Credit risk migration analysis of cooperatives

Abstract

Cooperatives struggled financially in the early 2000s because profitability declined. In 2002, Agway and Farmland Industries filed for Chapter 11 bankruptcy. Since then, bankruptcy events have been few, and cooperative profitability has rebounded. Credit risk is of particular importance to lenders, managers, and directors. Lenders are primarily concerned about their counterparty exposure to cooperatives while cooperatives are concerned about meeting their debt obligations and distributing patronage to their members. Conceptualizing the expected credit risk behavior of cooperatives over time and responses to exogenous favorable and adverse factors offer useful insight to cooperative managers. The objective of this thesis is to assess the historical evolution of agricultural cooperative credit risk. This will be done by studying the credit rating migration behavior of cooperatives using Markov chains. This research uses proprietary cooperative financial statement data of 155 cooperatives spanning from 1996 to 2014. The Credit Metrics component of Moody’s Global Agricultural Cooperatives Industry Rating Methodology is used to assign annual credit ratings. Unconditional transition probabilities and probabilities conditioned on the type of cooperative, the state of the economy, and the performance of the agricultural sector are estimated to compare differences in migration behavior. Non-Markovian behavior and time-heterogeneity are also examined. Cooperatives are more likely to experience a change in credit rating in the next period than remain unchanged. When the direction of the rating change is considered, cooperatives are more likely to be upgraded than downgraded. However, more annual instances of downgrades exceeding upgrades are observed across the sample. There are numerous instances where downgrades and upgrades span multiple rating classes. This occurs because cooperatives are often small and not diversified enough to withstand adverse challenges to their operations. Rating history impacts the direction of the subsequent rating change. Cooperatives that were previously unchanged in the prior period are more likely to retain their rating in the next, while those previously upgraded are more likely to be downgraded, and those previously downgraded are more likely to be upgraded. We find that the rating process does not follow a first-order Markov chain.