Decomposing USDA ending stocks forecast errors

The U.S. Department of Agriculture (USDA) publishes monthly Ending Stocks projections, providing an estimate of the end-of-marketing-year inventory of a particular commodity, which effectively summarizes its supply and demand outlook. By comparing USDA’s projections of balance sheet variables against their realized values from marketing years 1992/3 to 2019/20, we decompose ending stocks forecast errors into errors of the other supply and demand components. We apply a decision-tree-based ensemble Machine Learning (ML) algorithm, the Extreme Gradient Boost Tree (EGBT), that uses a gradient boosting framework and is robust to multicollinearity. Our results indicate that export and production misses are the key contributors to ending stocks projection errors. Because foreign imports of U.S. products are likely tied to foreign production deficits, we likewise investigate how U.S. export errors are linked to USDA’s foreign production and export forecast misses, country-by-country, and show that misses on production and export levels in China, Mexico, Brazil, and European Union cost USDA the most. Overall, our results make a strong case that better information about production expectations, both domestically and worldwide, will contribute to more efficient agricultural balance sheet forecasts.