The feasibility of securitizing disaster risk in agriculture using catastrophe bonds
MetadataShow full item record
This study makes an initial step towards designing catastrophe (CAT) bond products for agriculture making possible the transfer of natural disaster risks from insurance companies to investors/speculators in the global capital market. The CAT bond contracts were created in a typical agricultural setting - in this case, a setting where a crop insurance company writes crop insurance policies for cotton and peanuts in Georgia. The CAT bonds were priced based on state-level average yield data. Trigger density estimation determines the accuracy of the probability of the CAT bond being triggered and thus affects CAT bond prices. The proposed index triggers in the agricultural CAT bond contracts are based on percent deviation from state average yield. Yield detrending plays a crucial role in computing yield loss and therefore estimating trigger densities. The spline modeling procedure makes it possible to capture the stochastic components in the yield data for more accurate kernel density estimation. Quartic kernel density estimation with suitable bandwidth gives an appropriate curve for generating accurate probabilities of CAT bond triggering. The principal finding of the study is that it is feasible for crop insurance companies to issue insurance-linked securities. CAT bonds can be used in hedging catastrophic risk effectively in agriculture given the common practice that crop insurance companies seek to minimize the variance of the loss ratio. CAT bonds can reduce the variance of the loss ratio when issued optimally with respect to the number of bonds and contract specifications. CAT bonds were found to be feasible even in the range of normal losses commonly covered by crop insurance and reinsurance. Thus, the usefulness of CAT bonds in agriculture has been shown to be much greater than anticipated.