USDA Forecasts: A meta-analysis study

TL;DR

This meta-analysis examines the accuracy and market impact of USDA forecasts, revealing heterogeneity in study results but little evidence of publication bias, and provides insights into their bias, efficiency, and market relevance.

Contribution

It systematically synthesizes existing research on USDA forecast accuracy and market reactions, highlighting heterogeneity and assessing publication bias.

Findings

Mixed evidence on forecast bias and efficiency

Most studies find USDA forecasts influence markets

Some heterogeneity in study results

Abstract

The primary goal of this study is doing a meta-analysis research on two groups of published studies. First, the ones that focus on the evaluation of the United States Department of Agriculture (USDA) forecasts and second, the ones that evaluate the market reactions to the USDA forecasts. We investigate four questions. 1) How the studies evaluate the accuracy of the USDA forecasts? 2) How they evaluate the market reactions to the USDA forecasts? 3) Is there any heterogeneity in the results of the mentioned studies? 4) Is there any publication bias? About the first question, while some researchers argue that the forecasts are unbiased, most of them maintain that they are biased, inefficient, not optimal, or not rational. About the second question, while a few studies claim that the forecasts are not newsworthy, most of them maintain that they are newsworthy, provide useful information, and cause market reactions. About the third and the fourth questions, based on our findings, there are some clues that the results of the studies are heterogeneous, but we didn't find enough evidences of publication bias.