Three Essays on Agricultural Insurance and Farm Real Estate Investment

The subject of this dissertation includes studies on agricultural insurance and farm real estate investment. Portfolio risk in crop insurance due to the systemic nature of crop yield losses has inhibited the development of private crop insurance markets. Government subsidy or reinsurance has therefore been used to support crop insurance programs. The first essay investigates the possibility of converting systemic crop yield risk into "poolable" risk. The results indicate that the systemic risk in crop insurance can be eliminated by combining crop insurance policies across crops and countries. The second essay investigates farmland portfolio returns from a forward-looking perspective taking into account time-varying return and serial correlation. The results indicate that it takes a number of years for the expected return to reach the long-term equilibrium. From a forward-looking perspective, the attractive average return level observed historically can only be attained over a long investment period. The risk involved in the long investment period, however, is also considerably higher than the historical sample volatility. The third essay examines the predictive power of macroeconomic risk factors for farmland asset returns. Farmland value slightly increased in 2017 even though farm income was lower. This development suggests the rate of return required by investors for farmland asset has been reduced. One possible explanation for the decreasing required rate of return is an increased money supply. Previous research suggests that the money supply affects several macroeconomic risk factors through different transmission channels, which in turn influence investor behaviors and asset returns. Both linear and neural network models are used in this study to predict farmland returns. The forecast accuracy is compared across different models. The results indicate that farmland return prediction is significantly improved by adding capital market excess return as an explanatory variable. Adding additional risk factors, however, does not improve the prediction with the sample used in this study.