FORT COLLINS, Colo. (Reuters) – U.S. quarterly stock data from the Department of Agriculture has long been known to create waves in the market, but the recent reports have felt tsunami-like due to some unusually large and unexpected adjustments to previous numbers.
Wednesday could be a volatile trading day as the quarterly stocks data will be published along with U.S. planting intentions, which are highly anticipated this year since the last two corn and soybean harvests came up short of expectations. USDA will publish the reports on Wednesday at noon EDT (1600 GMT).
Analysts peg March 1 U.S. corn stocks at 7.767 billion bushels, a six-year low for the date and 8% below the three-year average. March 1 soybean stocks are seen at 1.543 billion bushels, a five-year low for the date and 35% below the three-year average.
U.S. wheat stocks as of March 1 are estimated at 1.272 billion bushels, a six-year low and 15% below the three-year average.
Market participants over the last year or so have been somewhat demoralized by the quarterly stock reports, as revisions to past data have been much larger than usual, throwing off the trade estimates and increasing traders’ exposure.
Most analysts derive their stock predictions with assumptions about use and disappearance during the quarter, but uncertainty over the prior stock levels has added extra, unwelcome difficulty to the estimation process.
For U.S. corn and soybeans, the December-February period is the second quarter (Q2) of 2020-21 and for wheat it is the third quarter.
The analyst guesses peg Q2 soybean use at a record 1.39 billion bushels, some 39% above last year and up 36% on the three-year average. That also tops the previous high for the quarter, set in 2014-15, by 16%. Q2 soybean exports and crush were about 20% and 10%, respectively, above the 2014-15 levels.
Q2 corn use is implied at 3.55 billion bushels, up almost 3% from the recent three-year average. Exports during the period should contend for a 30-year high, as they were up around 50% from the recent average.
But Q2 ethanol output, which accounts for more than twice the usage as exports, was down between 10% and 11% from the three-year average as the pandemic continues to weigh on fuel demand.
The feed and residual term on the corn balance sheet can make the corn stocks much harder to predict than for soybeans, since the two are lumped together. USDA’s official estimates show 2020-21 feed and residual falling 4% from the prior year, though some market participants remain skeptical of the elevated 2019-20 number.
USDA on Thursday said the U.S. swine herd was 2% smaller on March 1 compared with a year earlier, below expectations for similar year-to-year figures. March 1 cattle on feed was 2% above the year-ago levels, though chick placements for poultry meat production since January are down about 2% on the year.
NASS has said the most common reason for previous quarter stock adjustments is the inclusion of late-reported data, but that usually results in small moves. It is less common for the re-analysis of past data to indicate that a revision is needed, but those moves can be larger.
Tweaks to Dec. 1 corn stocks in the March report have been bigger than normal in the last three years. The 2017 number was raised by 50 million bushels in March 2018 and Dec. 1, 2018, corn supplies were lowered nearly 16 million bushels in March 2019.
The Dec. 1 corn revision was the largest last year, though that came after an about-face. In March 2020, Dec. 1, 2019, stocks were increased by more than 13 million bushels, but then 75 million bushels were lopped off in the June report. The current Dec. 1, 2019, figure is more than 61 million bushels below the one originally reported in January 2020.
Recent March adjustments of Dec. 1 soybean stocks have been relatively less impactful than those for corn. Over the last four years, “final” Dec. 1 soy stocks have ended up on average just under 5 million bushels larger than originally reported. Similar to corn, Dec. 1, 2019, soy stocks were adjusted in both March and June last year.
These Dec. 1 revisions are smaller than some of the other ones recently observed, especially for corn. June 1, 2020, corn stocks are listed as 221 million bushels smaller than originally stated, and Sept. 1, 2020, stocks are 76 million bushels below the initial number. Sept. 1, 2019, corn stocks ended up 106 million bushels heavier than reported in September 2019.
Luckily for the market, eventual March 1 corn stock revisions have been by far the smallest of any quarter, but that does not rule out an abrupt change in trend come June.
The trade is coming off two abnormally bullish stock reports for corn. The average analyst guess on Dec. 1, 2020, corn stocks was nearly 6% too high, while the Sept. 1, 2020, guess was 13% too high (17% if considering the adjusted Sept. 1 number).
Analysts’ June 1, 2020, corn prediction was 5% too low on report day, unusually bearish for that quarter, but it was only 1% too light compared with the revised June 1 figure.
In the latest quarters, analysts were within 1% of the June 1 and Dec. 1 soybean stocks, but their guess of the Sept. 1, 2020, figure was 10% too heavy, producing bullish price action.
March 1, 2020, corn stocks came in below analyst guesses, but the trade overestimated that number in the five prior years. The market is on a four-year streak of lowballing March 1 soybean stocks, though the 2020 miss was by less than 1%.
The trade bias on March 1 wheat stocks is more mixed, but the published number was bearish relative to the estimates in three of the past five years.
Editing by Matthew Lewis