The risk-balancing hypothesis (RBH) suggests that farms will take less business risk as their financial risk increases, but existing literature provides empirical evidence that the RBH might be invalid under certain circumstances. We present a unified model that explains the conditions under which the RBH holds or is invalidated by recognizing the role of latent heterogeneity among farms. We generalize the RBH idea and trace the source of credit risk back to latent heterogeneity among farms. We then apply recent literature to longitudinal data from a panel of Dutch farms and classify segments using a finite mixture regression fixed-effects model and find that the RBH may not apply to all groups in the same way.