An analysis of commercial bank participation in the Farmer Mac II loan sale program

Murray, Charles
Journal Title
Journal ISSN
Volume Title
Research Projects
Organizational Units
Organizational Unit
Journal Issue

This dissertation's objective was to provide a descriptive and empirical analysis of commercial bank participation in the Farmer Mac II loan sale program for guaranteed portions of U.S. Department of Agriculture (USDA) Guaranteed Farm Loan Program loans. The descriptive analysis summarizes reasons for participation cited by bankers responding to a series of survey questions. Participants indicate the following factors as important in their decision to sell loans: enhanced liquidity, improved profitability, reduced interest rate risk, added capacity to meet heavy USDA guaranteed loan demand, and ability to pass on better loan rates and terms to their borrowers. Nonparticipants say loan sales are unnecessary because of weak USDA guaranteed and overall loan demand, sufficient deposit and capital levels to fund USDA guaranteed loans, and a preference to hold the loans they originate. In general, they do not sell guarantees to buyers other than Farmer Mac. The empirical analysis uses a logit regression analysis to predict the probability of a commercial bank participating and to identify the factors useful in making that prediction. Five models are estimated. The first determines the probability of a bank selling any type of USDA guarantee to Farmer Mac, be it a newly originated Farm Ownership (FO) or Operating Loan (OL) loan or a "seasoned" FO or OL loan. The other estimations look at participation by each loan type. Experience selling loans into other secondary markets always has a large positive effect on the probability of participating. Banks that hold a larger volume of USDA guaranteed loans in their portfolio also have a greater chance of participating in each estimation. Greater USDA guaranteed FO and OL loan demand and reduced competition among USDA FO and OL guarantee lenders increase the probability of selling new originations. These two variables are less effective in distinguishing between banks that sell "seasoned" loans and those that do not. No rule of thumb applies for the other independent variables' effects on the probability of selling newly originated or "seasoned" FO or OL loans. The reasons for selling FO and OL loans appear quite different aside from the variables discussed above.