Disagreement in Farmland Value Expectations

The growth rate of the value of farmland is important to the agricultural sector. Real estate comprises 83% of farm sector assets, as well as 68% of farm sector debt (USDA, 2021). Farm real estate plays a large role in both sides of the accounting equation and land values – especially expected future land values – play a significant role in lending decisions. Evaluating these future land value expectations is the topic of this study. In the US, several organizations use surveys to elicit farmland experts’ expectations of farmland value. These expectations are presented in the aggregate, obscuring the potential underlying heterogeneity in the expectation formation process. Kuethe and Hubbs (2017) found agricultural lenders’ expectations are unbiased yet inefficient, and recently, Kuethe and Oppedahl (2020) found agricultural lenders’ expectations are conservatively biased. This study uses an expectation evaluation methodology from Davies and Lahiri (1995) and a newly-created panel of Indiana farmland experts from the Purdue Land Value and Cash Rent Survey from 2003-2022 to model heterogeneity in farmland value expectations. We find evidence of survey-wide under-prediction by farmland experts, consistent with Kuethe and Oppedahl (2020). In addition, we compare the future price expectations of lenders and appraisers, which may introduce friction in forming lending relationships. In addition, a key contribution of this study is the creation of a true panel dataset from past Purdue Land Value and Cash Rent Surveyresponses. The novel dataset may allow for future research to explore questions not previously possible, in absence of a true panel dataset