Correlated Losses in Federal Crop Insurance
Jonathon Siegle, Victor Gardner
The USDA keeps public data on the federal crop insurance program, and serves as a trove of agricultural data, albiet with messy inter-relations and endogeneity problems. Cross-crop correlations in losses are useful to investigate for both portfolio and actuarial purposes. This project explores models to find key explanatory variables, including crop pairs with correlated indemnities.