WASDE: Market impact and accuracy; Oh, really?
June 12 will be the next time USDA issues its World Agricultural Supply and Demand Estimate (WASDE), and the market is already anticipating what will be revealed in that report. And, in fact, the market is already issuing projections of what the USDA will report on such issues as corn and soybean production and carryover. While a recent USDA study says WASDE reports can move the market by substantial amounts on days they are issued, another economic study suggests that the accuracy of the information contained in WASDE reports is not always 100%.
Between the 8th and the 12th of each month, USDA staff members within the World Agricultural Outlook Board chair an interagency committee of specialists who assemble domestic and foreign production and consumption information on major crops, and release the information to a market that is ready to respond either bullishly or bearishly to the data. According to a USDA study published in the recent Amber Waves electronic newsletter, USDA’s Economics Research Service (ERS) looked at 350 WASDE reports and resulting commodity market responses from 1981 to 2012.
The report said, “ERS estimates that, following WASDE’s publication, the average report causes an immediate price change of about $190 per contract for cotton and soybeans and almost $140 per contract for hard red winter wheat at recent prices. These price changes, which correspond to a 1-day return on collateral of around +5 percent to a cotton and wheat trader and +8 percent to a soybean trader, signify that WASDE updates market expectations and is therefore informative.” To quantify the impact, the ERS study said, “In 2009, cotton and wheat traders held on average 100,000 contracts on report days; soybean traders held four times as many.”
In addition to the USDA study’s contention that the report had information that was of value to the market, the report also concluded, “Overall, findings reveal that markets value the situation and outlook information published in WASDE and rapidly incorporate that information into futures prices. Reports that included NASS crop survey data are very informative but so are most reports that do not contain production information for row crops. Lastly, WASDE affects multiple simultaneously traded contracts for the same commodity about equally, and the report’s impact increases during low-inventory periods.”
With the quantification of how the WASDE reports impact the market, and particularly in times when commodity levels are diminished, the accuracy of the WASDE estimates is being called into question. A trio of economists from Clemson University (Olga Isengildina-Massa), the University of Georgia (Berna Karali) and the University of Illinois (Scott H. Irwin) studied the WASDE reports from the 1987/88 crop year to 2009/20102, and agreed with the USDA study about market impact, by saying, “It has been widely documented that the release of many USDA reports moves the markets…volatility in corn and soybean futures markets increases about 7 times on report release days.”
Their analysis is that WASDE reports are not always 100% accurate, but they also point to the three reasons they found for the errors.
Tendencies of USDA forecast staff members:
The university economist say, “The first one deals with forecasters’ abilities and reflects behavioral characteristics and tendencies…Thus, forecasting behavior may be different with respect to predicting positive vs. negative change and extreme vs. moderate rate of change.” The economists say there is a personal reaction built into the estimation process that may have a positive or negative reaction to the numbers being calculated. And they surmise, “..reaction to positive vs. negative change in WASDE forecasts would likely be due to mis-calibration of forecasters’ reaction to information and can result in optimism as well as pessimism. Given the public nature of the forecasts and their ability to affect markets, these biases should be unintended.”
They found that, “corn and soybean forecasters tend to overcorrect previous mistakes. For example, a 1% overestimation in corn production in the previous year is likely to be followed by a 0.3% underestimation in current year. On the other hand, wheat forecasters tend to repeat their previous errors. Thus a 1% overestimation in wheat production in the previous year is likely to be followed by a 0.2% overestimation in current year.”
Incorporation of macro-economic data:
According to the university-based study, “The second source reflects data that are not (efficiently) included in the forecasts. Previous studies showed that USDA forecasts do not encompass simple time-series models. In our study we investigate whether USDA forecasts fail to efficiently take into account macroeconomic data and tend to show mistakes during certain economic conditions, which has not been analyzed in previous studies of these forecasts.” They say, “In our study we investigate whether USDA forecasts fail to efficiently take into account macroeconomic data and tend to show mistakes during certain economic conditions.” For example, they point to the economic changes in commodity value and production resulting from changes in policy from one Farm Bill to the next. That has pointed to more corn acreage and less wheat and soybean acres, which they contend the USDA forecasters did not efficiently incorporate into their estimates.
They discovered that not all of the macro-economic factors in the marketplace are efficiently incorporated in the WASDE reports—particularly exchange rate, oil price-related data, and inflation:
- “USDA forecasters appear to overestimate corn and soybean price and underestimate use (corn feed and residual and soybean exports) as well as supply (soybean production) during the periods of increasing oil prices.”
- And regarding the errors related to exchange rates, they said “Appreciation of U.S. exchange rate is associated with overestimation of corn and soybean price, underestimation of some domestic use categories (corn feed and residual, soybean crushings, and seed and residual, wheat food and feed and residual), underestimation in wheat ending stocks, and overestimation in wheat exports and seed use.”
- Higher inflation is associated with underestimation of corn ending stocks and wheat price, overestimation of wheat feed and residual use, as well as overestimation of soybean production and seed and residual use. Increasing tendencies for underestimation of corn exports and ending stocks, soybean production and wheat food, seed and ending stocks are detected during the Renewable Fuel Act (RFA) period (2005-2010), while wheat feed and residual use is increasingly overestimated during this time.”
Data use errors:
The agricultural economists say, “The third source of forecasting errors may stem from the data forecasters use. Most fundamental models of commodity prices rely on uncertain estimates of independent variables, and errors in those variables may cause errors in price and ending stocks forecasts.” They pointed to a pecan export forecast that was in error because of data processing errors. And since the market’s attention is fixed on carryover stocks, the economists report, “Due to the residual nature of ending stocks (total supply less total use), errors in ending stocks forecasts can be tracked down to their sources by regressing ending stocks forecast errors against the errors in all supply and use categories.”
They found that, “Analyses of data-related sources of error reveal that errors in ending stocks forecasts are mostly driven by errors in production forecasts across all commodities. Among use categories, corn feed and residual, soybean crushing, and wheat export errors have the biggest impact on endings stocks errors. Errors in price forecasts are caused by errors in U.S. ending stocks forecasts for all commodities and total use forecasts for soybeans and wheat.”
The ag economists studied all of WASDE’s reports on corn, soybeans, and wheat which comprise 90% of total US grain and oilseed production, and they added, “Knowledge of supply and demand forecasts accuracy is important because these categories serve as building blocks for price forecasts. Furthermore, supply and demand estimates are published within a set of other forecasts in WASDE reports that have been shown to affect the markets.”
So, with the next WASDE report being released June 12, the university ag economists say their findings can be used by the market to interpret the USDA data, if the market is fully aware of the flaws and inefficiencies in USDA forecasts or adjust for those.
USDA’s monthly reports on World Agricultural Supply and Demand Estimates (WASDE) have been shown to have an impact on the commodity market because of the importance of the information. However, the reports also have some accuracy issues, based on the personal biases of estimators, their incorporation of macro-economic data, and data processing errors. If the market is aware of those errors, then adjustments can be made.