ABSTRACT A common assumption is that multiple‐peril crop insurance markets suffer from market failures, thus justifying government intervention in the form of premium subsidies, operating allowances, and reinsurance agreements. One prominent rationale for intervention involves geographic correlation in agricultural production which leads to systemic risk in crop insurance portfolios. We measure the degree of systemic risk—and evaluate the effectiveness of risk pooling—in a hypothetical portfolio of insurance policies for U.S. corn and soybeans. We model dependence using vine copulas that capture potential asymmetries, tail dependence, and nonlinear associations. Our results indicate a reduction in overall risk when policies are pooled across space, decreased capital per policy held by the insurer to prevent ruin, and weakened tail dependence at moderate distances. Although the portfolio is subject to spatial dependence, systemic risk is unlikely to be the main impediment to market (i.e., private) crop insurance.
Fan et al. (Sun,) studied this question.