Next week (Thursday March 31) USDA will release the quarterly Grain Stocks report. Typically it is overshadowed by the Prospective Plantings report released on the same date. However, as Todd Gleason reports, it occasionally provides a surprise to the trade.
For soybeans, the stocks estimate is often very near the level expected by the market says University of Illinois Agricultural Economist Darrel Good. This is because we generally know how many soybeans are used at any point during year based off the magnitude of the domestic crush and the exports, both of which are tallied either by the government, the industry, or the two combined. The stocks estimate, says Good, really does indicated the magnitude of seed, feed, and residual use of soybeans in the previous quarter. Unlike corn, for which feed and residual use is a large portion of disappearance, seed, feed, and residual use of soybeans is a relatively small portion of disappearance during the winter months. However, he cautions, occasionally the March 1 stocks estimate provides a surprise.
Based on the average trade guess reported by news services, the March 1 stocks estimate has deviated from market expectations by more than 30 million bushels nine times and by more than 60 million bushels four times in the past 25 years.
The expected level of soybean stocks on March 1 this year can be calculated. The USDA’s Oilseed Crushings, Production, Consumption and Stocks report provides information for December of 2015 and January of 2016. The estimate for February will be released April 1. The National Oilseed Processors Association (NOPA) estimate of the magnitude of the February soybean crush by its members can be used to estimate the total February crush. For the nine months that USDA has provided soybean crush estimates (May 2015-January 2016), the USDA crush estimates have exceeded the NOPA crush estimates by 6.4 percent. Applying that ratio to the NOPA February crush estimate, suggests to Darrel Good that 483.1 million bushels of soybeans were crushed in the second quarter of the current marketing year. It’s possible to calculate the number of soybeans exported in the last quarter, too.
Based on a combination of USDA and Census Bureau export estimates, second quarter exports totaled just over 677 million bushels.
This leaves the seed, feed, and residual usage factor. That’s tougher to figure, but a much smaller number. If this year follows the average consumption pattern Good says it would be about 12 million bushels in the second quarter. So, 483 crushed plus 677 exported plus 12 fed equals roughly 1.173 billion bushels consumed in the second quarter. Subtract that from the first quarter stocks, plus the imports and you get 1.55 billion bushels of soybeans on hand March 1st in the United States. The Grain Stocks report March 31 shouldn’t vary much from this number, but it could says Darrel Good.
If the March 1 stocks estimate is surprisingly large or small, the accuracy of USDA’s 2015 production estimate may be called into question. The USDA has revised the previous year’s production estimate by varying amounts in 20 of the past 25 years based on the stocks estimate at the end of the marketing year (September 1). However, it would be pre-mature to question the accuracy of the production estimate based on the March 1 stocks estimate due to the large variation in the quarterly pattern of seed, feed, and residual use of soybeans.
The eight largest revisions in the production estimates following the USDA’s September 1 stocks estimate ranged from 1.1 to 3.5 percent. Only three of those eight large revisions followed a surprise in the March 1 stocks estimate that exceeded 30 million bushels. Conversely, of the nine years in which the magnitude of the surprise in March 1 stocks estimate exceeded 30 million bushels, only three were followed by revisions in the production estimate that exceeded one percent.