Understanding the impact of late season precipitation on the distribution of cotton yields provides insight into managing yield risks. This research combines Linear Moment Models with historical weather data to assess the impact of late season precipitation extremes on cotton production and revenue. The empirical analysis suggests that late season drought reduces both mean yield and variance. The shift in variance is coupled with an exchange of upside risk for downside risk implying that the variance reduction alone masks an important effect on producer’s risk. Revenue impacts suggest high revenue for irrigated acreage as compared to dryland acreage, and the late season drought impact on revenue shows that the use of irrigation causes increases in revenue as compared to dryland acreage.

The federal crop insurance program provides crop insurance for millions of acres and many commodities every year. The Risk Management Agency of the USDA is responsible for determining the premium rates for these covered commodities. Currently, the quality of soil is not considered when determining baseline yields and expected premium rates. This study utilizes the moment-based maximum entropy method to assess the effect of incorporating soil in the rate making methodology. Several moments of upland cotton yield in Arkansas, Mississippi, and Texas are conditioned on weather, irrigation, and soil control variables. Ultimately, I find evidence of mispriced premium rates for counties in all three states for both irrigated and non-irrigated upland cotton yield.