CHICAGO (Reuters) - The U.S. Department of Agriculture has commissioned an industry study on the agency's estimate of U.S. stockpiles of corn to address widespread concerns about the accuracy of the bellwether quarterly report, USDA chief economist Joseph Glauber said on Thursday.
"The presumption from a lot of people is the survey is wrong," Glauber told Reuters on the sidelines of a grain industry meeting in Chicago. "I have a project ongoing right now looking at that issue. I contracted out some work."
The study focuses on why trade expectations have diverged so much from USDA's reported data, Glauber said. But he would not give details about the USDA-commissioned industry survey or when results would be available.
Glauber acknowledged that grain industry complaints of an inaccurate count for the largest U.S. crop had persisted for the last couple of years. But he expressed confidence that USDA was doing its best to track stockpiles in a rapidly changing corn marketplace, where ever more grain is being held "off-farm" and less subject to the department's farmer surveys.
Disconnects between trade expectations and USDA stock estimates have occurred off and on historically but the frequency has increased in recent years, analysts say. Since March 2010, 12 of the past 14 quarterly corn stocks figures have been well out of sync with private estimates, generating huge price swings in Chicago Board of Trade corn futures markets.
"When NASS does its survey they essentially do a census of all the commercial operations," Glauber said, referring to the USDA's National Agricultural Statistics Service which does the count.
"So you would think with more grain being held off-farm, those numbers should be even better. Whereas on the farm they do it by survey, so there are some inherent sampling errors. But there's been a lot of theories why those expectations vary," Glauber said. "I don't know that there will be an answer."
"What I've been looking at is trade expectations relative to the reports," said Glauber, adding there have been "surprises" between expectations and the actual numbers.
"There's no question that a lot has changed in the corn market over the last 5-6 years. Ethanol is now a major factor. Trying to estimate feed use is difficult. We know corn for feed has declined, while distiller's dried grain have come in. That's one issue," he said.
Ethanol, which now consumes up to 40 percent of the U.S. corn crop, produces a byproduct called distiller's dried grain (DDG) which in itself has become a massive new source of livestock feed for home and export markets.
"With the corn stocks, increasingly a larger portion is being held off farm - some for ethanol production," Glauber said. "These are trends. To explain the discrepancies, I haven't found a smoking gun yet."
Glauber said USDA is looking at every "theory" it could about why traders were questioning USDA's corn stocks.
"For example, we are seeing a lot more corn being grown in areas where it wasn't grown before. Years like last year, there's more corn coming on early - more corn planted in the South," Glauber said. "As a consequence, there are thoughts that some of the new crop could be fed in the fourth quarter, which would explain the discrepancies. However, that should balance out when you get to first quarter stocks for the next year. We have seen a little of that."
(Additional reporting by Tom Polansek; Editing by Marguerita Choy)