Three essays on patronage refunds and capital structure of farm credit system associations

The objective of this dissertation is to investigate the patronage refunds decision in the Farm Credit System (FCS) lending associations, and simulate the capital structure change of a representative association dynamically. Patronage refunds distribution is a unique feature of financial cooperative organization and has important impact on the capital structure of FCS association. First, I utilize a logit model to analyze the decision of paying patronage refunds versus not paying patronage refunds. Increased loan size, higher capital ratio and larger interest margin are found to significantly increase the probability of paying patronage refunds. The bank district in which an association belongs also matters. Second, I examine the decision of how much patronage refunds is distributed in panel tobit models. Capital ratio and interest margin are still significant factors. Moreover credit risk and profitability play important roles in determining the patronage refunds amount. Further, my estimations evidence the bank district diversity and the temporal persistency in patronage refunds. Finally, I employ the System Dynamics modeling techniques and conduct a simulation analysis on the changing capital structure according to the decisions of patronage refunds, new loans issued, and interest rate charged on loans. Risk uncertainty in interest rate and loan default are particularly discussed. My simulation outcome supports the FCS association’s practice of building up capital reserves, yet warns the potential high volatility of loan interest rate when associations face risk shocks. For the first time, the research uses a panel data set that includes all FCS associations to conduct an empirical analysis. It offers new insights on what drives patronage refunds distribution and capital structure management in FCS associations.