Industry analysts seem to believe that the U.S. Department of Agriculture overstated new-crop soybean demand for the United States at its annual outlook forum in February, and this has led to a great deal of uncertainty ahead of the big crop report.

On Wednesday at noon EDT, USDA will publish its highly anticipated first look at world supply and demand outlook for the upcoming 2017-18 marketing year, and the U.S. soybean table will be one of the most closely watched items.

Last week, Reuters polled 19 analysts on their view of where USDA will place U.S. soybean carryout on Wednesday for both the 2016-17 and 2017-18 years.

On average, the analysts expect 2016-17 carryout will be reduced to 438 million bushels from the current 445 million, and that 2017-18 carryout will land at 555 million bushels.

If realized, the new-crop assumptions would be USDA’s highest initial carryout peg since the 2006-07 projection of 650 million bushels back in May 2006. Only one other initial projection in the past two decades (1999-00; 595 million bushels) was higher than the market’s current thinking. The 2006-07 marketing year eventually turned out the largest carryout on record - some 574 million bushels.

In May 2016, USDA revised the 2015-16 soybean carryout down to 400 million bushels, though that number ended up at 197 million by October. This was very close to the 2014-15 ending stocks of 195 million, which began at 330 million.

Analysts have been reminding the market all year of this recent late-season carryout trimming trend, often citing the rock-solid demand to support the idea of its possible return in 2016-17.

But a trend reversal has yet to be observed this year, which could set the stage for near-record soybean supply in the United States in 2017-18. And based on the sizable market predictions of new-crop carryout, analysts’ confidence in the unbreakable demand narrative has started to waver.

Number Crunching

The U.S. soybean harvest was unexpectedly large in 2016-17 at 4.3 billion bushels, and 2017-18 expectations place that crop very close at 4.25 billion bushels. The market should already be in tune with what USDA will print for new-crop production on Wednesday since the agency has already published trend yield and planting intentions earlier in the year, and they do not typically change for the May report.

Two potentially back-to-back monster crops have made it very hard for analysts to pare down their view of soybean stocks, which weigh heavily on Chicago soybean futures as supply expands. USDA also could not find justification to cut down on new-crop supply when it published its preliminary balance sheet for 2017-18 in conjunction with its outlook forum in February, as it placed the carryout at 420 million bushels.

The minor categories on USDA’s balance sheet – such as imports, seed and residual – typically will not change much at all between February and May. Assuming this and combining the implied new-crop production with the average poll estimates for old- and new-crop carryout, total soybean use in 2017-18 would be 4.16 billion bushels.

This is above USDA’s current 2016-17 figure of 4.08 billion bushels, but it is lower than the agency’s February new-crop projection of 4.21 billion. This suggests that analysts are less confident on 2017-18 U.S. soybean crush and exports – the two uses of the oilseed in the country – than they were a couple of months ago.

If USDA were to maintain the same fractions of use as it did in February, total use at 4.16 billion bushels implies 2017-18 exports around 2.1 billion bushels and crush at 1.92 billion. USDA’s February new-crop numbers were 2.125 billion and 1.945 billion bushels, respectively.

While not all analysts are applying a bearish feel to demand, some think that new-crop exports and crushings will fall below this year’s levels in Wednesday’s report. At least four of the 18 analysts who provided estimates for both years appear to believe USDA will target 2017-18 total soybean use below its current stance for 2016-17.

On the other hand, a couple of analysts see USDA placing total soybean use in 2017-18 some 5 percent above the current 2016-17 projection, demonstrating just how wide of a view the industry has on the domestic soybean market going forward.

In the past six years, the maximum that USDA’s May projection of total soybean use has ever deviated from the preliminary numbers in February was 1.9 percent to the upside in 2016-17, and 1.5 percent to the downside in 2012-13.

External Factors

With world soybean stocks expected to maintain at high levels, the excess supply will have to be assigned somewhere – most likely to the United States or its Southern Hemisphere counterpart.

Analysts expect that USDA will raise 2016-17 world soybean ending stocks to an all-time high of 87.53 million tons from 87.41 million last month. They also predict the agency will peg 2017-18 at 86.59 million tons – a slight decline on the year but still a hefty number, well above the 77.13 million from 2015-16.

China, the world’s leading soybean buyer, can cut into some of these stocks. The country has notched record monthly imports for each of the first four months of 2017, and the total volume for the period is up 18 percent on the year.

USDA has Chinese imports at 88 million tons for the current marketing year, but based on the recent activity this could be revised upward. In late March, USDA’s Beijing attaché placed 2017-18 Chinese demand at 89 million tons, but history suggests this number is likely to be higher on Wednesday.

Another factor leading the U.S. soybean demand discussion is the abnormally slow pace at which Brazilian farmers have been selling their bin-busting crop based on lower-than-expected profits. As of May 5, farmers there had sold 50 percent of the crop compared with 67 percent one year ago and a five-year average of 65 percent.

Brazil has produced a record-large crop of at least 111 million tons – meaning if local farmers are unwilling to sell for long enough, the excess global stocks could end up in Brazil. The world’s largest soybean exporter is not typically known to hold on to their beans, making the recent hoarding trend unusual.

Through the end of April, the United States had sold a larger volume of soybeans on the 2016-17 books than USDA’s 2.025 billion bushel export projection for the same year. This, along with the slow Brazil sales, makes the case for an increase to U.S. old-crop exports on Wednesday and could potentially bolster new-crop exports too, pending how long USDA believes the Brazilian farmer might hold out.