Stress Testing Projected Capitalized Farmland Values

Abstract

This study initially presents historical trends in both the capitalized value and market value of farmland in the eight states comprising the Corn Belt and Lake States production regions as defined by the USDA. An econometric analysis of annual real cash rents per acre prior to determining the capitalized value of farmland in the eight states is then conducted. Two distributed lag models were hypothesized. The comparison of regression results of these two distributed lag models indicates that current year real cash rent can be best explained by current year real net farm income, lagged real net farm income over a period of years, and real cash rent in the previous year. A spreadsheet simulation model is used to project capitalized farmland values in each state as well as regional averages over the 2012-2015 period. These projections reflect alternative assumptions regarding future trends in real net farm income at the state level as well as the rate on 10-year constant maturity U.S. government bonds to assess the potential sensitivity of capitalized farmland values under adverse economic conditions. The projected trends in capitalized farmland values under two alternative stress scenarios reflecting higher interest rates levels and lower net farm income levels indicates that capitalized farmland values are particularly sensitive to interest rate fluctuations since cash rent expectations of landlords are based on current and lagged historical profit performance.