Report accuracy takes center stage
The accuracy of USDA corn reports has become a hot topic during the past two years as USDA and trade numbers have varied considerably, sending commodity prices on a wild ride.
"When the corn inventory numbers are off 200 million to 300 million bushels [from trade expectations], it has a dramatic impact," says Peter Meyer, publisher of Opening Print. Such divergences have created a high level of market volatility, in addition to questions about the validity of USDA grain stocks numbers.
For example, Meyer doesn’t think USDA’s estimate of an 850 million bushel carryout was on target. Furthermore, he questions whether the difference between old crop and new crop reflects reality. "[Surveyed] elevators don’t really differentiate between old crop and new crop," he says. Commingling old crop and new crop became a big issue in 2010.
"The level of frustration with government grain stocks reports is very high," echoes Dan Cekander of Newedge USA. He challenges USDA numbers on ethanol, which he says can cause up to a 200 million bushel difference in supply. He does not take issue with production reports from USDA’s National Agricultural Statistics Service (NASS), however.
When it comes to predicting yields, Meyer is concerned that USDA does not include the past few years of lower than trend-line yields, thus setting expectations too high in the projections made at USDA’s Ag Outlook Forum. NASS forecasts are based on survey results, not trends.
Despite the frustrations, the trade sector realizes there’s still value in government reports, illustrated by the fact that one of USDA’s January reports created limit moves six times in a row. The cattle feeding number was 2% below what the monthly report showed, Meyer explains, and as a result, it created higher corn stocks than the trade was looking for, moving prices lower once the news hit trading pits.
USDA defends its process. To tally the August crop report, which includes the first yield estimate for row crops, the department counts stalks and ears in its 5,000 field plots in various production areas and surveys 27,000 farmers about their yields, says Joseph Prusacki, director of statistics for NASS.
The Secretary of Agriculture does not participate in the forecast and estimate process. He or his representative signs the report and is briefed later, Prusacki says.
One question USDA often receives is why it doesn’t use more satellite data to compile its production estimates. Forecasting yields using satellites is a bit tricky, Prusacki says. "Satellites can only give us planted area and growth, but they can’t see harvested area or tell how much crop is in bins. On soybeans, furthermore, there is no correlation between growth and yield."
- March 2012