Modeling debt choice in agriculture: the effect of endogenous asset values

Purpose The purpose of this paper is to investigate the endogeneity of asset values and how it relates to farm financial stress in US agriculture. The authors conceptualize an implied measure of farm financial stress as a function of debt position. The authors posit that there are variations in the asset values that are beyond the farmer’s control and therefore have implications on farm debt. Design/methodology/approach The framework recognizes the endogeneity of return on assets (ROA). It uses a non-parametric technique to approximate the variance of expected ROA (VEROA). The authors model the rate of return on agricultural assets and interest rate with a formulation that focuses on macroeconomic policy. Further, the authors use a dynamic balanced panel data set from 1960 to 2011 for 15 US agricultural states from the Agricultural Resource Management Survey, and information from traditional state-level financial statements. Findings Estimation of linear dynamic debt panel data models accounting for the endogeneity of ROA and VEROA is a challenging task. Estimated variances are unstable. Hence, the authors focus on variance specification that uses the residuals squared from the ARIMA specification and non-parametric estimators. Arellano-Bover/Blundell-Bond generalized method of moments estimation procedures, although may be biased, show that VEROA has a negative and significant effect on the total amount of debt in the agricultural sector. Research limitations/implications The instruments used in this analysis are lagged regressors which may be weakly correlated with the relevant first-order condition, hence not properly identifying the parameters of interest. Future research could include the identification of better instruments, potentially use of sequential moment conditions. Originality/value Unlike previous study, the authors use non-parametric approximation of VEROA. The authors model the rate of return on agricultural assets and interest rate with a formulation that focuses on macroeconomic policy. Second, the authors make use of a large dynamic balanced panel data set from 1960 to 2011 for 15 agricultural states in the USA. To the best of the authors’ knowledge, this study is one of the few that provides evidence on risk-balancing behavior at the agricultural sector level, of the USA.