exports a modest 25 million bushels higher than USDA at 975 million, but offset by lower domestic usage ideas. We’re estimating 2020/21 U.S. wheat ending stocks at 948 million bushels, largely in line with USDA.
The USDA issued their first by-class balance sheet estimates of the year today, as well, which can be seen at the end of this post. The most-notable year-over-year change is for hard red winter with 2020/21 ending stocks estimated at 423 million bushels, down nearly 100 million from last year’s 521 million, which should come as no surprise with production down 123 million bushels from 2019/20. While down solidly from last year and the lowest since 2015/16, HRW stocks at that level are still quite substantial from a historical perspective, having ranged from 138-387 million bushels over the 14-year period from 2001/02-2014/15. This year’s SRW ending stocks were estimated at 103 million bushels, essentially unchanged from last year’s 105 million, with HRS stocks put at 270 million bushels vs 280 million last year. Durum stocks are also near unchanged at 40 million bushels vs 43 million last year.
Overall, the U.S. wheat balance sheet continues to constructively tighten, but is still nowhere near levels considered concerning. Demand simply needs to prove stronger than expected to create any true concern for the balance sheet. U.S. wheat exports have ranged from 778-965 million bushels in five of the last six years and it will take an export program stronger than that to generate much in the way to true fundamental interest this year. Today’s strength in CBOT wheat looked heavily fund-influenced, once again, as they have been estimated net buyers in seven of the last nine trading sessions.
USDA’s global revisions were fairly limited this month, with the EU wheat crop necessarily lowered again by 1.5 MMT to 139.5 MMT (154.9 MMT last year) and exports lowered 1 MMT to 27.0 MMT (38.0 MMT last year). Russia’s crop was ticked lower by 0.5 MMT to 76.5 MMT (73.6 MMT last year), but exports were left unchanged at 36.0 MMT (34.5 MMT last year). Other major exporters were left unchanged.
USDA Monthly Supply/Demand Balance Sheet Revisions
