Rain-delayed planting and hot, dry growing weather in the U.S. Midwest resulted in the government reducing its corn-crop estimates in 2010 and 2011 by the most for any two-year period since 1984.
The U.S. Department of Agriculture’s final production estimate last year was down 7.4 percent from its initial forecast in August, and this year the drop was 4.7 percent. Output was overstated by 25.5 percent in August 1983 and 0.2 percent a year later, USDA data show, after a period of adverse weather. In 2004 and 2005, the department understated the size of the crops by an average of 7.2 percent each year, as cool conditions and ample rains resulted in higher yields.
“The forecasts have had a margin for error much larger over the past few years, but that’s a reflection of the adverse weather,” Peter Meyer, the publisher of OpeningPrint LLC and a former agricultural specialist for JPMorgan Chase & Co., said in a telephone interview. In each of the past two years, Meyer accurately forecast smaller crops in August than the government predicted.
This year, the USDA has cut its forecast four times in as many months, leading critics such as Darrel Good, a University of Illinois agricultural economist, to question the reliability of the government estimates. The price variability of corn futures in Chicago is about seven times higher on days the USDA released crop-production reports than for other trading sessions, Good said in a report earlier this year.
The biggest two-year variation between the USDA’s initial forecasts and final crop estimates occurred in the early 90s. In 1993, flooding left final production 17.1 percent below the August projection, and in 1994, cool, wet weather resulted in an 8.3 percent increase, USDA data show.
USDA crop-production estimates are made each month from August through January with the exception of December. The USDA’s National Agricultural Statistics Service surveys 11,000 to 27,000 farmers for each report, according to Joe Prusacki, the director of statistics at the agency. For most reports, two samples are also taken from about 1,900 fields.
“The USDA has an impossible task every year, and the last few years, the variability increased because of the weather problems,” Dale Durchholz, the senior market analyst for AgriVisor LLC in Bloomington, Illinois, said in a telephone interview. “People have to remember that the reports are just a sample of the total, and they have to anticipate there will be sampling errors based on weather conditions.”
In June, at the start of the hottest Midwest summer in 50 years, corn futures for July delivery touched a record $7.9975 a bushel on the Chicago Board of Trade. The grain tumbled 17 during that month as the government said U.S. farmers planted 1.8 percent more than analysts were expecting, and stockpiles as of June 1 were 12 percent higher.
Estimating grain inventories has become more difficult as corn producers increased on-farm storage and ethanol output rose to a record last year, boosting the use of ried distillers grain, a byproduct of biofuel production, in livestock feed, said Tim Emslie, the research manager for Country Hedging Inc. in Inver Grove Heights, Minnesota.
As of Dec. 1, 2010, farmers increased storage capacity for corn and other grain crops by 10 percent to 12.515 billion bushels since 2005, according to the USDA. U.S. ethanol production consumed a record 5.021 billion bushels of corn in the marketing year that ended Aug. 31, the first time more of the grain was used to produce fuel than to feed livestock and poultry, department data show
“The methodology may be the best they can do statistically over the long run, but actual physical conditions from year to year are going to drive how big the error is every year,” Emslie said. The USDA’s quarterly grain inventories “report is happening against a major logistical shift,” that is causing larger errors because of crop losses, increased farmer storage and rising ethanol production, Emslie said.