Three essays on systemic risk and rating in crop insurance markets

This dissertation investigates several issues surrounding systemic risk problems in crop insurance markets. The first essay deals with the important issue of spatial aggregation and crop yield risk in weather derivative hedging. Previous studies identify limited potential efficacy of weather derivatives in hedging agricultural exposures. In contrast to earlier studies which investigate the problem at low levels of aggregation, this study finds that better weather hedging opportunities may exist at higher levels of spatial aggregation. Aggregating production exposures reduces idiosyncratic risk, leaving a greater proportion of the total risk in the form of systemic weather risk which can be effectively hedged using relatively simple weather derivatives. The aggregation effect suggests that the potential for weather derivatives in agriculture may be greater than previously thought, particularly for aggregators of risk.;The second essay investigates the loss performance of the Federal Crop Insurance program. Historically, government insurance programs tend to be ineffective at segregating risks, leading to markets that are inefficient. In the case of the Federal Crop Insurance program, rates are set non-competitively. This study develops a spatial econometric model of loss experience and finds evidence of geographic misratings. The results also suggest that substantial actuarial cross-subsidization is resulting from the apparent rating inequities, which has a variety of welfare implications.;The third essay examines the actuarial implications of the Loss Cost Ratio (LCR) ratemaking methodology employed by the Federal Government for setting base rates for the program, and identifies specific conditions required for the LCR methodology to result in unbiased rates. Specifically, constant relative risk through time and other restrictive requirements are required. These requirements are tested against a unique farm-level data set. The results indicate that the conditions required for the LCR to produce unbiased rates are violated for a large number of actual farmers. Implications include that the current ratemaking produces biased rates that are 75%-180% in excess of actuarially fair premiums in Illinois. A simple correction function is proposed and illustrated.