Credit Risk and Financial Performance Assessment of Illinois Farmers: A Comparison of Approaches with Farm Accounting Data

Pro forma financial performance evaluation of agricultural producers is an important issue for lenders, internal management and policy makers. Lenders strive to improve their credit risk management. Internal management is interested in understanding the financial impacts of alternative strategic decisions. And policy makers often assess the magnitude and distributional effects of alternative policies on the future financial performance of farm business. Data limitations are a major impediment in assessing farm financial performance. Most traditional farm operations are private firms and thus, public traded equity information which can be converted into market valuation change is not available. Moreover, historical loan performance data on agricultural loans such as past due and defaults are not readily available. These aspects present substantial methodological issues when establishing an independent variable to use in assessing future performance. Credit risk modeling and financial performance assessment have been remotivated and gained unprecedented academic attention in recent years(Barry 2001, Kachova and Barry 2005, Saunders and Allen 2002). However, some of the new approaches and models have limitations when applied to agricultural producers. Adapting the models and approaches to utilize the available information of farm business needs careful attention and validation. In this paper, Altman’'s Z' score model and Z" score model are applied to farm accounting data for the detection of farm operating and financial difficulties. i.e., farms with high credit risk. The well-developed and widely used Altman models have not been applied to agricultural data. The results are compared to an experienced based credit risk migration model (Splett, et al) and a logistic, lender-based model (Featherstone, Roessler, and Barry 2006). The experience based model is a primary model used in the current farm credit analysis. The logistic model is claimed for better statistical prediction accuracy and no binding assumption on multivariate normality (Altman 1968). Results from each of these models are compared across a common database of Midwestern grain farms.Farms are grouped into different categories with different levels of financial. Instead of focusing on farm loan defaults, earned net worth growth rate (ENWGR) and term debt coverage ratio (TDCR) are used as two major indicators for financial stress situation of farm credit quality.