Does USDA Information Have Value?
Does USDA Information Have Value?
Another group of studies, shown in table 7, applied informational value tests to assess whether information released in the reports gets us closer to knowing the final estimate, thereby reducing uncertainty in the markets. Several of these studies (e.g., Garcia et al., 1997; Isengildina-Massa, Karali, and Irwin (2020) implemented the test of informational value developed by Baur and Orazem (1994), focusing on the reduction in the market’s supply forecast variance resulting from the introduction of the government forecast. For example, Garcia et al. (1997) found that the largest reduction in corn supply forecast variance was associated with August crop production report with subsequent crop production reports having only marginal value. Even though the relative accuracy of USDA and private forecasts in this study was similar, a significant price reaction to USDA reports, implied that USDA forecasts are perceived as less risky and more objective by market participants. Isengildina-Massa, Karali, and Irwin (2020) confirmed the importance of August crop production report in reducing corn market’s supply forecast variance, but also showed an increasing value of October crop production reports in post 2002 subperiod. In addition, these authors argued that Prospective Plantings and Acreage reports play a much bigger role in reducing supply uncertainty than Crop Production reports since 1983. Furthermore, these authors find that the information value of USDA forecasts has increased over time and was the strongest in the most recent 2002-2019 subperiod.
Another approach to demonstrating informational value is by assessing whether advanced knowledge of the information in these reports would allow traders to correctly position themselves in the market. For example, Garcia et al (1997) use the test developed by Henriksson and Merton (1981) to demonstrate that traders can correctly determine market direction based on the information contained in corn and soybean production reports over 1971-1992, thus indicating that these reports have market timing value. McKenzie (2008) used a Hamilton-type (1992) approach to demonstrate that “there were periods when having advanced knowledge of the August report would have significantly adjusted rational agent expectations, augmenting information already embodied in futures prices. (p. 365)” Milacek and Brorsen (2017) developed trading models based on knowing the WASDE report in advance to estimate potential trading returns from using WASDE report predictions in the days before the report. Their findings reveal that the perfect foresight trading signal generated an average daily return of 1.11 cents per bushel for corn with July, September and October reports generating the highest returns.
McKenzie and Singh (2011) demonstrated that hedging stored grain over USDA report days is extremely important and beneficial from a risk management perspective to reduce potential losses due to large price movements on report days. Even though the volatility of basis was statistically highest on report days, hedging does lead to lower losses than storing unhedged cash grain, as shown in figure 12. Additionally, Abbot, Boussios and Lowenberg-DeBoer (2016) used dynamic multi-period Monte Carlo simulation of the inventory adjustment models for the U.S. corn market to estimate the value of WASDE reports and its components. Their results show significant value to market participants from the WASDE reports, roughly $301 million or 0.55% of overall corn market value. The results also show significant value for each forecasted component of the reports: area ($145 million), yield ($188 million), production ($299 million), demand/stocks ($300 million) and exports ($320 million).